Time delay compensation of distributed multiple microphones in

Time delay compensation of distributed multiple microphones in
Time delay compensation
of
distributed multiple microphones
in recording
-an experimental evaluation
Theresa Ann Leonard
Department of Theory
Faculty of Music
McGiII University, Montreal
November, 1991
A Thesis Submitted to the Faculty of Graduate Studies and Research in partial
fulfillment of the requirements for the degree of
Master of Music.
J
© Theresa Leonard, 1991
T ABLE OF CONTENTS
Acknowledgements ...................................................................... ili
. 0 fD'lagrams ............... '" " ..... , .................. '" .......................... IV
.
LIst
Abst:ïd.ct
............................................................................. V
Résumé
............................................................................. VI
1 OVERVIEW OF CLASSICAL MUSIC RECORDING PRACfICE: THE NEED
FOR BLENDING MAIN MICROPHONES WITII SPOT MICROPHONES
1.1
Classical Music Recording Considerations ................................... !
1.2 Spot Microphone Considerations ............................................. 5
1.3 Negative Effects of Spot Miking ............... . .. ........ . ................. 6
1.3.1 Acoustic Phase Cancellation ............................................. 7
1.3.2 Corn b- Filter Effects ....................................................... 8
1.3.3
Microphone Off-Axis Coloration ....................................... 8
1.4 Focus of Work .................................................................. 9
2 MIXING WITH DELA Y COMPENSATION
2.1 Time I:>elay Mixing ............................................................ Il
2.2
Previous Work in This Area ................................................. 12
2.3 Delayed Spot Microphones ................................................... 13
3 OVERVIEW OF FACfORS CONTRffiUTING TO GREA1ER
UNDERSTANDING OF AUDffiLE EFFECTS OF DELAY COMPENSATION
3.1 Psychoacoustic Comiderations ............................................. 16
3.2
The Role of First Reflections ............................................... 17
3.3 The Perception of Depth In Recording ...................................... 19
3.4
l
Musical Instrument Radiation Patterns ...................................... 21
l
4 RESEARCii PERFORMED: EXPERIMENTAL METHODS AND
IMPLEMENTATION
4.1 Procedure ..................................................................... 24
4.2 Methodology ................................................................... 25
4.2.1 A Calcuhttion Model .................................................... 2H
4.2.2 Temperature and Humidity Effects on the Speed of Sound. . ..... 29
4.3 Listening Test Recordings .................................................... 30
5 LISTENING TEST RESULTS .................................................... 40
5.1 Analysis of Data .............................................................. 40
5.2 Imerpretation ofComments ................................................... 43
6 SUMMARY AND DISCUSSION ............................................... 48
6.1 'Time Delay Recording Technique: A Tooi For Balancing ................ 48
6.2
Potential Problems
............................................................ 49
6.3 Future Considerations
........................................................ 52
7 BIBLlOORAPHY .................................................................. 56
8 APPENDIX ......................................................................... 59
u
ACKNOWLEDGEMENTS
1 would like
10
thank my advisor, Dr. Wieslaw Woszczyk, for his inspiration and
advice in begmning this research and for taking the time 10 keep in contact with me from a
distance.
1 also appn!Clate the efforts of Kevin Elliou for his helpful guidance through the final
stages of this work and for his understanding 10 giving me time to complete this papl~r
while at The Banff Centre.
T'hank you 10 all who took part in the Listening Tests.
To my mother for her continuous support of my work, 1 dedicate this thesis.
iii
LIST OF DIAGRAM~,
1
1. McGill Symphony Orchestra: Pollack concert hall; January 25,1989 ......... 32
2. McGill Symphony Orchestra: Pollack concert hall; April 1,1989 ............. 34
3. The Banff Centre: Project Studio Recording; December 2, 1989 ............... 36
4. The Banff Centre: Project Studio Recording; January 29, 1990 ............... 39
5. Potential effect of reverb sound with spot microphones ........................ 50
iv
ABSTRACT
In the search for improved reproduction of a classical music perfonnance, the technique
of time delay compensation is shown to he a useful 1001 for balance in recording. ThIs
paper investigates the importance and val1dIty of small rime adju),lments in recordIngs to
compensate for vanaoon 10 distance between spot microphones and a mam stereo pair.
Conventional recording techniques, psychoacoustic considerations and lechnological
aspects of the use of delays are researched in order to determine therr validity in timbrai
Improvement.
Multiple microphone set-ups are used ta record hoth large orchestral works and
smaller-scale cla'lsical compositions where auxiliary microphones may be net"Aied to ensure
an optimum balance in the final mlx. Small Ume delays <Ut! denved from calculations
involving the distances between microphones, the speed of sound, and humidity and
temperature readIngs from the hall. Proper synchronization of these delays is deslrable w
preserve phase coherence and combat comb-filter effects. Precise delay units are used to
compile musical excerpls for listening tests.
The results reveal any change in sound quality and provide a basis for investigating botJ-.
the positive and negative effects through objective study of the value of time delay
compensation in the live recording reproduction of classical music perfonnances.
1
v
RESUME
Dans le ca~ de recherches sur la façon d'améliorer la reproduction de la musiqut~
classique, la te:chnique de l'égalisation de temponsauon revêt une gmnde ullhté pour
l'équilibre de l'enregistrement. Dans cet arucle. l'auteur étudie l'Importance et la valeur d.?'
petits ajustements temporels visant à com~nser l'écan de distance entre le!'> mlcropholle . .
isolés et la pJincipale paire stéréophomque L'auteur analyse les techniques
d'enregistrement classiques, les paramètres psych~acoustiques et
lc~
aspects
technologIques de l'emploi des temporisatiom pour déterminer leur valeur dans
['amélioraüon du tImbre.
La mif~e en place de microphones multiples sen à enregistrer les oeuvres pour grand
orchestI'l~ et
les compositIons claSSiques pour formanon plus restremte où Il
~e peut
qu'on ait besoin de microphones auxilialre~ pour assurer un meilleur équIlibre dan .. le
mixage final. Les petits retards temporels sont urés des calculs
~ur
la dl'ltance cntre lc'i
microphones. la vItesse du son et les relevés d'humidité et de température dan~ la 'Ialle
La bonne synchromsation de ces retards est souhaitable pour préserver la cohérence de
phase et lutter contre les effets des filtres en peigne. Les dlsposmf'l de tempon'iatlon
précis servent à compIler des extraits musicaux pour les es'mls d'écoute.
Les résultat" révèlent le moindre changement dans la qualIté du 'Ion et 'icrvent de
fondement à l'étude des effets positifs et négatifs par l'étude objective de la valeur de
l'égalisation de temporisation pour l'enregistrement en direct de concens de musique
classique.
vi
Chapter 1
OVERVIEW OF CI.ASSICAI. MUSIC RECORDING PRACTICE: THE NEED
FOR BLENDING MAIN MICROPHONES WITH SPOT MICROPHONES
1.1 Classical Music Recording Considerations
The recording of a classical music perfonnance is generally most successful when
Ibtcners are least aware of any signal processing by the recording technicians. There are
argument!'> as to whether the reproduced sound should be as close as possible a replica of
the perfonnance, or whether the recording engineer should improve on the performance
where feaslble via si? 'al processmg.
AIthough there is no speclfic recording technique for classical music, the microphone
technIque wIll have a great effect on the reproduced sound. The choice of a particular
[echmque wIll depend on vanous Clfcumstances. Three basiC philosophies developed
regarding microphone placement Wlth the mtroduction of stereophonie recording: 1)
l'oinCldent stereo,
~)
spaced omni, and 3) multiple microphones. Each technique has its
own uses in the recording studio, sorne with advantages over others 11. certrun sItuations,
'\uch as a lIve event when the room
whit:h
lhe~e
IS
not farniliar. There are three basic categories into
philosophies faB: single-point, main microphone pickup (mis includes
..:oincldent, near coincident and spaced microphone techniques); smgle point with auxiliary
:ll1crophones, and muln-microphone technique. In the flfSt case, one stereo microphone is
_l~cd
to record the sound as accurately as pos:,.ti)le. This is considered to he the minimalist
.tpproach. The second case uses one main stereo microphone as well with addltlonal spot
mil:rophones to support '.lJeak instruments or groups. This require:; mixmg, as does the
thtrd case where each
m~trumental
group is recorded with its own rmcrophone.
Case two may offer an advantage over case one and three, especially where the hall
tS not familiar, by using spot microphones for emphasis. Also, by adding an
appropriate lime delay to the spot mIcrophones. whlch wIll be
thl:!
~ense
of depth can he Improved. 111lS
microphone techmque to he explored
In
mve~rigatlon
dISl'U~~cd ln
will bc regarded
the same manner
a~
a~
Chapter 2.
another
more convcntlonal
recording techmq ues.
A proper
under~tandmg
of the effects of miCrophone placement
I~
neœ"ary when
making a decI~ion with respect to the sm: of the hall. the dIstance l'rom the prot-,'Tam ,ource
and the dynalTIlc range of the program. A recordmg
clo~e
Will
only glvl the lmprC\\IOn of hemg
to the perforrnanœ if a large amount of high frequency mt0mlatlOn l' pre ...ent
Moving away from the
~ource.
the reverberabon
Incr~:\ses,
dlmmishes, and the dynamlc range compre!\se~ It
1...
dlftïcult to ehnunate the vanable,
why one recording sound, better than another-even when
same haJl.
ThIS
the hlgh frequency content
u~mg
could he due to the nllcrophone placement,
a~
to
the ... ame mam paIr In the
m~trument
placement, the
Fletcher-Mumon effect, proximlty effect, or anothcr acoustlcal phenomena The cngmeer
must quanrify the se results as much as IS
pos~lbk
but remember that the caf 1\ the ultlmate
Judge of the recorded quahty.
Studies by W. Woszczyk 1 and B. Bartlett 2 provlde a clear methodology on acou!o,t!cal
and umbral analysis of both mulo-microphone and
~mgle
mIcrophone
techmque~
on
~ingle
~nstruments. The sound of a musical mstrument or of a complete orchc~tf'J 1" fomlcd h} the
direct sound waves and the reflections from the celimg and walb of the concert hall The
acoustical qualtty depends on the ratio of the dIrect to rcvcrberant
;Jerception of depth. Studies of the VlbratIOnal
charJctensric~
~(,l1nd a~ doe~
of the
orche~tra
the
can deterrmne
the microphone choice and placement 10 the ~ame manner a~ the study of vlbrauonal
characteristlcs of each mstrument. Both close and distant mI<.,Tophone plaœrnent~ can he
1 Wieslaw R. Woszczyk, "Improved Instrument TImbre Through MIcrophone
Placement," Recording Engineer/Producer,Vo1.10, (October ]979),78-95
2 Bruce A. BartIeu, "Tonal Effects Of Close Microphone Placement," J.AudlO
Eng.Soc., Vo1.29, No.10, (October 1981), 726-738.
2
u~ed
to dI~cover where the mInImum number of microphones should he located in order to
reproduce a rea~onable balance of the total spectrum.
AgaIn, hy
u~Jng \pot
microphones
In
combmauon wah the maIn pdlr on an orchestra,
\mall ume delay'> can be Implernented by calculatmg the d\~tances between rrucrophone
cap ... ule~ wlth reference to the speed of sound and hurmdtty and temperature readings from
:he hall
When we IIsten to musIc. we hear a combination of the sound proJuced by the
Jn\trllment~ and the resonance of the room or hall. Denon's anechOlc recordmg samples 3
arc mvaluablc in demonstranng the cffect of dIfferent microphone techmques by removing
lhl\ hall reson,Ulce They also demon~trate the effect of adding resonances from dlfferent
hall~
to the
~ame
ml\leading \mcc
plece of mU~IC recorded In an anechOlc chamber. The results can he
'~c
re\onance) whlch
Room~
not the case In nOlmal recording SItuatIOns
usually do the opposHe of what recordmg engineer~ would hke. They decrease
the low frequcncy
~eparation
1\
InItial reflectlol1s are not pre<;ent (even wlth the addItIon of the room
~cparauon
The oppo~lte
~paclomne~s
Llstening
I~
[0
and Will
sometlmc~
rather mcrease the hlgh frequency
deslrahle especlally with comcIdent recordmgs which may lack
the orchestra or ensemble a~ they rehearse
In
the room and
companng 11m -;ound wlth the Inmal mIcrophone setup can he extremely beneficial. In a
Jead room, one main mIcrophone paIr cannot acnl.l.,ve a blended quality as it can in a livt'
hall. Nomlally a close muluple mI!ang technique would destroy depth but In a dead roorn
lht! reflectlons are nO(
to u~e ~pot
pre~ent
to provide a sense of depth. ln thls case it may be necessary
mlkmg techmques to get away from the negauve
effect~
of the room and
improve overall balance.
Our heanng mechamsrn is responsible for localizmg and judging acoustical sound
quahty. In order to simulale spatIal depth, our ears need the appropriate order of direct
3 Denon, Anechoic Orchestral MUSlC Recordm~, (Japan:Nippon
Columbia Co. Ltd., 1988).
3
·-------------------------------a.....................a
i
sound, first reflections and reverb. In a good ~~ereo recordl!1g, the se early reflections mu~t
be captured with their correct placement to ~m1Ulate natural heanng. They suppon the
original sound and help us to locahze the sound source with respect ln distance
In order to evaluate the perfonnance of the mmn mIcrophone pair, wc must consider the
music, the instruments, the room and the purpose \)f the rcconhng Experimental research
In
choosing a main microphone was conducted by Marun Wohr in Germany. After
conducting listenmg tests as to the best ChOlce for 3 mmn microphone lt wa\ concluded thal
there is no clearly favorable mam microphone technique. Although there
IS
no "best"
solution d~fferent factors can contribute to making a Wl~e choice
In choosing a main microphone for an overall plckup, there are a !1umber )f possible
approaches.
The~e
can be dlvided mto three main groups: comcldent, near-coll1cldcnt and
spaced-pair techmque. They includc XjY, A/B. MIS, ORTF, Blumlc1l1 and the Dummy
Head method. Dependmg on the technique, !ocalizauon can bc achlevcd through 1I1temlly
differences, rime differences or a combination of the IWO. Anglmg cardiold lmcrophones
with the diaphragms together ln a cOlncident arrangement (X/Y, M/S,Blumkm) producc.,
intensity differences. Spacing cardlOld or omm-directtonal
mlcrophone~
(NB,dummy
head) produces ume differences hetween channel~ Angling and spacll1g cardlOid
microphones (ORTF) produces both intensity ar.d time dJfferenccs betwcen the channe!s.
We use a variety of physical eues to de termine the location of a ~ound source Bcslde lime
and intensity differenees at the two ears, changes m the spectral compo<;lt1on of sounds due
to head diffractIOn, pinna effects or sound source movements can abo Influence the
pereeived direction of the source. The more of the~e cues aval!able, the more sure and
accurate the sound source locatIon will he.
4 Martin Wohr, "Untersuchungen zur Wahl des Haupunikrofonverfahrens",
(Munchen: Bildungswerk des Verb,mds Deutscher Tonmeister, Bericht 14,
TQnmeisterta~un.g, 1986).
4
Since dassicaI music encompasses a large dynamic range, the microphones should have
very low noise and distortion. Due to the size of an orchestra, in order for the tonal
qualities of ail instruments to be reproduced equally weIl, the microphone should ideally
have a wlde, flat response at aU angles of incidence and the polar pattern should ideally be
the same at aIl frequencie~.
The fmal consideration in any
mil~ing
situation is the sound. The engineer must decide
whether or not the sound adequately represents the original sound source through
locahzation, depth and presence, as weIl as clarity and balance of the individual
components. Although there is no specific method for miking classical music, it would
appear that spot microphones could he advantageous especially with the option of time
delay compensation.
1.2 Spot Microphone Considerations
In recording an orchestral work in an unfamiliar room or hall, spot microphones may
he used for Improving balance, proximity, reverheration or depth (with time delay
compensation) to embellish the sound from a main microphone pair. They also make up for
the lack of visu al cues of the listener at a live performance. In this manner we are taking
advantage of both close and distant miking techniques.
The spot microphone can serve to improve clarity, dermition and the high frequency
transient information when combined at an appropriate level with the low frequency and
reverherant energy captured by the main microphone pair. For optimal spot microphone
placement the timbre and localization of the instru:l.lent shonld he listened to through the
main pair before deciding on panning and the placement. Il is important to solo each spot
rTIlcrophone, to detennine whether it compliments the sound of the instrument through the
main pll-k-up, and to listen to the combination of microphones for dramatic changes in
sound
5
When considering the complete orchestra, a pair of woOOwind spot microphones are
\
.l
placed up over and high above the woc1winds and a bit in front to add more front edge or
presence to the woodwind and brass instruments, not to change the overall balance. The
microphone acceptance area can be matched to the width of the group. A spot microphone
for a soloist in front of the orchestra can be placed about one meter from the soloist and and
mixed in at a low level-just enough to add definition. In spot mikmg the percussion. the
microphone(s) can be placed about one meter above the timpani or overall perclissive
sections in an orchestra but much closer for a drumset such as in Llstening Test record mg
B. If only one spot microphone is used for emphasis over the nrnpani in the orchestra an
omni-directional pickup may be used (Listening Test A: recording no.l). The percussion
section is also separated sornewhat from the rest of the orchestra, therefore off-axis
coloration should not he a problem.
Because of radiation patterns, the close miking perspective
work~
better on piano,
percussion and bass th an on strings, brass and woodwinds. The effect of single close
microphone placement on the timbre of an instrument would suggest that the instrument
should be miked only as close as necessary. The closer the microphone is placed. the more
selective the output will he. Most musical instruments are meant to sound best at a distance
since an instrument rruhates different tone qualities in ditferent directions and dlfferent
spectra are produced from different parts of the instrument. A flat-response microphone
does not necessarily provide the most natural reproduced sound.
1.3 Negative Effects of Spot Miking
Before deciding whether or not to use spot microphones, the engineer must be aware of
the negative effects that can result from this process. In: "Hauptmikrofon und
Stützmikrofone--neue Gesichtspunkte für ein bewahrtes Aufnahmeverfahren,"5 Günther
1
5 Günther Theile, "Hauptmikrofon und Stützmikrofone--neue Geslchtspunkte
für eio bewiihrtes Aufnahmeverfahren," Bericht 13, (Tonmeistenagung: 1984), 170.
6
Theile states that the non-delayed spot microphone has a negative effect on the simulation
of spaciousness. For simulation of spatial depth the ear needs the appropriate values which
are given by the temporal order of duect sound, tirst reflections and reverberation. If the
~pot
microphone is not delayed, the listener may perce ive the direct sound as coming from
this microphone wnh the main microphone simulating reflections and reverb, thus
distorting the temporal order of the impulse. This is a result of the direct sound from the
mrun pair being late due to distance. What should be perceived as first reflœtions from the
spot microphone are arriving too soon.
There are sorne cases where these "negative" effects may prove beneficiaI as a result of
any number of circumstances. For example, clarity can be improved due to the
s~t
microphone signal arriving ahead of the main microphone signaI. However, this can detract
from the spaciousness of the sound.
1.3.1 Acoustic Phase Cancellation
Phase cancellation is not only produced by dry or direct sounds but aIso by reflected
sound which room surfaces can reflect back into the microphone. If two waveforms are
combined and their relative phase altered, a new waveform will result from both
constructive and destrucùve interference and can cause distortion. Proper use of
Burroughs' 3-to-l rule will minimize acousùc phase cancellation. 6 This states simply that if
Instrument #1 is located one foot from its microphone, then instrument #2's microphone
must be placed at Ieast three feet away from microphone #1. This mIe can he followed as a
genemi guideline however, it does not take into consideration that different instruments will
be playing at different intensiùes. Thus the ear is the best judge in deciding whether or not
the spot microphone is placed correctly or is beneticiaI to the recording.
6 Lou Burroughs, Micro.phones: Desi~n and Application. (Plainview, New
York: Sagamore Publishing Company,Ine., 1974), 115.
7
1
1.3.2 Comb-Filter Effects
Time differences between direct and reflected sound (usually floor reflections between 1
and 15 ms) result in comb filter distortion through dips and peaks of response thal can
color the sound. This results from supenmposing the direct sound on its retlections,
causing cancellations and augmentations within the response range. The resulting
frequency response corresponds to a comb-filter curve.
When recording using multiple microphones, tonal quality can also he detenorated due
to cOlllb-filter effects. The result 15 a fonn of amphtude distortion, due to the sound of an
inso'ument heing plcked up by many microphones at different time intervals resulting in
acoustic phase cancellatlon and distorting the frequency response.
Although there is agreement as to the existence of comb filters, there may he
disagreement as to their relative subjective importance in certain situations.
1.3.3 Microphone Off-Axis Coloration
When using multi-microphone techniques, the engineer must he aware that each
microphone will receive off-axis sound leakage from instruments it was not mtended to
pick up. Although the microphone may have an acceptable frequency response on axis, it
can have a colored or uneven response to sounds reaching it from other directions. It is
important to keep in mind that the published frequency response of a microphone most
often refers to sounds arriving directly on axis. Coloration can occur not only when several
instruments are recorded with multiple microphones, but aIso when many instruments are
recorded with one microphone.
The polar response plot of a microphone shows the sensitivity of the microphone at one
frequency band, plotted for every angle at which sound arnves at the microphone. An
omni-directional microphone may he omni-directional at low frequencies but becomes a
1
weak super cardioid at high frequencies. This is due to a narrowing of the polar pattern at
8
high
frequencie~
because of diffraction, which results in a dulling of the sound off axis.
Uni-directional microphones tend to have a flatter random-incidence response than ornnidlrectional~
because off-axis sensitivity is diminished at hlgh as weIl as low to ITI1d
frequencie~.
However, a cardiOld pattern does not totally dampen the frequencies at 180°.
Wh en choosmg a mi<.:rophone, the random incidence response is ? good indication of the
tonal coloration of the microphone to th" reverberation.
When the spot microphone is used in combination Wlth the main pair its purpose is to
rein force the transient character or add presence to the direct signal, thus drrectional
microphones work best in capturing this on-axis response.
Microphone interference is not as great a concern wh en using only a few spot
microphones throughout the orchestra in addition to the main pair due to the distance
between the microphones.
1.4 Focus and Objectives of this Work
ln order to detennine the value of small time delays in live classical music recordings, 1
have compiled Listening Tests from recordings of the McGill University Orchestra and
smaller chamber ensembles recorded at the Banff Centre. This paper will concentrate on
subjective impresslOns of the effects of time delays revealed through listening tests.
Having reviewed the considerations for c1assical music recordmg and spot miking,
chapter 2 introduces the concept of mixing with delay compensation and discusses
previous work in this area. Chapter 3 presents an overvlew of factors contributing to a
bett(~r
understanding of tim"" delay compensation.
Listening tests, described in chapter 4, indicate the validity of the time adjusttnents.
Comments from the listeners help determine the effect the delays have on the recording (eg.
If frequency response problems are improved by the time adjustment), and whether
subjective preferences are separate from this. Participant'i in the Listening Tests inc1ude
experienced musicians, composers, producers and recording engineers.
9
Chapter 5 gives an analysis of the data from the Listening Tests and the problems
encountered. The results and comments from the pruticipants determine any change in
sound quality and provide a basis for concluding the validity of the time delays; whether or
not there is an optimum time delay (should the delay arnve in step Wlth the main pickup),
an apparent increase in depth, and whether these delays are audible and helpful in both
large orchestral works and smaller-scale ensembles.
A final discussion of the results, in Chapter 6, summarizcs the value of time delay
compensation and how this techmque should be viewed as a recording tool. Potential
problems as a result of this technique are discussed and suggestions are made for areas that
should he explored in future work.
l
10
Chapter 2
MIXING WITH DELAY COMPENSATION
2.1 The Concept of Delay Compensation
A challenging yet relatively unexplored area of recording research deals with the use of
compensating ume delays (5-30 ms) in multiple microphone recording in order to
synchronize the direct sound from the close ffilcrophone with the direct sound of that
Instrument arriving at the main microphone pickup. This techmque is claimed to enhance
the perception of depth which contributes to the impression of a realisnc sound stage.
Major recording companies show great differences 10 their appToach to recording a
claSSlcal orchestra. Sorne believe in a minirnalist approach while others prefeT to use
multIple microphones and multitrack techniques wh en recording classical music for extra
control of balance and definition. As stated earlier, this can result in comb-filter effects.
'The amount of comb-filtering will depend on the levels of the summed signals and the
width of the common frequency band. Despite this effect, an optimal balance may require
the use of multiple microphones-especially when recording in rlifficult acoustic
environments.
This study was inspired by an excellent article: "Digital Time-Coherent Recording
Technique," by Takeaki Anazawa and Yukio Takahashi, 7 in which the authors determine
the detectable and pennissible limits of the comb-filter effect on various instrumental
sounds in multiple microphone recordings through delay compensation of the direct sounds
(see section 2.2)
7 Takeaki Anazawa and Yukio Takahashi, "Digital Time-Coherent Recording
Technique." Audio Engineering Society Preprint, 2493 (H-2), (October 1987), 1-8.
11
•
1
Changes in recording techniques are a direct result of digital recording practices whlch
capture the subtleties and wide dynamic range of orchestral perfonnances. Although the
clarity of digital recording is forcing a trend back to simpler techniques, the dcvelopment
large-scale integrated CircUIt technology has made it possible to delay
~Ignals
very
or
preC\~ely
with no effect on frequency response or detenorauon of sound quahty, thu~ promctmg the
use of spot mIcrophones in addition to the mam stereo pair.
2.2 Previous Work in Time Delay Compensated Mixing
In his article: "A Different Way to Record Classical Music," 8 Jurg Jecklin first
proposed that recording techniques for natural music are improved by tlme-delaying the
spot microphone signaIs.
In lapan, the Denon recording company designed a time-aligned digital console to
correct the anticipatory effects of spot microphones. In the
proce-;~,
Takcaki Anazawa and
Yukio Takahashi concluded that sound quality deterioration was greatly reduced
a~ a re~ult
of digital tinte-coherent recordmg and that it was important to sec ure a delay compensal1on
precision of 1 ms. Tt was also concluded that deterioratlol1 to the ~ound l~ audible even
when the level of the non-delayed spot microphones is 10 dB below the level of the main
pair. 9
Very tinle study has followed regarding the necesslty and accuracy of ume-coherency in
improving the tonal quality of recordings. The effect of crosstalk between microphones and
the effect of delay compensation on reinforcing the hall tone has yet to be detenmned. Long
8 Jürg Jecklin, "A Different Way to Record Classical Music," J. Audio
Eng.Soc., Vo1.29, No.5, (May, 1981), 329-332.
t
9 Takeaki Anazawa and Y ukio Takahashi, "Digital Time-Coherent Re<.:ording
Technique," Audio Engineering Society Preprint, 2493 (H-2), (October 1987), 3.
12
term liMening
15
a necessary factor in detetmining the improvement and effect this
technique has on the overall balance.
As referred to earlier, Denon manufacrurers have since produced a compact dise
"Anechoic Orchestral MusIc Recording" showing thl! effects of time I.!day compensation in
an anechOic chamber. They have demonstrated the effects of this technique using nmecoherent recording with vanous mIcrophone setups. The anechOiC recordings help 10
demonstrate that e\ en without Ole room reflecuons, a close perspective may require spot
mH.;rophone~
~pot
to improve upon the pickup from the main pair. However, the addition of
microphones with room resonance can be misleadmg smce support mIcrophones bring
out spectral components not present at the rruun microphone due to the lack of reflecuons.
Theile 10 suggests that when using spot ITIlcrophones not only must ume delay
compensauon be used, but an increment must be added to the calculated delay in order that
the direct sound from the spot microphone arrives with the first reflections at the mam pair.
2.3 Delayed Spot Microphones
By using carefully measured delays, it is possible to synchronize the arrival time of the
sound from a spot micr<Jphone and the main pair. If sounds do not arnve at OOth
microphones simultaneously, there will be sorne acoustic phase cancellation which can
result in frequency cancellanon thus reducing the volume and changmg the color of the
Instrument. It is important to listen to the combination of the microphones for drasnc
change~ 10
the sound, especlally when using spot microphones in closer proxirnity on a
srnaller ensemble. The different path lengths from source to microphones result in different
phase relationships between the microphone outputs for the same sound. Also, the
supported instruments can be reproduced with considerable precedence and sound too
much up front, thus disturbmg the spatial quality of the recording. In order not to affect the
10 Günther Theile, "Hauptmikrofon und Stützmikrofone--neue Gesichtspunkte
fûr I!in bewahnes Aufnahmeverfahren," Bericht 13, (Tonmeistenagung: 1984), 170.
13
spatial sensation and depth perspective of the instruments whe'1 usmg spot
microphone~.
the signals from these microphones must be delayed by the travel ume of the 'iignal to the
main microphone.
Using carefully measured delays with the spot microphones maintams the recording
properties of a good stereo main pair with respect to the arnval of direct sound,
fIr~t
reflecl10ns and reverberation. It also allows the engineer to increase the level of the
signaIs without bringing the supported mstrument too much up front.
Becau~e
~pOI
our ear~
need the appropriate order of direct sound, tirst reflectIons and reverb.to sImulate :.-.patial
depth, these early reflection'i must he captured with their correct placement lt would
therefore seem that time delay compensation.would he the best solution when usmg "pot
mIcrophones.
It is possible to measure the delay so that the effect belween the direct ..,olmd and tiN
reflections remain the same. Since early reflections carry most of the dl~tance tnfOn1lallon,
and the first sound gives the location, il would seem 10 theory that the delayed
~pot
"Ignat
should arrive in time with the firsr. reflectlons. Usmg Thetle's suggc\t10n of adding an
increment to the delayed calculation would accomplish this. This could abo relate to the
earlier mentIoned theory by Ohve and Toole that a strong reflecuon can have a po"nive
effect in renewing the precedence effect.
In an earlier article: flThe Subjective Effects Of Flrst Reflecnons In Concen
Need For Lateral RefIectlons", Il Barron suggests that spatiallmpres~ion
1S
Hall~- The
pnmanly a low
frequency phenomenon whlch depends again mostly on lateral sound energy beJow 400 H/
arriving between 10 and 100 ms after the direct sound. This
sugge~t~
that lime delay l'i
important since it is the reflected sound which should have a high level of lateral vclocity.
Barron's work also revealed that spaciousness depends on the level of low frequencies.
Il Barron, M. "The Subjective Effects Of First Reflections In Concert HalJ!.The Need For Lateral RefIections." lSound Vib. (1971),475-494.
14
Theile also points out that it may have a positive effect to delay left and right spot
~ignals
differently in order to simulate sideways ref!ections which do not change
localization but do increase volume. This supports the earlier work by Barron regarding the
importance of low frequency laterai refleetions for simulation of spaciousness. Delaying
spot signals would therefore seem benefieial by incr\.!a"iing the chance for lateraI reflections
at low frequencies.
15
------------------------------------.........
r
Chapter 3
OVERVIEW OF FACTORS CONTRIBUTING TO A GREATER
UNDERSTANDING OF AUDIBLE EFFECTS OF DELA Y COMPENSATION
3.1 Psychoacoustic Considerations
It is imponant to examine avaliable psychoacoustu.: evidence to suppon
compensatmg ume delays. In order to do so, It is neccssary
10
th~
use of
first look at the way our
hearing mechamsm works.
Psychoacom.uc~
mdicate that the real study of sound
Although the ear 15 a hlghly develope.d organ,
1t 15
powers of the brain. The arnval ume of a sound
take~
place between our Iwo ear"
useful only whcn coup\cd wlth the
i~ u~ed a~
a loc.ltIon elle hy the ear-hntl/l
mechanism ("precedence effeet") ThIS bemg the case, delaymg the
~ound
l'rom the "'pOI
mKTophone so that It coincIdes wlth the arnvaI ume at the mam pIckup wou Id
~eem
necessary in maintaining a correct rime relaùonshIp
Although natural heanng 1S more satisfactory when listcmng in a hall at a dI'Itanœ from
the source, mIcrophones cannot focus
111
on sounds as our ears can We may he left wÎlh a
nice sound but wIth no presence or sc i1se of definiuon from a parucular Instmment. The
lack of visuaI eues, which also have a large mfluence on auditory locaillauon, mu~t he
compensated for in miklllg a sound source.
U~ing
small time delay'l on the '>pot
microphones to compensate for the distance from the main pair can help the
compensate for the lack of visual eues
10
h~tener
ereating an IllusIOn of depth.
Stereo heanng must be considered as being d1fferent from natural heanng
~ll1ce
in stereo hearing hears two copies of the sound rather than one copy of the dIrect
from the source. Level and time differences at the hstener's ears are not the
1
16
~ame
each car
~ound
a..'1 tho~e
from the loudspeake~, thu~ various stereo techniques are used to ensure that the
loudspeaker ~ignals produce eues that are compatible with natural hearing.
In reviewmg the proce~s for natural hearing, lt is ObVlOU~ there are intensity and time
differencc~
~ource
between the t'wO ears due to amplitude dlfferencc'l because of localizauon of the
and the
phy~\(:al
dlfference 10 arnval tIme due to the spacmg of our ears. There are
also dlfferencc'> III the -;pectmrn of the
~ound
entenng the
t'YY\1
car canals partly due to head
diffractIon.! 2 Delay . . a~ ... mall a~ a few hundred mlcroseconds (because of the physlcal
bump~,
ndges and caVH1es of the pmna) are used as locauon cues by the ear-brain
rnlXham~m Thl~
belOg the case, the use of urne delay compensation should he easily
percelved by our ear-bram mechanŒm.
The Fletcher-Munson curves or Equal Loudness contours arr
!'Ien~lt1vJly
LI '" .:ries
of graphs of
of the ear versus frequency at dIfferent loudness levels. These are helpful in our
unde~tanding
of complex tones, showing that the rate of increase in loudness with an
IOcrca~c 10 mten~lty i~
frequencle~,
greater for low frequencies and to sorne extent for very high
than for muidle frequencies. "IbiS mformation
i~
also useful in our
understanding of varymg intensiues in eomb-filtering effeets where multiple microphones
may combme weIl only at eertam frequencies. Here again small time delays may help in
controlhng the overallievei of an instrument or section in eombination with a main pair.
3.2 The Role of First Retlections
The dclayed signal from a spot microphone aets as a type of refleetion in that it arrives
after the direct sound from the main pair. First reflectlons play an important role in
recording. Together with the reverberation they provlde a signifieant sound clue to the
12 Sams, R,~fsrenee Data FQr En~ineers; RadiQ. ElectrQnies. Computers and
CQmmunicatiQ.llli, seventh edition, (Indianapolis, Indiana: Howard W. Sams, 1985),
34.
17
room's size, proportion and wall structure.The impression of the room size mcréase~ with
\
•
an increase in the delay time of the reflections.
The frrst ex peri me nt to use simulated reflections in order to better understand the roIe of
first reflections was conducted in 1950 by Helmut Haas, who mvestigated the effec! of
single short-delay reflectlons with speech. Thls has
SlIlCC
become hetter known a~ the Ha"s
effect which states that reflcctions occumng within approximatcly 40 mIlltscconds of the
direct sound (exact orne dependent on frequency content and envelope of the sound)
become fused with the direct sound. Haas noted changes In loudness, ~ound quality and
body of the sound due to these reflectlOns. He also discovered that fu~ion will occur even
If the closely timed echo cornes from a different direCtiOn than the onglllal ,>ource and the
location of the total sound
tS
determined by the locatIOn of the first ,>ound. The ~econd
stimulus would have to be made sufficlently louder (15dB) above the first ~ound before
overriding the precedence factor. 13
Whlle direct sound dctermines localizaoon, early ret1ectIon~ carry O1ost of the dl~tar.c(..
information. The later diffused reflecuons are the reverberation, they glVe mfonnal!On
~ourcc
about the size of the room and sorne information about the distance of the
but nOl
It!-.
direction. Reverberation time is not the only determinant of acousocal quahty. The carly
reflection sequence has an important influence on subjectively audible qualily difference"
JO
diffe:ent rooms or halls.
When listening to a live concert, the ear receives a half dozen or more reflected
repetitions of the original sound. This early sound field also pre~enh front, back, 1efl,
right, top and bottom views of the sound. The!.e VICWS provldc the h"lef'l
varieties of tone color produced fro: ' the spectrum of the instrument.
'f
OUf
WIth the I11hny
auditory ~y~tcm
is capable of organizing this informatIon into a sing .; percept. It is Important that the
13 Brian C.J. Moore, An Introduction to the Psyçholo~y of HearIn~, Second
Edition, (London: Academie Press Inc., 1982), 163.
I
18
recording engineer be able to recognize the subtle qualities of an instrument (due to its
precise construction) by these tirst reflections in the hall, and he able to evaluate the total
blend of these qualities with those of other instruments. This can only he done through
very criticallistening.
Reflected sounds rarely have the same spectrum as direct sound since the high
frequeney content is nonnally reduced. Experiments by Sean Olive and Floyd Toole 14
agam indtcate that lateral reflections generate much more of a sensation of spaciousness
whereas vertical reflections are more apparent as having an effeet on timbre. The focus of
this anicle however is on the effect of reflected sounds in stereophonie reproduction in
typical TOoms. Results show that delayed sounds arriving from the same direction as the
direct sound are often less audible than when arriving from other directions and a strong
reflection following a reflection-free interval can have a positive effect in renewing the
precedence effect. Once again this supports having the delayed spot signal arrive in time
with the fifst reflections.
3.3 The Perception of Depth in Recording
Before trying to simulate spatial depth through small rime delays it is important to
understand how depth is perceived with our hearing mechanism.
Our binaural hearing capability is largely responsible for the perception of depth.
Therefore in order to simulate spatial depth, the stereo signal must contain elements similar
to the ear signais for normaIlistening. Again, the order of direct sound, frrst reflections and
reverb at the main pair must he similar as for naturallistenmg.
It is important to consider that the acoustics of a room have an influence on the depth
perspective. With a fixed microphone distance, the more reverherant the hall, the fanher
back the rear of the orchestra will seem. By moving the main microphone further back in a
14 Sean E. Olive and Floyd E. Toole, "The Detection of Reflections in Typical
Rooms," Audio Engineering Society Preprir.+, 2719 (F-1), (November, 1988), 1-15.
19
hall, the ratio of the front to back distance will decrease and the front of the orchestra will
become softer in relation to the level at the back. Varying the height allows us to change the
front-to-back perspective. When using cardioid microphones bass instruments a1ways
seem to be doser on the recording than wh en listening in the hall due to the disumcedepending compone nt. Many factors are mvolved when judging a mit-lophones senSinVH)
to the ratio of direct to reverberant sound energy 10 a room. This wlll depend on bo:h the
instrument and microphones directional characteristlcs as weIl as the the position of the
microphone to the instrument and the directional charactenstics of the room.
Air absorption can conni.bute to the se.lse of distance
10
an acoustlc environ ment.
Also,as we move away from a sound source the high frequency content is reduced. These
high frequencies can thus be manipulated in a recording to change the apparent front-tobackdepth.
In his anicle: "Spaciousness and Localization m Listening Rooms and Their Effect'\ on
the Recording Technique",15 Griesinger says that by electronically lficreasing separatIon dt
low frequencies, coincident microphone techniques can produce supenor recordings. He
proposes a technique of spatial equalization which works by applymg a bass boost to the
L-R signal and/or a bass cut to the L+R signal. In this manner the low frequency spatial
impression of the hall is emphasized without affecting imag1l1g at higher frequencies. Till . .
can be related to Barron's earher (iiscovery of the effect of lateral reflections at low
frequencies in relation to spaciousness. The imponance of the abihty to control the
directivity of low frequency information is apparent. A microphone technique propo\ed by
Wieslaw Woszczyk1 6 appHes the technique of second order gradient unidirectionahty to
15 David Griesinger, "Spaci.ousness and Localization in Listening Rooms and
Their Effects on the Recording Technique," J.Audio Eng.Soc., Vo1.34, (April
1986), 255-268.
16 Wieslaw Woszczyk, "A Microphone Technique Applying The Principle Of
Second-OrderGradient Unidirectionality", Audio Engineering Society Preprint (Los
Angeles, May 1981).
20
extend the high directivity of the pickup to low trequencies since low frequency energy is
not easily dissipated in a room. This is achieved by creating four pressure points to be
measured by the microphone configuration. Whereas pressure microphones sarnple the
acoustic field at one point only, first-order gradient microphones will measure the pressure
at two points while second-order gradient microphones will measure the output at four
points. What Woszczyk has proposed is an arrangement of two cardioid microphones
wherem a second-order system will he effective at low frequencies while a transition to a
frrst-order system will occur with increasing frequency. Further research into this
technique could reveal new possibilities with respect 10 improvement in spaciousness by
the advancement of directional microphone pickup at low frequencies, thus de'leloping yet
another tool in sound recording.
3.4 Musical-Instrument Radiation Patterns
By studying the vibrational characteristics of the orchestra, in the same manner as
studying the vibrational characteristics of each instrument, microphone choice and
placement can he bener determined. This applies to both the main microphone and spot
microphones placed within the orchestra. Although time delay compensation may help
when combining the sound from the spot microphone with that of the main pair, it depends
largely on the placement of the spot microphone. Capturing a cenain frequency component
may make up for a lack of presence of that component al the main pair.
The timbre of a musical instrument is made up of the various spectra radiated from
different parts of the instrument. 1. is essential to know the radiation patterns of the various
instruments in arder to better understand microphone placement. especially where spot
microphones are involved.
Briefly, string instruments have very complex directivity patterns. Each string group
has a nondirectional characteristic for the low frequencies with the exception of the double
1
21
r
bass. At higher string frequencies, there are regions of preferred radiation that are different
for every partial and can alter dramatically with frequency. Due to the complexity of the
radiation of string instruments, the frequency response at close range can result in comb
filtering. Where spot mikmg is necessruy, placement is crucial and carefullistening IS
required when combming the close microphone with the main pair.
Woodwind instruments have definite multi directional patterns since the sound is
diffracted in all directions through the open tone holes. In general, the components that he
below a certain cutoff frequency of each instrument will radiate in an omni-directional
pattern. For components near this cutoff, the T'ddiation will be dise shaped while
components above this will radiate In progressively smaller angles from the heU in the
direction of the hall. Bere microphone placement is more imponant than type since the
timbre changes more depending on direction. Placing a pair of spot nucrophones over the
woodwind section in an orchestral recording can add definition to this section without
getting too close to any one instrument. Adding delay compensation can also increase the
depth perspective.
Brass instruments are much more omni-directional in tenns of sound radiation sin ce the
direction al characteristic is largely symmetrical about the bell. Radiation becomes narrower
with increasing pitch which results in a duller sound the further off axis the microphone
IS
placed. The characteristics of the brass instruments help them to overcome distance in
being placed at the back of the orchestra. They frequently sound closer than other
instruments located at the same distance from the microphone. Although spot microphones
may not be necessary on the brass section in an orchestral recording, using delay
compensation on the woodwind spot microphones also helps in achieving a greater sense
of depth for the brass instruments.
The recognition of musical instruments depends on transients and the structure of the
total sound envelope. In recording we must balance the sharp transient detail with the rich
J
harmonie information in the reverberant environment. Here the timbre of the instrument is
22
~-
the result of the interaction between the direct and reverberant sound fields. Microphone
sensitivity to this interaction is a result of the directivity of the sound from the instrument,
the position of the instrument to the microphone and distance from the microphone, the
polar pattern of the microphone and the room characteristics.
The direct sound is important forc1arity and defmition especially for short tones yet the
total spectrd of an instrument is measured best in a reverberant room where multiple
reflections combine to produce a full ensemble sound. Since musical instruments produce
low frequencies mostly omni-directionally, very tinle of tbis energy arrives at the
microphone as direct sound. The percentage of low frequency infOlmation is much greater
in the reverberant field and adds wannth and fulness to the sound.
In looking at overall orchestral radiation, we must take into consideration that the
arrangement of the orchestra was chosen for obvious reasons due to balance and the
spectral radiation of sound from instruments. As we move higher up in a hall, we perceive
more high frequencies since the high frequencies of many instruments, especially strings,
radiate upward rather than forward. Most orchestral instruments have a preferred axis of
sound radiation that is taken into account in various orchestral seating arrangements.
23
a
Chapter 4
RESEARCH PERFORMED: EXPERIMENTAL METHODS AND
IMPLEMENT ATION
4.1 Procedure
In order to investigate the subjective impressions of time delay compensation, 1 have
created a set of listening tests to better detennine their validity as a recording tool.
The purpose of the listening tests is to demonstrate whether small compensating rime
delays can be detected in the context of both large and small classical musIc ensembles
through changes of balance, depth, tImbre, level or other unknown factors.
Comment~
should detennine which factors are subjectively audible. They should aho reveal whether
the delays do affect depth and wh ether optimum delay~ exist in varying situation~.
The subjective judgemeL~ of acoustical quahty can he very elusive due to the number of
variables which can affect the recording. Employing a set of listening tests for a group of
trained musicians, composers and sound engineers can help ln eliminating a number of
prior assumptions. My initial experience in working with
delay~
soon indicated that the
most suitable subjects were those who were both used to listening to music and had sorne
experience with audiological tests. A group of founeen mdividuals were tested, seven of
whom are professionally tramed recording engineers. The rest of the group was comprised
of composers, ml.lsicians and producers, only one of whom is not a profession al.
Thi~
person was added for subjective testing purposes. Although the number of subJect!> used
was small, the variation in response between the listeners was smaller th an expected.
Listening tests were given individually over the same monitors in the
~ame
room. The test
required concentrated listening. Instructions were given on the tape; each excerpt was
indexed for convenience and to allow extra time for comments where necessa..-y.
24
In addition to using one main pickup for a live to two-track stereo recording, a
backup multitrack tape was made using various main stereo pairs and individu al spot
microphones for accenting individual Instruments or sections. Afrer each recording
temperature and humidity readings were taken in the hall as weB as precise measurements,
descTibed in the following section, for calculating time delays.
4.2 Methodology
ln order to measure the results as objectively as possible, 1 chose to divide my test into
four separate sections. Listening Test A deals only with the orchestral recordings from
McGill. In most of the 10 sections making up this test, the subjects were asked to compare
various excerpts to one reference (the main microphone pair). In this manner 1 was able to
add varying amounts of time delay to the spot microphones (which are always compared to
the main pair) or add no delay to test for depth perception, optimum delay time and the
audibility of comb-filter efftcts or other relevant factors. Spot microphones are used on
only the necessary instrumental gr.')ups and not more than four spot microphones are used
at a time. This was found to he best in controlling the overall balance.
Listening Tests Band C are both recordings of smaller groups in the Project Studio at
the Banff Centre. In the se tests, the emphasis is on longer listening excerpts and on
differentlatmg between main stereo pairs and close miking possibilities in this room. In
both cases one main pair is chosen as the best pickup. To this 1 added a spot microphone to
each instrument or group of instruments in order to reinforce the balance of any or all
instruments and to avoid the negative effects of the room. The spot microphones in this
case are aIl added to the main stereo pair fIfSt without delay compensation then with delay
compensation. The excerpts are longer so as to demonstrate the long term effeet on balance
and not initial comparisons as may he the case in Listening Test A. These tests also
demonstrate the results of using delay compensation on both a good (Test C) and bad (Test
B) recording from the same hall.
25
,
Listening Test D is another extreme in showing the difference in using very shon NB
excerpt eomparisons. This test is used to show whether a difference can he perceived in
very short rime intervals.
Instrumentation: Delay units used in experiments:
McGiII Recordings:
Eventide Ultra Harmonizer H3000 (Digital Delay program):
Left channel: 1.26 ms residual delay
Right channel: .2 ms residual delay
Roland E-660 Digital Parametric Equalizer(Digital Delay PlOgram):
Both channels: .2 ms residual delay.
Banff Recordings:
Lexicon 480L Digital Effects System (Twin Delay Program):
Both channels: .2 ms residual delay
A pulse generator and oscilloscope were used in order to chee!.. for residual delays
within the units used. The .2 ms residual delay is minimal for both the Lexicon 480l. and
the Roland Digital units. A 1 ms delay was subtracted for the left channel calculated delays
for the Eventide in order to compensate for the residual delay.
AlI units were tested for Freque·.lcy response, 1HD (Total Hannonic Distortion) and
!MD (1nterrnodulation Distortion). Graphie displays and results are located in Appendix D.
Specifications given for eaeh unit are as follows:
Lexicon 480L: Frequency Response: 20 Hz-20 kHz, +0.5 dB,-! dB.
THD&Noise, <0.015%@ 1kHz limit level (+18 dBm unity gain)
IMD, <0.05% SMPTE lM @limitlevel
26
Eventide H3000: Frequency Response: 5 Hz to 20 kHz +/-1 dB, +/-.5 dB typical.
Distortion: .01% (.007% typi..::al)@ 1 kHz, 1 dB below clipping in "pitch
change" mode, 0 shift, levels aU at 0 dB.
Roland Digital Parametric Equalizer E-660: Frequency Response: 20 Hz-20 kHz, +0/-3
dB.
THD <0.015% (1 kHz at rated input)
AlI units show good specifications in all areas for the purpose of these tests-(see
Appendix D)
The amount of delay is varied in both the positive and negative directions to investigate
whether or not there appears to he an optimum time delay and to further investigate
microphone crosstalk and the effects of over compensation.
The listening level was kept constant between the non-delayed and delayed spot
microphone. This was achieved by connecting the oscillator to "multitrack in", at the
desired channel on the patch bay (set to "O"VU), and to the left or right channel of the delay
unit. The retum of the unit Was then patched 10 the desired channel, using the unit's level
setnng to match the "0" VU setting. Since comb-filtering drops sorne of the energy of the
sound, the level may vary between dela}ed and non-delayed signaIs. In a few cases (see
listening test solutions), 1 varied the level of the delayed spot microphone for comparative
purposes where 1 felt the level difference was too noticeable. Also, increasing the delay
will reduce the apparent amplitude of the signal. The calculated level, which was initially
determined subjectively for the non-delayed spot, is always used in an eartier excerpt
before any change is made in level.
27
T
1
4.2.1 A Calculation Model
H
IMINNIÇ.
The fonnula used for calculating the time delay between the spot microphones and the
main pair as follows:
H represents the height from the audience floor to the main microphone capsule.
hs represents the height ofthe stage (l(>6cm)
hl represents the height from the audience floor to the spot microphone capsule.
h2 represents the height of the spot microphone capsule from the stage floor.
h3 is the difference in the height between the spot microphone capsule and the
microphone capsule.
Therefore,
h3=H-hl
hl=hs+h2
h3=H- hs-h2
d="h3 2 + L2 (from A2= B2+C2 )
d=" (H-hs-h2)2+L2
Time= d/v(wet)
t
28
mam
If the temperature reading is 2TC and humidity is 40%,
from the CRC handbook of Chemistry and Physics we know that
the velocity of sound at 40% humidity (20·C- 2k frequency) is 343.95 rn/sec.
and 343.56 rn/sec. for 0% humidity (20·[email protected]).
We also know the ve!ocity of sound in dry air at 27·C is 347.44 rn/sec.
V(wet)=V(dry) x v 40% humidity(20·C)/vO% humidity(20·C)
V(wet)=347.44 x 343.95/343.56=347.84 rn/sec
If d=6 m, then T=6/347.84=17.2 ms
4.2.2 Temperature and Humidity Effects on the Speed of Sound
Tight control of temperatllre and relative humidity must accompany the use of very
smaU rime delay increments to improve room response. In implementing small rime delays
to compensate for differences in distances between microphones, temperature and humidity
readings must be used in calculating the speed of sound, which is used in the overall
calculation of the amount of delay.
The speed of sound is dependent on the temperature and humidity reading in the hall.
Tiny changes in the se conditions will affect the phase relationship of direct and reflected
waves. Both temperature changes and moi sture affect the density of air and therefore the
speed of sound in air. Since moist air is less dense than dry air there is an increase in the
speed of sound with an increase in humidity. AIso, an increase in sound absorption due to
the humidity reading will cause a decrease in reverberation time where surface absorption is
low; this is not significant for frequencies below 2 kHz. The frequency response of a
microphone would not he significantly altered by humidity changes; there would he a small
increase in absorption only at very high frequencies.
An acoustical delay is a direct result of the finite speed of sound. For normal
r
tempemtures (20-22°C) sound travels 344 rn/sec. This will vary with both temperature and
29
humidity changes. For the purpose of this expenment the steady suue reading at the end of
,
cà
the performance is used since it can slightly affect the calculation for the speed of sound
The CRC Handbook of Chemistry and PhYSlcs 17 pubhshes results of varying temperamn:
and humidity readings on the speed of sound.
4.3 Listening Test Recordings
1. McGill Symphony Orchestra: Pollack concert hall; January 25,1989
Microphone placement for recording: (see Diagram no.l)
Mozart: Oboe Concerto in C, K. 314
Bruckner: Symphony NO. 7
ln order to capture the best overall pickup, 1 set up four main microphone palTs during
rehearsal. Since the performance look place on two consecutive mghts 1 was able to
improve on the microphone placement for the main pairs and spot microphone~ for the
second performance (which was used for the hstemng tests). My goal was to capture th~
best overall sound with one main microphone pair and use spot microphone~ to glve
detinition to the woodwind, string and percussion sections without de~ltroymg depth A
hypercardioid 1S used on the bass section to glve extra definition of the bow on the
~trJng~.
Precise measurements were taken between microphone stands and the helght of each
microphone was calculated from both the audience and stage floor each mght. A
temperature and hurnidity reading of the hall was taken during each night's performance.
For the initial mil(, 1 dec1ded between a stereo Blumlein (AKG C422), an AKG 414
Blumlein, and the closely spaced B&K 4003's as the maIn pair.
The C422 Blumlein configuration did not produce an even plckup between capsules,
and so was not used. The 4003's provided a more spacious reproduction whlle the AKG
414 Blumlein setup provided a doser, more intimate sound reproduction. This worked
17 R.C. Weast, Ed., "CRe Handbook of Chemistry and Physic!"", Boca Raton,
Florida: CRC Press Ine., 1979.
t
30
especially weil on the Mozart piece in helping to distinguish quiet musicallines being
;.>layed simultaneously by the orchestra. It was also easier to judge the relative distances of
different sounds in the reproduced image Wlth the Blumlem pair. Thus the addition of
delayed spot microphones lends itself better to this technique rather than to the spaced omm
paIr where the time difference would have to he considered separately for each channel due
to the greater spacing between capsules.
The addition of the B&K 4006's worked weU in combination with the 4003's in
creating a 4-omm wall technique, yet they caused phasing problems when combined with
the precise imaging of the Blumlein pair.
Having chosen a main pair, 1 experimented with the addition of the delayed spot
microphones. In aU cases the faders were brought up discretely Gust before they became
audIble as additional microphones). The woodwmd microphones were placed 2 meters
apart above the wind section and angled toward the left and right center of the sectionjust
befofe the brass. This was most helpful for the deep seating arrangement of the Bruckner.
The addition of the solo microphone (Mozart), helped to stabilize the instrument and
improve the attack without adding much more finger noise. The spot microphone on the
first row of basses improved articulation of the sound of the bow on the strings and helped
tO
capture the lower notes which the double bass is unable to radiate effectively. The string
microphones were used in an attempt to give body to the string sound. Being placed behind
the strings, the microphones gave a slightly richer sound than when placed in frOnt,
especially for the violins. A single percussion microphone was used above the timpani
(Bruckner). The U87 was set to omni since it was the only microphone used and provided
a better overall pickup.
31
Diagram No.l
1
1. McGill Symphony Orchestra: Pollack concert hall; January 25,1989
U87 PERCUSSION
WOODWINDS
KM84'S
~
cl
KM84
KM84
STRINGS-L
STRINGS-R
MSS
-=0
BASS
US7
SOLO~
~AKG414's
TBlumlcm
OB&K
U 4003
B&K Q
B&Kî
4003 0
QB&K
U 4006
4006
STEP 1
. 'H- 556
M'
h
AKG 414' s (Bluml'
am M'lcroplone:
em confi19uranon)
. m
v (wet)m/s
d
~otmics
h3
L
l.percussion (U87) 0 2.5m
9.3m
347.84
9m
2.strings-L (KM84)
6.83
347.84
2.lm
6.5m
3.woodwind-L
6.48
347.84
2.2
6.1
(KM84)
4.woodwind-R
2.1
6.36
347.84
6
(KM84)
5.strings-R (KM84)
5.7
347.84
2.15
5.3
347.84
6.Bass (M88)
8.4
3.8
7.5
7. solo (U87)C'J
4.2
347.84
3.75
1.2
T
26.9ms
19.6ms
18.6m~
18.3ms
16.4ms
24ms
1lms
Note: The two sets of spaced B&K omni microphone pairs as well as the
AKG C-422 were not used for the experiments.
t
32
2.McGill Symphony Orchestra: Pollack concert hall; April 1, 1989
Microphone placement for recording: (see Diagram no. 2)
Bruch: Violin Concerto
Mahler: Symphony No.!
During the rehearsal, both a Blumlein pair (M130's) and an ORTF (AKG 414's) pair
were placed on a large stand quite high over the orchestra and angled for the best overall
pickup and to eliminate sorne of the noise frcm the audience picked up by the figure-eight's
hack lobes. An initiallistening test between the se two main pairs and a spaced pair of B&K
4006's resulted in my choice for using only the ORTF configuration for the concert.
The Blumlein pair, although capturing an "open" ambient hall sound, lacked breadth.
Although they gave a good sense of depth, the total sound lacked power. The ORTF pair
proved to he a better choice. There was less rQOm feh, yet the solo violin was more
integrated with the orchestra. The placement was generally off-axis to the brass (this suited
the repertoire) and placed higher above the strings for a more even pickup. The overall
string sound was much clearer and the pickup much more immediate on the peicussion
section. Since the percussion microphones are added only for attack (low end frC<.J.ùency
rolled off at 100Hz), a high-pass filter was used on all four microphones. Again, 1 placed a
spot microphone on the first stand of basses (mic. approximately 50 cm from instrument)
for articulation, and a solo microphone to help localize and define the attack of the violin. A
pair of microphones were placed at an angle over the woodwind section (1.75 meters
apart). The snot microphones were mixed initially without delays and at a low level just to
the point where they added clarity to the overall mix. This was a subjective choice on my
part. The levels were not changed with the addition of the delay for listening test purposesunless specified as such.
,
\-
33
Precise measurements were taken between microphone stands as weil as the height of
each microphone from both the audience and stage floor each night. A temperature and
humidity reading of the hall was taken during each night's perfonnance.
Diagram No.2
McGill Symphony Orchestra: Pollack concen hall; April 1, 1989
TùD.}.
'1':1.&) . lCM84
'tI87 EH8..
Jcre.
~
U87
lue.
l'
WOOl>WDD>S
I:Mt4'S
l'
HICK40
==0
Q=:o
H.~
H88
BASS
V87
AXG 414',
OllTl
. ) H=606
M'
- m
am mIcroplhone: AKG 414' s (ORTF confi19uratlon
v(wet)
spot mics
h3
L
d
l.harp (MKH40)
4.25
9.5
347.84
8.5
2.percl (U87)c:>
10.6
347.84
3.5
10
3.timpl (KM84)
9.5
347.84
3
9
9.5
347.84
4.tim~2 (KM84)
3
9
5.perc2 (U87)c:>
9.5
347.84
3
9
6. W oodwind-L(KM84) 1.9
6.7
347.84
6.5
6.7
347.84
7. W oodwind-R(KM84) 1.9
6.5
347.84
8.4
9.4
8.Bass (M88)
4.3
9.5010 (U87)c:>
4.1
347.84
3.1
2.7
34
T
27ms
30ms
27ms
27ms
30ms
19.5ms
19.5ms
27ms
12ms
1
"
3. The Banff Centre: Project Studio Recording; December 2, 1989
Microphone placement for recording: (see Diagram no. 3)
Stravinsky: L'histoire du soldat
For the recording of A Soldier's Tale, 1 set up three main microphone pairs in the
Project Studio at the Banff Centre. As can be seen from Table 3, a spot microphone was
used on each of the six instruments as well as five spot microphones to coyer the
percussion section placed at the back of the studio. The initial placement of the instruments
was chosen for best compatibility with the main microphone.
After listening carefully to each microphone pair, 1 chose the Dummy Head
configuration as the main pickup. Il provided the best localization af the group yet there
was a slight hole in the middle of the reproduced image. Using spot microphones in
combination with the main paif eliminated trus. Because 1 used spot microphones on all
instruments in this case, the spot microphones were panned to match the position of the
instruments at the main pair . 1 then did a mix using only the spot microphones, creating an
even balance at a low leveJ. The main pair was mixed in at a higher level with the spot
microphones at a lower level so as to add c1arity to the mix. While maintaining the spot
levels, calculated delays were added to the spot microphones and recorded on empty tracks
so as to create both a delayed and non-delayed multitrack mix.
1
35
y
1
Diagram No.3
The Banff Centre: Project Studio Recording; December 2, 1989
Stravinsky: Soldier's Tale
...oJ
K
L
i
i
F
E
~
G
~
~
?
H
12.13m
D
?
"
"
/
c~
Ay
X
B
<
Y
11.02
)
Three main pairs are represented by A,B, and C.
A=Sennheiser Dummy Head (using two B&K 007 microphones)
B=Blumlein (Beyer M130 microphones)
C=X!Y (AKG 422 Stereo microphone)
The Dummy Head was chosen as the main microphone in combination with the spot
microphones.
Humidity: 35% Temperature:23·C
V(wet)=V(dry) x v 35% humidity;(20·C)/v 0% (20·C)
=345.12 x 343.89/343.56=345m!sec
Due to the method of measurements laken for recording sessions 3 and 4, L is calculated
from L2=B2+C2 (B=AX-X; C=A Y- Y); h3=H-Z; d=../h3 2+L2
1
36
~--~----------~~----nr----~~~~~~-.---~--~~--Microphone
X
Y
Z(helght) h3
L
d
T
f
A-Dummy Head (B&K4007' )5.07(AX) 4.10(AY) 2.21(H)
!
~B~-~B~IU~ml~e~m~(~B~e:l~'e:T~M~13:0:"s~)~~5~.0:'7~::~3.~.8~:8~~~~:2~.6~1~~=;:==~~=::::::=::=-_-=::
C- X/Y (AKG C-422)
D-Vln (U87)C'J
E-Comet(TLM170)
5.07
4.10
4.11
4.50
5.70
6.44
2.26
1.60
1.33
-l
1.87m 1.97m 5.7m~ ,
2.53
2.68
7.8ms ~
MF~:C~1~m~·n~e~t(~(A~K"~G~~~:)----~4~.5~5---h7~.4~9--~1~.3~10~~~~I~~3~,.4~3-+~3.5~5-+~lG~
. 3n~1
G-Bassoon (B&K 4011)
5.30
7.96
0.77
1.44 3.47
3.76
1O.Yms 1
H-Bass (BeyerM88)
6.33
6.00
0.40
1.81
2.66
3.2
9ms
I·Trombone(AKG414)C'J
6.46
5.25
0.88
1.33 1.36
1.9
5.5m~"
hJ~-S~n~ar_e~(S~e~n~n.~Ml*)~__~~4-..4~8__-h9~,.6~7~~~O~.6~0__~~1.~61~~5~.6~~5~.8~~16~.8~J
hK~-~T~om~s~·(~A~K~~G~,4~14~)_C'J____~5~.l~3~~'9~. 4~4~~-.O~.4~6~~~1.~75~~5~~~4-+~5~.6~~16~'~~~i
L-Cymbai (KMI40)
5.40
8.84
1.54
.67
4.75
4.8
13.9m~
M-Orum (M88)
6.35
9.59
1.35
.86
5.64
5.7
16.5ms
N-H. Hat (KMI40)
5.97
10.18
1.50
.71
6.15
6.2
17.9ms
37
L
.61m
.88
r
l
4. The Banff Centre: Project Studio Recording; January 29, 1990
Microphone placement for recording: (see Diagram no. 4)
Pecou, Thierry: Un temps jusqu'au bout de la fibre.
An X/Y AKG C-422 and a Dummy Head were set up for choice in determining a main
pair. Spot microphones, as can he seen in Table 4, were used on ail instruments or groups
of instruments with exception of the DX-7 (l used a direct pickup rather than mike the
amplifier). The four string and four wind players were arranged 10 such a manner as to
allow for one pair of microphones to he used on each group. A pair of ribbon microphones
were used to record the piano. These were placed under the left and nght side of the sound
board which provided a nice pickup and good isolation. In this case the AKG C-422 stereo
microphone was chosen over the Dummy Head configuration as the best overall pickup.
Having matched the position of the spot microphones with the main paIr, 1did a mlX using
only the spot microphones to obtain the best balance. 1 then brought up the maIn
microphone pair and made a final subjective judgement as to the levels of each mi<"ï"ophone
for the best overall balance. A mix was done using the main microphone only, the mam
microphone with spot microphones and the main microphone with delayed spot
microphones (keeping the same levels).
38
Diagram No.4
4. The Banff Centre: Project Studio Recording; January 29, 1990
Pecou, Thierry: Un temps jusqu' au bou~ de la fibre.
FR HORN
TPT
()o-
TUBA
A
PIANO
i
oboe
'""0 HARP
clar
flull!
jl
'\
WINDS-L
DX7-dorect
WINDS·R
'\
STRINGS·L
-0
via
vin
vin
bassoon
jl
B~S
vcI
STRINGS·R
~ DUMMYHEAO
:r
XIV AKG c-422
B&K4007's
~
Humidity: 35% Temperature:23°C
V(wet)=V(dry) X V 35% humidity;(20·C)jv 0% (20°C)
=345.12 X 343.89/343.56=345m/sec
Note: distance measured directly from each microphone capsule to the main microphone
capsule in this case.
The AKG Stereo microphone (X!Y) was chosen over the Dummy Head microphone as the
. pau
.,ln corn b"matIon W1·th the spot nucro
. plhones.
mam
spot microphones
T
d
V(wet)
1. Strings-L(U87) C')
3.4m
9.9ms
345m/sec
2.~~trings-R(U87)C')
345m/sec
10.2ms
3.5m
3. Bass(M88)
3.4m
345
9.75ms
4. Piano-L(M130)
S.8m
345
17ms
5. Piano-R(M130)
6.1
345
17.7ms
6. Woodwind-L(Sony C38)
5.02
345
14.5ms
7. Woodwmd-R(Sony C38)
4.80
345
14ms
~. Harp(M88)
7.95
345
23ms
9. DX-7 (dir.)
345
23ms
7.95
10. Fr. Hom(B&K4011)
Sm
345
23ms
11. Tpt.(AKG414)CJ
6.60
345
19ms
17_. Tuba(AKG414)C')
6.80
345
19.7ms
39
Chapter 5
LISTENING TEST RESUL TS
Initially the Listening Tests were devised to detennine whether 1) the use of
compensating time delays does add depth, 2) the delay is perceivable by the participants
and 3) whether there is an optimum time delay. From the results of the Listening Tests, it is
obvious that the use of delay does add depth regardless of the subjective choice of the
listener. That the delay is perceivable 1S suggested by the consistency of differences being
noted. An optimum time delay seems preferable when delaying more than one section of
the orchestra-in most cases with four spot microphones in combinallon with the main pair.
Both the overall calcuiation for preferences and the indlvidual comments of the listeners
must be taken into consideration in the analysls of the test results. 1 will summarize the
results in section 5.1 and explain these results in more detail through the interpretation of
comments in section 5.2.
5.1 Analysis of Data
Listening Test C: shows the most positive result for lime delay compensation as a
useful recording too1. A good stereo recording should capture accurate imaging, good
acoustics, tonal accuracy and depth. Listening Test C captures these elements with a stereo
microphone pickup in "excerpt 2". Five excerpts are used to demonstrate two different
main microphone techniques and to c1arify the effects of delay compensation.
40
Listening tests reveal the following:
1) all participants chose main pair no. 2 to be the best main microphone pickup.
2) sorne participants chose main pair no. 2 with spot microphones as an audible
Improvement over an a1ready good stereo pickup and aIl heard the effect of the spots
despite preference (see Appendix C)
3) all participants noticed an improvement in depth in excerpt 4 and 90% (including all
recording engineers) chose excerpt 4 with time delay compensation as the best choice.
This shows a positive result for time delay compensation with an a1ready good main
microphone pair. Both Listening Tests Band C were recorded in the same room and both
use longer listening examples, which seem to work best in detennining overall preferences
with respect to balance.
By contrast, Listening Test B proved to be the least acceptable recording. The main
pickup is not g<X>d: although comments reveal that the listeners hear an improvement in
depth, individual preferences are very divided. Although comb-filter effects can deteriorate
the sound, Listening Test B reveals that time delay compensation will not necessarily
improve the recording and an increase in depth may actually be detrimental to the
recording. No reverb was added to this recording in order to better determine the effect of
delayed spot microphones.
The comments as to the preferred technique differed and listeners were divided on the
benefit of the addition of delays to the spot microphones, even though the majority heard
the addition of depth with the delays added (see Appendix C). Auxiliary microphones were
needed to ensure an optimum balance in the final mix
xcerpt no. 3 (spots added
without time delay compensation) was preferable for balance with the proximity of the
close microphones. lbis could be partially due to the circular placement of the musicians in
front of the main pair. ln this case the closer instruments can mask the more distant ones
The further instruments sound too distant or dull in a not 50 reverberant space. Because of
1
41
1
this seating arrangement, the extra presence of the instruments with the spot microphones
added is more important than resulting comb-filtcring effects.
In Listening Test A, multiple microphone setups were used to record large classical
works. General conclusions from Listening Test A reveal that not only is time delay
compensation detectable, it does increase depth. Changes in sound quality with rcference to
the main pair were highly detectable by all participants. A delay was almost always
preferable in this set of tests and there was a majority preference, with few exceptions, for
one excerpt.
A calculated delay is preferable whenever four microphones are used in combinaùon
with the main pair. In Test III: 100% preferred the calculated delay for the combination of
both string and woodwind microphones with the main pair, and 80% preferred
th~
calculated delay again with strings now 2 dB lower in level (excerpt 6). This seems to he
due to an improved sense of balance throughout the orchestra (see Appcndix C).
Delaying only ont. spot microphone or a microphone pair (woodwinds or strings)
resuhed in greater differences of opinion hetween listeners as to the preferred amount of
delay. Using spot microphones on the string secùon did not necessanly Improve the
balance as the strings sound very strident in general. Another problem may arise Wlth the
"optimum" rime delay being optimum only for the instrument(s) belllg supponed whereas
the delay may have a negative effect on surrounding instruments. Funher diSCUSSion of thi~
matter can he found in section 5.3.
Where level changes were made, they are always viewed as an improvement by the
majority with respect to balance. This was mentioned cartier with reference to Test III. In
Test V: again 100% preferred the calculated delay and 75% of the listeners opte<! for the
calculated delay with an addition al change in level for bass and percussion spot
microphones (see solutions). In Test VIII,90% preferred the calculated delay, 80% of
whom preferred the changes made in level (all sound engineers preferred thlS change
level ). Test IX is one funher ex ample with 85% hearing the calculated delays as an
42
III
f
'\.
improvement in the percussion section and 75% of this group preferring the adjustment in
level.
Listening Test D demonstrates that even with very shon A!B comparisons differences
were heard by a large majority for aU examples and delays are generally preferable to nondelayed spots. However, comments indicate that the excerpts are too short for the listeners
to define the differences. An example of this can be seen when comparing Listening Test
C, excerpt 4 to Llstening Test D. Whereas 90% of alllisteners chose excerpt 4 in C, only
35% chose the same excerpt as a preference for a much shorter comparison in Listening
Test D.
5.2 Interpretation of Comments
In order to gain a better perspective on what the listeners reaIly heard, a list of
comment,; have been compiled in Appendix C. Both individual responses and overall
judgements support the above analysis of the tests.
In interpreting the results of these comments, 1 will begin with Listening Test A (Tests
1 through X) TEST 1 supports the idea that the delay is perceived by the listeners. Various
recording engineers reference to the horns s being "rounder", "louder" and "further back"
with the lime delay shows that the perceived image is best for this instrument. Other
listeners are commenting on different sources which also support the perception of the
delay anà the increase in depth. This is obvious with such comments as "the brass section
could he more prominent" or there being "more strings than before? .. this is the best
naturaI balance".
Number
n demonstrates diversity in personal preferences. The use of time-delay
compensation as a recording tool in this case is a matter of preference which would require
longer tenn listening. Here, spot microphones and/or delay cou Id he used to improve an
already bad pickup for the strings. Although an improvement in balance is noted by several
43
listeners in excerpt 6, a slightly larger majority prefer excerpt 4. The perception or' the delay
and an increase in depth arc supported by the comments.
As stated earlier, an optimum (calculated) delay is preferable whenever four
microphones are used 10 combination with the mam pair; this is found in Tests III, IV, V,
VIII and IX. With Test III, the 80% preference in excerpt 6 seems to be due to th..!
'balance' factor, with comments referring directly to a "better" or "ntl'Cr" balance. The
comments listed for Tests V and VIII are similar and point toward an overall improvement
in balance. The four microphones used in Test IX work very weIl for the percussion
balance but not necessarily for the overall balance. Sorne comments indicated that the mmn
pair alone was "best" yet others feel that excerpt 4 bcst supports the perCUSSlOn
instruments. One particular comment supports both: "In terms of volume and spatial
balance, 1 defmitely prefer the main pickup; in tenns of percussion timbre and sOlmd,
excerpt no. 4 seems to best support the percussion instruments, but they are way in front
of the rest of the orchestra and much louder".
Test IV also reveals a majority preference for the addItion of delay to the four .:ipot
microphones, yet long term listening should be used 10 better determine ove raIl balance as
the comments are varie.d in tenns of why excerpt 2 is preferable.
Tests VI and VII, dealing with 1 or 2 spot mIcrophones, Ix>th reveal varied subjective
preferences. There is no consensus through comments as to what detenrunes a better
balance, despite a majority preference for one excerpt. In Test VI, excerpt 4, comments
indicate that the listeners hear the presence with the solo
~pot
microphone yet differ in
opinion as to its value. Comments in the se two sectIons support the earlier analysl~ that
delaying only one spot microphone or a microphone pair (woodwtnds or strings) re~ults
In
greater differences of opinion between listeners as to the amount of delay.
Where level changes were made, comments reveal an improvement by the maJonty with
respect to balance. This is demonstrated in Tests I11,V,VlII and IX. In Test III, where the
t
string spots are lowere.d m level by 2 db. the comment'i refer ta an improvement in the
44
overall balance ... "like no. 4 but better balance";" ... nicer balance between sections". This
same reference to balance is found with level changes made in Tests v,vm and IX through
such comments as: "...clearer, and weIl blended/balanced. If; "good depth and solo blend";
"best blend"; "best balance".
Listening Test H, as stated earlier, reveals that tlme delay compensation will not
necessarily improve the recording, In excerpt 4, comments support the addition of spot
microphones (though by a small percentage) without the addition of delay ... "percussion
more detined, v\olin closer,woodwinds are too present... If; "better images of separate
instruments ... "; "cleaner in percussion, dry sound"; ".. .improved placement ofvIOlin".
Comments in favour of the addition of the delay to the spot microphones demonstrate
that the delay is heard: "definition is there without 10ss of depth ... "not much different from
excerpt 4. 1 preferred 4 but can't pin-point the reason. It was more 'natural', but lhis
sounds a little deeper".
Obviously both points of view are subjective. Long term listening is required in order to
better determine the results. The v~Jue of time delay compensation as a recording tool is
demonstrated by the ability of the listeners to discern an improvement in presence by the
addition of spot microphones, an increase in depth and in sorne cases an overall
Improvement in balance.
As stated earlier, Listening Test C shows positive results for time delay compensation
as a useful recording tool where the main pickup alone has captured an already good stereo
recording.
Comments reveal that sorne participants chose main pair no. 2 with spot microphones
as an audible improvement over what they considered an already good pickup:
"spot on bass? more realistic, good sound, much clarity-beautiful".
"cello clearer, strings clearer, french horn clearer, piano a bit weak, basses better-more
defined"
45
T
l
1
Other comments reveal that the listeners hear the addition of non-delayed spot
microphones and point out both positive and negative effects. This results in vruying
subjective opinions:
"flatter stage (less sense of depth and space). Piano and bass both more present and upfront. Not as spot-lighted as no. 2 in tenus of precise instrument placement, but still easy
to detennine where each pomt source is".
"highlights certain strands of the melody but texture is less homogeneous th an excerpl
no. 2"
"string section is clearer now "
"narrow image-instrument.; on top of each other"
One listener comments on the effect of comb filtering on th~ clarinet: "my
guess is no delay to spot mics; clarinet shows comb filtering-sounds duB on
lower frequencies; sorne brass fIoats".
AlI participants noticed an improvement in depth in excerpt 4 and 90% (includmg aIl
recording engineers) chose excerpt 4 with time delay compensation as the best ChOlce:
"Now we have depth! ... If; "best, a11 timbres are clearer"; "preferable to excerpt 3", "better
balanced imaging"; "wider imaging and spatially better"; ".. has the m081 detail";
"Thl~
sounds best, most natura1. Good stereo spread, al! instruments clear, present and defined,
without over emphasis ... ".
Listening Test D... as stated earlier, reveals that differences
majority with very short
w~re
heard by a large
A/B comparisons The delays are generally preferable to non-
delayed spots yet the excerpts are too short for any consistency ln tne
hstener~ re"pon~e\.
An exarnple of the vanety of comments to support this statement i~ eVldenr from Test 1:
"Quite a subtle difference ... "; "not much difference ... no delay?"; "CaIl hear slight
difference; perhaps lower instruments a bit more prominent in B"; "oboe louder in B";
"more orchestral sound in B ".
46
In conclusion, although there are subjective differences of opinion through the various
comments, the results show that the use of compensating time delays 1) adds depth, 2) is
perceivable by the participants and 3) an optimum time delay seems preferable when 4 or
more spot microphones are used in combination with the main pair. This seems especially
true for orchestral rccordings where careful choice of spot microphones can improve the
balance without destroying depth. There is no consistency in the comments to suppon
Theile's theOl"y of adding an increment to the calculated delay for optimum results. This
th~ory
rt'fers to over-compensating (larger) delays in order that the signal arrive at the main
microphone in time with the first reflections (as opposed to strictly compensating (exact),
or under-compensating (smaller) delays. Long tenn listening and level changes may also he
necessary in making subjective decisions for improving balance.
47
Chapter 6
SUMMARY AND DISCUSSION
The results of the listening tests show time delay compensation to he a useful tool, with
both positive and negative results. From a positive viewpoint, it can provide extra
spaciousness, reverb and sustain of the sound by improving the sense of depth. l'hese
same positive attributes can become undesirable 10 certain situations, destroying
perspective and manipulating depth by pushing the present clements further back. Although
the delay does align the sound al the spot nucrophone with the arnvaJ lime of thal sound al
the main stereo pair, this may not be desirable or it may not he the most imponant factor to
consider in a given situation. Improving balance, proximity or reverb may he more
beneficial to the recording.
6.1 Time Delay Recording Technique: A Tooi For Balancing
When combining the output of severa! microphones, the audibllity of comb-filter effecls
is a major factor in the subjective assessment of a recording. Results from the listening
tests confmn that the addition of spot microphones in remforcing the direct sound can be
important for clarity and definition, and there can he an audible improvement in the sound
quality and depth perspective with delay compensation of the spot microphones. However,
comb filter effects may not be the most important aspect as far as a ~pecific recording
15
concemed. The spot microphones may he used to enhance proximity or improve the
balance without the use of delay compensation.
Time delay recording is important as a recording tool because a noticeabie difference CWl
be achieved in the time relation regardless of how desirable the result may he. ft b
t
48
important for the recording engineer to he able to list both positive and negative effects in
any given situation before making a decision as to whether or not to use this techmque.
As a positIve effect, time delay compensation can provide extra spaciousness, reverb,
balance and sustain of the sound. However, depending on the recording, it can destroy
perspective and manipulate depth by pushing the closest elements further away. In certain
situations, extra reflections and reverb can destroy the clarity of the direct sound.
Long tenn listening IS necessary for assessment of balance. Upon hearing more reverb
or a brighter sound, our initial reaction may be that this is better without our taking the rime
to really concentrate on the overall or long tenn listening effect of the perceived balance.
This seems to be the case in reviewing the results of Listening Test A, in comparing the
different excerpts to the main pair. Due to the length of the excerpts listeners seemed to
immediately respond to spaciousness, brightness and loudness as being immediately
preferable. The excerpts rnay not be long enough to determine the effect of the bnghtness
or other changes in sound quality on the overall balance.
Although the use of a spot microphone may not effectively simulate early reflections,
this may not be the intent. In sorne cases the presence or inrimacy achieved through the use
of multiple microphones will help in reinforcing the direct sound, especially when the room
tone is not desirable. It can improve the dynamic range and the overall balance.
It should be the goal of the engineer to produce a captivating presentation of the music.
Although there is no one "right way" to accomplish this, it is important that we learn and
experiment with as many recording techniques as possible so that the relevant knowledge
and recording 100ls available will help us, especially in problem situations.
6.2 Potential Problems
One very important discovery that seems to have been overlooked in previous research
is the increac;ed effect of reverberation or amount of reflected sound, with the use of rime
delay compensation. Although we have been dealing mostly with the direct sound,
49
1
depending on the situation, rime delays can have a great deal of effect on the recording due
J
4.
to their potential effect on reflected sound. Thus an optimum delay may only be good for
the instrument concemed but damaging for surroundlng groups.
It is irnponant to make a decision as to what we are hstemng for. The Denon anechoic
CD experiments show that the technique does work in term~ of spaciousness yct may hun
the close perspective of the strings. The use of time delay compensation in an anechok
situation must be regarded as a differem kind of tool that IS u~ed for a dlffercnt purpo~e.
We lose the perspective of the music without the interaction of the room It can he
~ecn
from the following diagram that although spot mlcrophone~ are not usro specifically on the
strings, other spot microphones influence the sound of such a large section. The
microphones in this case can be thought of as destructive
"flashlight~"
~pot
of sound. There b
no proof that the use of delays does not extend the reverb Through lime delay
compensation of the spot microphones, the strings are also delayed, not only by the
woodwind microphones, but also by the percussion microphones. In sorne recordings,
again depending on numerous factors, this effect may be more obvlOus than in other ca~cs.
woodwind mics
/
fi
Q\----1\
1
- -_
ft
\
\
\
\
. . . - strings ..
.,-
strings~ bass
mie
y
soloist
mam parr
1
50
The quality of the recording seems to he a major factor in determining the validity of the
time delays. The results from the Stravinsky recording indicate that the poor quahty from
the overall pickup was a prime factor in varying subjective impressions.
Mixing the dIrect sound from multiple microphones will not automatically aclueve the
proper balance of reverberant sound. This may not he as crucial in the case of mam
microphone-spot microphone techniques if the reverberdIlt sound is balanced 111 the main
pickup and spot microphones are used for improving clarity and deflmtion of the direet
sound.
Further research into the pickup of fIfst reflections at the spot microphone mu!>t he
investigated and could pose dangerous problems to the sound in certain situations;
especlally in the case of floor reflections which can be destructive to the timbre of the
instrument depending on the room.
According lO Anazawa, it is desirable to secure a delay compensation precision of 1 ms
or better. In the case of aligning the direct sound from the spot microphone, comb filtering
effects may he more detrimental with very short delays (1 ms) , especially with short
sounds. Thus Ume delay compensation would have to he very precise. This can he difficult
to ensure due to the many variables that can effeet the final measurement. The direct sour.d
may contribute to first reflections which mayor may not he beneficial depending on the
sound quality of the room.
The potential problem in sending the signal through the ND and DIA filters in digital
delays is most important for fidelity of the direct sound. The accuracy of AID conversion ,~
the heart of the digital recording process; the quality of the delay unit is most important.
There is the danger of overloading the delay unit depending on the dynamic range of the
direct sound, if the device is not properly set. Although there should he no effeet on the
frequency responlle,the transients may cause overshoot and ringing that may occur when
the sound passes through the converters.
51
.
6.3 Future Considerations
As stated earlier, the room plays an important role in carrying timbrai infonnauon.
Room response problems must therefore be investigated as a major factor in detemllOmg
the validity of compensating time delays. ObvlOusly precise calculauons are only valid if
there are no inherent delays of the sound because of room response. 1 have included two
graphs that were created using Sound Designer II software l8 . They represent the respon~c
of the Project Studio to a one millisecond impulse of pink noise sent through two speaker-;
placed back to back to capture the full sphere of sound radiation in the room. Two B&K
microphones (4007) are used to capture the response of the room to the pmk
noi~e. Thi~
test was done following the Stravinsky: L'hlstOlre du Soldat record mg in the Project Stud!(\
at the Banff Centre. The speakers were placed in the approximate locatlon of where the
clarinetist sat. The far microphone position is in the location of the main stereo pair
n.s
meters) with the close position movmg the second B&K microphone one-half mCler do . . ~r
The results reveal an inherent delay aiready present in the room at 15 ms wher..: the
signal appears as high as 50% of the level of the impulse. This
IS
espccially notice able in
the far position.(graph #2)
t
18 Sound Designer Il software, Version 2.0 ©1990 by Digidesign Inc.
52
g
"j3'
"a '.
~
0
a
3
CD
CI
~
~
III
•
Z
0
:::It
if
.ë3
.....
1.
...
~
en
,...
C
&.
0
~
(ft
..
1.
CI
1..
0
•
~
\"
0
\,.it
8!.-
~
Ct
\;;
'a.
~
'"
ft
0'
CIl
GI
i
~
0"
::J
bot
0
a
"8::1
CIl
~
n•
0
CIl
n
2.
ra
"8~.
c.
0
::1
S3
"....
l51P
.....
-.
~
l'
•
Z
0
0
3
'"
.~
11\
~
i"
....
!P
-....
SI)
~
~.
-g
~.
0~
rD
....
en
....
c:
(')
_.
\?
c.
0
8'"'
3
d
~
0
::s
~
0
1
Pr
ë.
!'l
'8...
......::s
tII
0
o
o
31/1
t
C7\
<:)
"A
<:)
"0
~
~
0
~
lA
0
,0
1
0
0
~
l,
"CI)
0
54
~
0
~
0
0
~
<.
1\1
~
In the future this technique may become a more flexible t001 in sound recording by
being a built-in function on a recording console. In this manner spot microphones will
provide not only the possibility of helping balance through changes In level and
equalization, but also in time relationships-influencing both the direct sound and reverbth us adding a new dimension in the balancing process.
Adjustments in perspective, imaging, loudness and balance are possible through easily
accessible changes in the rime relarionship of the spot microphones to the main stereo pair
This requires critical listening skills on the part of the sound engineer. Perhaps the
possibility of the opposite relationship-in delaying the main pair to emphasize the
importance of the spot microphones, would be beneficial in another situation. With so
many factors contriburing to the overall sound, the ear must be the ultimate judge.
Finally, these tests have convinced me of the advantages of using multiple microphon~5
as a tool for improving balance. This technique is not meant to replace other microphone
techniques but can he viewed as yet another method which can give very good results in
sorne applications.
1have made reference to many recording techniques, new discoveries and suggestion)
for improvement through careful analysis throughout (his paper. Without going into detall
or having tried to prove any one technique as being superior to another, l hope 1have
demonstrated the way each has to be considered as a valid recording tool in the art of SOUf(
recording.
55
BIBLIOG RAPHY
...
Anazawa, T. and Takahashi, Y. "Digital Time-Coherent Recording Technique." Al.l.diQ
Society Preprint, 2493 (H-2), October 1987.
En~ineerin~
Anderton, C. The Di~ita1 Delay HandbQok. New York: Amsco Publications, 1985.
Barron, M. "The Subjective Effects Of First Reflections In Concert Halls-The Need FOI
Lateral Reflection s." J.Sound Vib. (1971)
Bartleu, B. Introduction 1'0 Professional Recording Techniques. Indiana: Howard
Sams & Company, 1987 .
w.
Bartleu, B. "Microphone Off Axis Coloration." Recording EngineerlProducer, Vol. 1L
pp.82-93, June 1980.
Bartlett, B. "Recording Techniques: Simple Stereo Microphone Techniques." db
Magazine, pp.8-14,Sept./Oct. 1986.
Banlett, B. "Tonal Effects Of Close Microphone Placement." lAudio Eng.Soc., Vol
No.lO, pp.726-738, October 1981.
Benade, A.H. "From Instrument To Ear In A Room: Direct Or Via RecordlOg."
Eng.Soc., Vo1.33, No.4,April 1985.
Bohn, D.A. "Environmental Effects On The Speed Of Sound." J.Audio
Vo1.36, No.4, pp.223-231,April 1988.
2~,
J.Aud~,!
En~.Soc.,
Burroughs, L. Microphones: Design and Application. Plainview, New York: Sagamor.!
Publishing Company,lnc., 1974.
Coeon, C. "Comparative Stereophonie Listening Tests. "J.Audio Eng.Soc.,VoI.20,
No.l, pp.19-27, Jan./Feb.1972.
Cross, L. "Stereo Microphone Evaluations: An Overvlew of Techniques and Subjective
Assessments." Recording Engineer/Producer, pp.24-29, August 1987.
Denon. Anechoic Orchestral Music Recording. Japan: Nippon Columbia Co. Ltd.,
19x:~
Dilley, W. "Recording Perspective." AudIO Magazine, pp. 34-38, April 1961.
Eargle, J. Handbook Of Recording Engineering. New York: Van Nostrand Reinhold
Company Inc.,1986.
Everest, F.A. The Master Handbook Of AcoustIcs. Blue Ridge Summit,PA: T<:.b
Inc.,1981.
BO::>l\~
Faulkner, T. "A Classic Case." Studio Sound, pp.66-71.
Gerzon, M. "Stablishing Stereo Images." Studio Sound. pp.60-64, December 1974.
56
(
Griesinger, D. "SpacÎousness and Localization in Listening Rooms and Their Effects on
the Recording Technique." J,Audio En~,Soc., Vo1.34, pp. 255-268, April 1986.
Jecklin, J. "A Different Way To Record Classical Music." J,Audio
No.5, May 1981.
Katz, B. "Achieving Depth Perception in Recording." db
1981.
Lambert, M. "Recording The Classical Orchestra." Mix
1989.
En~.soc.,
Ma~azine,
Ma~azine,
Vo1.29,
pp. 52-55, August
pp. 1 10-117, January
Lipshitz, S.P. "Stereo Microphone Techniques: Are The Purists Wrong?" Presented at Ii-'r_
78th Convention of the Audio Engineering Society, l.Audio En~.Soc., Vo1.34, ~o.9,
September 1986.
Meyer, J. Acoustics And The Performance Of Music. Frankfurt/Main: Verlag Das
Musikinstrument, 1978.
Meyer, 1. "Directivity Of The Bowed String Instruments and Its Effects on Orchestral
Sound in Concert Halls. "Journal of the Acoustical Society of America, Vol.5l, April
1972.
Moore, B.e. An Introduction To The Psycholoey Of Hearin~. Second Edition. London:
Academic Press Inc., 1982.
Olive, S.and T oole, F. "The Detection of Reflections in Typical Rooms. "AlliliQ
Eo~ineerin~ Society Preprint. 2719 (F-l), November 1988.
Opolko, F. and Woszczyk, W. "A Combinative Microphone Technique Using Contact
And Air Transducers." Audio En~ineerin~ SocietY Pre.print(149l), May 1987.
Petersen, G. "Orchestral Miking: Two Views." Mix
Peus. S. "Microphones and Transients." db
Ma~azine,
Ma~azine,
pp.68-71, October 1987
May 1977.
Preston, B. "A Coïncident Microphone Technique." Studio Sound, pp.34-36, Nov.
1977.
Repka, C. "A Guide To Coincident Mikes." Audio
197ft
Ma~azine,
pp.41-51, November
Sams. Reference Data For En~neers: Radio. ElectrQOics. Computers and
Communications, seventh edition. Indianapolis, Indiana: Howard W. Sams, 1985.
Snape, C. "DID/Denon." StudiQ SQuDd, pp.56-64, July 1988.
Streider, R. and Dooley, W. "Basic Stereo Microphone Perspectives A Review."
AydiQ EDe. Soc., VoI.33,No.7/8, pp.548-555, July/Aug. 1985.
J...
Theile, G. "Hauptmikrofon und Stützmikrofone--neue Gesichtspunkte für ein bewahrte~
AufDahmeverfahren." Munchen: Bildungswerk des Verbands Deutscher Tonmeister,
Bericht 13. Tonmeistenaiun~, 1984, pp.170-184.
57
.
Turabian, K.L. A Manual for Writers of Tenu Papers. Theses. and Dissertations, Fifth
Edition. Chicago: The University of Chicago Press, 1987 .
Weast, R.C., Ed. CRC Handbook of Chemiso:y and PhysiCS. Boca Raton, Florida:
Press Inc., 1979.
Cl~(
Wohr, M. and B. Nel1ess~n. "Untersuchungen zur Wahl des Hauptmikrofonverfahren,. '
Munchen: Bildungswerk des Verbands Deutscher Tonmeister, Bericht 14,
Tonmeisterta~un~, 1986.
Woram, J. "The Mathemaùcs of the Microphone, Parts 1 and 11." db
July 1981.
Ma~az1!le,
June &
-,:--___ ' The RecQrdin~ Studio Handbook. Plainvlew, New York: Elgar Pubhshmg
Company Inc., 1982.
Woszczyk, W.R. "A Microphone Technique Applying The Principle Of Second-Order
Gradient Unidirectionality. ".Audio En~neerin~ Society Preprint. Presented at the 691h
Convention of the Audio Engineering Society, Los Angeles, May 198 1.
____ . "Improved Instrument TImbre Through Microphone Placement."
En~ineerlProdueer,Vol.I0, pp.78-95, Oetober 1979.
B~cordJng
"Multimierophone Pickup of Solitary Acousttcal Instruments for Smgle
Channel Transmission." Audio En~ineerin~ Society Pr~rint( 1491), May 1987.
-,:------c--.
58
APPENDIX A
LISTENING TEST A
I.Mozart Oboe Conceno: Woodwind spot microphones (listen for horn
entries)
8 excerpts (aIl odd numbered excerpts are the same...comment on the
differences you hear with the even numbered excerpts which include the
woodwind spot microphonE"J In combination with the main pair)
Please rate the even numbered excerpts to the main pair :
1=same or little difference
2=better
3=worse
4=best(if applicable)
DAT
Jrrdex #'s
1. main pair reference
2
2. comments:
3
3. main pair reference (same as 1)
4
4. comments:
5
5. main pair reference
6
6. comments:
7
7. main pair reference
8
8. comments:
II. Mozart Oboe Concerto: 7 excerpts
String Spot Microphones
9
1. main pair reference
1()
2. comments:
Il
3. main pair reference (same as 1)
59
12
4. commcnts:
13
5. main pair reference
14
6. comments:
15
7. main pair reference
III. Mozan Oboe Concerto: 6 excerpts
Woodwind and String Spot Microphones
16
1. main pair reference
17
2. comments:
18
3. main pair reference
19
4. comments:
20
5. main pair reference
21
6. comments:
IV. Mozart Oboe Concerto: 2 excerpts
2 woodwind, 1 bass, 1 solo (oboe) spot microphones (main pickup
included in both)
22
1. comments:
23
2. comments:
60
V. Bruckner: Symphony No. 7(not 3 as stated on the tape): 6 excerpts
2 woodwind, 1 bass, 1 percussion spot microphones
24
1. main pair only
25
2. comments:
26
3. main pair only
27
4. comments:
28
5. main pair only
29
6. comments:
VI. Bruch: Violin Concerto: 5 excerpts
Solo VioUn Spot Microphone
30
1. main pair only
31
2. solo mie added; comments:
32
3. solo added; comments:
33
4. solo added; comments:
34
5. solo added; comments:
61
r
vn. Bruch: Violin Concerto: 5 excerpts
Woodwind Spot Microphones: ail jnclude main pair
35
1. comments:
36
2. comments:
37
3. comments:
38
4. comments:
39
5. comments:
vrn. Bruch: Violin Concerto: 3 excerpts
2 woodwind, 1 bass, 1 percussion spot microphones...
"ail include main oail:
40
1. comments:
41
2. comments:
42
3. comments:
62
IX. Mahler: Symphony No. 1: 4 excerpts
4 percussion spot microphones
43
1. main pair only
44
2. spot mies added; comments:
45
3. spot mies added; eomments:
46
4. spot mies added; comments:
X. Mahler: Symphony No. 1: 4 excerpts
2 wood\tind, 1 bass, 1 timpani spot microphones
47
1. main pair only
48
2. spot mies added; eomments:
49
3. main pair only
50
4. spot mies added; comments:
63
LISTENING TEST B
DIFFERENCES IN SOUND RECORDING METHOD
I. The following 5 excerpts are used 10 de mon strate the sound quality
obtained by using vanous rmk.ing techniques and to clarify the effecls of
delay compensation. 1 have included five excerpts from Stravinsky's
L 'histoire du soldat, using 3 miking techmques plus auxiliary
microphones and delay compensation effects.
51
1. Stereo Microphone XfY recordmg technique Oevel difference); comments:
52
2. 81umlein recording technique (level difference); comments:
53
3. Dummy Head technique (time difference); comments:
54
4. Dummy Head with spot microphones; comments:
55
5. Dummy Head with spot microphones; comments:
LISTENING TEST C
56
1. main pair no. 1; comments:
57
2. main pair no.2; comments:
58
3. main pair with spots; comments:
59
4. main pair with spots; cornments:
64
LISTENING TEST D
This test i~ based on an A-B comparison test The excerpts are short.
change you hear between the two.
~,ote
any
Pause Dat Tape Between Excerpts for more time..... you may use thf index numbers
to repeat these short excerpts
60
1. comments:
61
2. comments:
62
3. comments:
63
4. comments:
64
5. comments:
65
6. comments:
66
7. comments:
NOTE: excerpt no. 8 incorporates the main pair and spot microphones in both A
&B!
67
8. comments:
65
APPENDIX 8
CONDITIONS
LISTENING TEST A
I.Mozart Oboe Concerto: Woodwind spot microphones (listen for hom entries)
8 excerpts (all odd numbered excerpts are the same ... comment on the differences you hear wllh
the even numbered excerpts which include the woodwmd spot microphones in comblllation
with the main pair)
1. main pair reference-A KG 414's (Blumlem)
2. comments: -no delay
3. main pair reference (same as 1)
4. comments: -calculated delay (18 ms)
5. main pair reference
6. comments: -shonened delay by 2 ms (16 ms)
7. main pair reference
8. comments: -longer delay (21 ms)
II. Mozart Oboe Concerto: 7 excerpts
String Spot Microphones
1. main pair reference
2. comments: -no delay
3. main pair reference (same as 1)
4. comments: -calculated delay (19/16 ms)
5. main pair reference
6. comments: -shoner delay (16/13 ms)
7. main pair reference
.'
66
III. Mozart Oboe Concerto: 6 excerpts
Woodwind and String Spot Microphones
1. main pair reference
2. comments: -no delay added to 4 spot microphones
3. main pair reference
4. comments: -calculated delay for spots
5. main pair reference
6. comments: -calculated delay with strings 2 db lower level
IV. Mozan Oboe Concerto: 2 excerpts
2 woodwind, 1 bass, 1 solo (oboe) spot microphones (main pickup included in both)
1. comments: -no delay to aIl 4 spot microphones
2. comments: -calculated delay for al14 spot microphones (see table 1)
V. Bruckner: Symphony No. 7(not 3 as stated on the tape): 6 excerpts
2 woodwind, 1 bass, 1 percussion spot microphones
1. main pair only -AKG 414's (Blumlein)
2. comments: -no delay on 4 spot microphones
3. main pair only
4. comments: -calculated delay on all spots (see table 1)
5. main pair only
6. comments: -delayed with change in level for bass and percussion spot (bass up 1 db: per:,
down 2 db)
67
..
VI. Bruch: Violin Concerto: 5 excerpts
Solo Violin Spot Microphone
1. main pair only -AKG 414's (ORTF)
2. solo mie added; comments: -no delay
3. solo added; comments: -calculated 12 ms delay
4. solo added; comments: -10 ms delay
5. solo added; comments: -15 ms delay
VII. Bruch: Violin Concerto: 5 excerpts
Woodwind Spot Microphones: all include main pickup
1. cornments: -no delay to spot microphones
2. comments: -19 ms delay ( 19.5ms actual calculated delay)
3. cornments: -20 ms delay
4. cornments: -23 ms delay
5. cornrnents: -15 ms delay
VIII. Bruch: Violin Concerto: 3 excerpts
2 woodwind, 1 bass, 1 percussion spot microphones ... all include main pair
1. cornrnents: -no delay to spots
2. comments: -spots added with calculated delays (see table 2)
3. comments: -same as #2; yet woodwind spots down 2 db in levell bass up 1 db
68
IX
,jler: Symphony No. 1: 4 excerpts
4 percussion spot microphones
1. main pair only -AKG 414's (ORTF)
2. spot mlCS added; comments: -no delay
3. spot mics added; comments: -calculated delays (see table 2)
4. spot mics added; comments: -calculated delays but lower level by 2 db.
X. Mahler: Symphony No. 1: 4 excerpts
2 woodwind, 1 bass, 1 timpani spot microphones
1. main pair only
2. spot mics added; comments: -no delay
3. main pair only
4. spot mics added; comments: -calculated delays (table 2)
LISTENING TEST B
DIFFERENCES IN SOUND RECORDING METHOD
1. The following 5 excerpts are used to demonstrate the sound quality obtained by using
various miking techniques and 10 clarify the effects of delay compensation. 1 have included five
excerpts from Stravinsky's L'histoire du soldat, using 3 miking techniques plus auxiliary
microphones and delay compensation effects.
1. Stereo Microphone
XIY recording technique (level difference); comments:
2. Blumlein recordmg technique (level difference); comments:
.,. Dummy Head technique (time difference)~ comments:
4. Dummy Head with spot microphones; comments:-no delay
5. Dummy Head with spot microphones; comments: -calculated delays (see table3)
69
,
LISTENING TEST C
,
1. main pair no.1 -Dummy Head (using B&K 4007 microphones at ears)
2. main pair no.2 -AKG C-422 Stereo microphone
3. main pair with spots -AKG Stereo microphone with spot mies. ( no delay)
4. main pair with spots -AKG Stereo microphone Wlth spot mics. (calculated delays-see table
4)
LISTENING TEST D
This test is based on an A-B comparison test.
1. comments: A-main pair; B-with woodwind spots (no delay)
2. comments: A-main pair; B-with delayed woodwind mies.
3. comments: A-main prur; B-stnng spot microphones (no delay)
4. comments: A-main pair; B-woodwind, bass and solo spot microphones (calculated delays)
5. comments: A-main pair; B-woodwmd, bass & percussIOn spot mies. added (ca1culated
delays)
6. comments: A-main pair; B-solo mic added (no delay)
7. comments: A-main pair only
B-main pair with spot microphones (calculated delays)
NOTE: excerpt no.8 mcorporates the main paIr and spot mkTophones in bOlh A & B!
8. comments: A-maIn pair with spot microphones (no delays)
B-main pair with spot microphones (calculated delays)
70
APPENDIX C
RESULTS-PREFERENCES AND COMMENTS
LISTENING TEST A
I.Mozan Oboe Concerto: Woodwmd spot microphones (listen for hom entries)
8 excerpts (all odd numbered excerpts are the same ...comment on the differences you hear Wl.·
the even numbered excerpts which include the woodwind spot microphones in combination
with the main pair)
1. main pair reference-A KG 414's (Blumlein)
2. comments: -no delay
~.
main pair reference (same as 1)
4. -calculated delay (18 ms) ... 13%
comments:
5. main pair reference
6. -shortened delay by 2 ms (16 ms) ... 20%
comments:
7. main pair reference
8. -longer delay (21 ms) ... 67%
comments: "French horos rounder, oboe fine here" .... recording engineer
"Homs sound "best" here, but may he too prominent (as in loud), but finest
sound." .....recording engineer
"less present.. .. horos further back.... most delayT' ..... sound engineer
"more ambiance with horos, low strings good" .... musician
"brass section should be more prominent".......composer
"more strings than before? .. this is the oost natural balance" .... composer
II. Mozart Oboe Conceno: 7 excerpts
String Spot Microphones
.......... too strident.. .. 5%
"strings sound too strident throughout all examples for my
taste" ....composer
1. main pair reference
2. comments: -no delay ....
"strings advanced" ... composer
"enhances the viola tine but not advantageous to
others" ... composer
"horns louder because of spot mics"... recording engineer
.l main pair reference (same as 1)
4. -calculated delay (19/16 ms) ..... .55%
comments: "louder string tutti' s, almost distorted; more depth with first cello".... musician
"flatter dimension"... recording engineer
"delayed string spots seem to make oboe recede, adds depth" ..recording eng.
71
5. main pair reference
6. -shorter delay (16/13 ms) .... 40%
comments: "most definition ... after hearing the delay in no. 4 this would seem the tlattest
sound" .....recording engineer
"good balance" ....composer
"better balance" .... sound engineer
"clearer in lower strings (viola, cello)'· ... composer
"very slight difference again, but strings sound "muddled" ... individual
instruments less distinct. Not as clean or "gcxxi" a sound as no. 4.
7. main pair reference
III. Mozart Oboe Concerto: 6 excerpts
Woodwind and String Spot Microphones
1. main pair reference
2. -no delay added to 4 spot microphones
comments:
"french homs present, strings clm,er, not much depth" .. recording engineer
"bnghter, much more definition of french hom, clarity of oboe" ... composer
3. main pair reference
4. -calculated delay for spots ..... 20%
comments: "more depth, defmition of slrings very good" ... recording engineer
"like no. 2 but strings seem better placed, int~grated mto overall
pickup" .. recording eng.
"noticeable delay in string mics." ... recording engineer
5. main pair reference
6. -calculated delay with strIngs 2 db lower leveL. ....... 80%
comments: "like no.4 but better balance".... recording engmeer
"less woodwindlhorns, oboe sounds better thls time; mecr balance betwecn
sectIons" ... composer
"better defimtJon but flatter depth of field" ... recording engineer
IV. Mozart Oboe Concerto: 2 excerpts
2 woodwind, 1 bass. 1 solo (oboe) spot microphones (main pickup included in both)
1. -no delay to all4 spot microphones ... 23%
comments:
"wmds (and brass) are better integrated Wlth strings in excerpt nO.l
-ln terms of space and timbre as well; strings sound lighter" ... recordmg eng.
2. -calculated delay for aIl 4 spot microphones (sec table 1)..... 77%
comments: "has bener depth of field-but sorne funny ambil!nce slappmg
72
"
(delays?)" .. recording eng.
"best, hear more of the high strings"... composer
"we hear the bass better" ... composer
"oboe reclines-distracting" ... recording engineer
"no. 2 has more clarity and spaciousness and sounds better" ...recording eng.
"oboe tloats, horos are louder, strings are more defined but not
better" ... recordmg eng.
V. Bruckner: Syrophony No. 7(not 3 as stated on the tape): 6 excerpts
2 woodwind, 1 bass, 1 percussIon spot microphones
1. main pair only -AKG 414's (Blumlem)
2. comments: -no delay on 4 spot microphones
3. main pair only
4. -calculated delay on all spots (see table 1).... 25%
comments:"Woodwinds much better. Nicely imaged and balanced. Bottom too heavy but
not 50 muddy" ... re('ording engineer
"Woodwinds sound "behind" the rest of the orchestra in terms of space, but
timbre 15 nice" ... recording engineer
5. main pair only
6. -delayed with change in level for ba;,s and percussion spot (bass up 1 db: pere. down 2 db)
75%
"more of no. 4 (clarinet and flute more defmed, trompet richer, strings
round)ith percussion focused now"... recording engineer
"heard the hom thlS time"... musician
"c1eaner, clearer" ... composer
"bottom end and timps best in this version. Clearer, and well
blended/balanced.Woodwinds now sound too dry and forward" ... rec.eng.
"tighter low end" ... recording engineer
73
VI. Bruch: Violin Concerto: 5 excerpts
Solo Violin Spot Microphone
1. main pair only -AKG 414's (ORTF) .........07% doubling unpleasant yet #5 more like
chorus
2. solo mie added; comments: -no delay ...... 07%
"solo mic causes vIn. to move to centre~ space is different for the solo compared to the
orchestra'·... recording engineer
Hrosin noise dlstracung from soIOlst" ... recordmg engineer
"this seem3 to have the hveliest sound" ... muslcian
3. solo added; comments: -calculated 12 ms delay ... 07%
"viol in too present" ... same comment by two recordmg engineers
"this sounds a little unnatural- hard to say why-confused vlolin
positioning?" ... recordmg engmeer
4. solo added; comments: -10 ms delay .....64%
"sound of the viohn is !>omewhat between Ex. 2 and 3 (both in tenns of timbre and
spaClOusness; orchestra gain, depth and nchness is more mtegrated with the
soloist" ... recording engmeer
"this sounds very ~imilar to no. 2 (viohn ncher, blgger, and evcn more forward: bull
like this better)-coherent image" ...recording engmeer
"Image stable for vlolin, now plZZ. is really audible in orchestra, violin
rounder" ... recording engmeer
"more resonance, more 'aIry"' ... muSICtan
"more clarity for woodwmds" .. recordmg engmeer
5. solo added; comments: -15 ms delay ... 5%
"vln tone munder-and clearer" ... composer
"vin tone seems more rounded" ... composer
"doubling less "unpleasant" here compared to excerpts 3 and 4~ almost like
chorusing ... recording engmeer
"lower stnngs get lost; orchestra is not as defined" .. recording engineer
VII. Bruch: Violin Concerto: 5 excerpts
Woodwind Spot Microphones: all mclude mam pickup
1. comments: -no delay to spot microphones ..... 15%
2. comments: -19 ms d~lay ( 19.5ms actual calculated delay) ... 20%
"woodwinds clearer than no. 1" ... composer
"more of brass and timpani"... composer
"woodwinds sound 'supported' by spot microphones" ... recordmg engineer
3. comments: -20 ms delay
r,
4. comments: -23 ms delay ..... 45%
"airy violin, orchesrra too boomy" ... musician
"best woodwind balance" ... musician
"best balance here; percussion good" ... recording engineer
"1 like this balance between clarinet 1vin. soloü.t" ... recording engineer
'''slap' of delay not so noticeable here" ... recording engineer
74
5. cornments: -15 ms delay .... 20%
"woodwinds increasingly more present (no. 3-5)"... composer
"a bit muddy" ...composer
"hke no. 4 (wmds weIl mtegrated and positioned, still very focused and clear)-but a
little better; 1 hke the sound and balance of this one the best"... recording engineer
"Balance of an groups best here" ... recording engineer
"flute too promment, clannets 10st, brass loud" ... recording engineer
VIII. Bruch: Vlolin Concerto: 3 excerpts
2 woodwind, 1 bas~, 1 percussion spot microphones ... all include main pair
). comments' -no delay to spots-lO%
"most appealmg of three excerpts"... composer
"best" ... composer
2.
comment~:
-spots added with calculated delays (see table 2) ... 10%
3. comments: -same as #2; yet woodwind spots down 2 db in leveV bass up 1 db ... 80% all
engmeers included here
"better than no 2; especially percussion"... recording engineer
"good depth and ~olo blend"... recording engineer
"Ilike thIS best; wmds natura! and not too forward-well blended. Bass is weak though
and timps a Intle muddy" .. recording engincer
.. ~till not a lot of bass defininon; clarinet and timpani bener here" ... recording eng.
"best blend" ... rccordmg engineer
"depth of field seems best here but left heavy ambiance'· ... recording eng.
75
IX. Mahler: Symphony No. 1: 4 excerpts
4 percussion spot microphones
1. main pair only -AKG 414'8 (ORTF) ..... 15%
"lower timpani very close but 1 liked thlS excerpt best overall" ... recordmg engineer
"In terms of volume and spatial balance. 1defimtely prefer the mam pickup~ in temlS
of percussion timbre and sound, excerpt no. 4 ~eem~ to best ~lIppon the percussion
instruments, but they are way m front of the rest of the orche~tra and much louder.....
..recordmg engmeer
"best-rest have unnatural close percu~~lOn sound whlch takes away l'rom the re~l of
the orchestra" ... composer
2. spot mics added; comments. -no delay ... m
3. spot mlCs added; comments: -calculated delays (see table 2) ... 10%
4. spot mics added; comments: -calculated delays but lower level hy 2 db ... 75%
"better percussion depth of field-thls actually sound.., better than the main pair"
... recording engmeer
"best balance" .. recordmg engmeer
"'this sound~ best: percussIOn clean and defined, better placed-not ~o
forward" ... recordll1g engmeer
"this is preferable, sound IS not choked" .. recordmg engineer
"percussion ~eems most mtegrated" ... compo~er
X. Mahler: Symphony No. l' 4 excerpts
2 woodwmd, 1 ba~s. 1 timpam spot mlcrophone~
1. main pair only
1. spot mICS added; comments: -no delay . .40%
"bass and timpani clearer-best"... composer
3. main pair only
4. spot mies added; comment~: -calculated delays (table 2) ... 60%
"strongeJ hom secùon, stronger crescendo in timpam" ... musician
"less presence than excerpt 2" ... musician
"good balance, more definitIon" ... rccording cngineer
"definitely the best !o.ound. Tlmpani improved, bas~ ~till weak" .. recording engineer
"bass sounds up front, l hke that french hom~ sound choru'i-Y from the delay" ..
recording engmeer
"more woodwinds and tlmpam" .composer
"woodwinds are less present than in excerpt 2 hut ~till more present than ln excerpt 1.
Bass section ~ounds more sub~tantIal than in both cxcerpt~ 1 and 2 , (impanI ~lmilar to
2 but clearer. 1 would choose no. 3" ... recordmg engineer
"best-timpam better in both excerpts 2 and 4-bass too clo~e?" ... recordll1g engllleer
.
76
LISTENING TEST B
DIFFERENCES IN SOUND RECORDING METHOD
1. The followmg .) ex ;erpt., are u~ed to demonstrate the sound qualny obtained by using
vanous miking technll;'le~ and to clarify the effect~ of delay compensation. 1 have included five
excerpts from Stravm~ky '<, L'histOire du ~oldat. using 3 mlkmg techniques plus auxiliary
mIcrophone., afld delay eompensauon effeet~.
1 Stereo MIcrophone XIY recordmg techmque Oevel dIfference); ~omments:
"~mall ~oul1d .,tage, Ilot mueh depth, dry" ... recordmg engmeer
2. DIumlein recordmg technIque Oevel difference); eomments: 15%
"good stereo Image. vlohn a bIt dlstant" ... composer
"horns more pre\ent, more depth, small sound ~tage" ... rec. eng.
"1 hke thlS better than excerpt l-Ies~ artIfiClal soundll1g stereo" .. rec. eng.
J. Dummy Head techmque (ume difference), cornments. 5% ... but nonces depth in 5
"better [han excerpl 1 but not so mce as 2" ... compo~er
4. Dummy Head WIth "'pot mIcrophones, comments:-no delay ..... 50%
"percuS~lOn more dcfined, vlOlIn clo~er, woodwmds are too present,
~tereo picture "hghtly dlstoned, my guess IS spot mlking with no
delay" .. reeordlOg engmeer
"better than excerpt no. 3'- .. composer
"better images of separate mstruments" ... muslclan
"very wide yet pereus~lon better"... recordmg engmeer
"cleaner In percussIOn, dry sound" ... composer
"sirmlar to excerpt no. 3 (flallmage), but lmproved placement of viohn.
Aho seern~ hke nore depth on percus~lOn and brass. Quite
Ilkeable" .recordmg engmeer
"exœrpt no. 4 is best followed by no 2; woodwmd~ too distant 10 all
~elecuon~ (e~peclally bassoon), but clannet and bass are better in
4". recordll1g engineer
5. Dummy Head wlth spot nucrophones; cornrnents: -calculate.d delays (see
tahlc 3) ....... .40%
"defmmon 1~ therc WlthOUt lo~s of depth, ombre of instruments is not
dlstoned by do~e mIes wlth delay-Ievells cruclal" ... recording engineer
"Ies~ effectIve than exccrpt no. 4" ... muslcmn
"not much dlfferent From excerpt 4. 1 preferred 4 but can't pinpoint the
reason. It was more 'natural', but thlS sounds a latIe
deeper" ... reeordmg engmeer
"hest- appropnate for thlS musie" ... composer/muslcian
"
77
LISTENING TEST C
1. main pair nO.1 -Dummy Head (usmg B&K 4007 mIcrophones at ears)
"piano, bass and viohn sound 'out of place', too far back. ln general, the stereo in13ge
is indistinct, hard to tell where each Instrument IS" .recording eng1l1eer
2. main pair no.2 -AKG C-422 Stereo microphone
"cOlPr:ident? No separation, but good fre4uency rco.;ponse" .. recordmg cng
"mue better sense of space and depth than 00 1 Stnng~ hetter defined, piano more
present. Stereo Image more coherent, very precl,e placement of ~ound ~ourœ-;, ,tl~o
good balance between Instrument)). Sound clear and natural" rccord1l1!! eng1l1L'er
spot~ AKG Stereo fillcrophone wnh ~pot mlC~ ( no delay)
10<4
"spot 00 bass? more reahstlC, good sound, much clanty-beautlful" wmpo~cr
"flaner stage (le~s sens~ of depth and space). Piano and bas, hoth more pre-;en! Jnd
up-front. Not as 'spot hghted' al) no. 2 m tenn, of precI~e tn~trument placement, hUI
still easy to detennme where each point ~ourcc \,," rccord1l1g cnpnccr
"my gue~s l~ no delay to ~pot nue", clannet ..,how, comb liltcnng-,ollllth dul\ on 1(1\\_1
frequencies, ~ome bras . . float<;" rel'. eng
"cello clearer, strlng~ clearer, french horn clearcr, piano a hlt wcak, ha, ... e~ heller-mOll'
defined'·. composer
"stnng section IS clearer now" . compmer
"hlghhghts cenmn strand... of the melody but texturc 1.., k ... ~ homogcnol1' than c\..:crpl
') "
.
_ ... mUS1Clan
"narrow image-ail m~trument~ on top of each other". rec. eng
3. mam pair wlth
4. mam pmr wnh spots -AKG Stereo rmcrophone WIth
~pot
mIe, (calculated delay'-'>cc la!
4) ...... 90o/c
"now we have depth! Clannet IS dIstant hut defined, frequency fe'iponse ,eem\
accumte. Wlsh ma.m paIr had been omm~-cotncldent due ... n'( !-.howca~e lhl~
demonstrauon as weil" .recordmg engmeer
"best, a11 tImbres are c1earer". composer
"preferable to excerpt 3" .. composer/mu~lclan
"better balanced Imagmg" ... muslclan
"wlder Imagmg and spaually bener" ... recordlOg englOeer
"low oboe or bassoon IS louder, WhlCh 1~ I"cer-(unplea'oiant room ,ound on oboc
later)-has the most detail"' ... recordmg engmeer
"This sounds be~t, most natumL Good .:\tcreo spread, allmstrument\ c1ear, pre..,ent
and defined, WIthout over ernphasls. A bIt of "butld up" 10 the mlddk,?". rec. eng.
1
78
1
LISTENING TEST D
1. comments: A-main pair; B-with woodwmd spots (no delay) ........... 100% hear differer . t
"No depth in A" ... composer
"Quite a sl'btle dlfference. Wmd~ a httle more defined in the B excerpt, but sound too
dry for the string~ (I.e. the wmds seem forward in comparison Wlth the
'Itrings)" ... recordmg eng.
"Liked the bit more french horn in D'· ... rec. eng.
"not much difference ... no delay?" .. recording engmeer
"Woodwmds better In B".. composer
"Can hear sltght dlfference; perhaps lower Instruments a bit more prominent in
13" .. composer
"oboe louder ln B" ... recordmg engIneer
"more orchestra ~ound in B" ... composer
2. comments. A-mam prur; B-wuh dela)ed woodwind mics
a......... 20% prefer main pair
"stnng section tno present In B" ... composer
"prefer A" ... composer
B.......... 30% prefer delayed woodwmds
"B has delay; more depth" ... Andre
"woodwmd delay? French hom sounds mce, deep" ... recording engineer
"Improved definition anri tone of solo oboe ln B. Strings sound muddled compared to
mam paIr refeœnce. Hm ''\ and wll1ds may be a bit too loud?" .. rec, eng.
"beuer depth In B" .. composer
50% unsure
"sound IS the same" ... rec. eng.
"hear no change" ... composer
.1 comments: A-mam pair, B-smng spot microphones ( no delay) .......... 80%
"B-betler base response" .. muslcian
"B-woodwmds closa" ..,rec. eng.
"B-has close InICS-nO delay" ... rec. eng.
"B-audible chorusmg fmm stnng spots-delay?" ... rec. eng.
"B-can't really hear a dlfft'rencl? Except the violins (hard left!) may he clearer and more
defmed In B" .. rec eng.
'1 .. 20%
"A and B sarne ..... composer
"no dtfference".. composer
"first a bIt bright'! Left speaker emphasized"... composer
4 comments: A-main pair; B-woodwind, bass and solo spot microphones (calculated dela> l
13 .. 70%
A ..... 20%
?( uns ure ) 10%
"solo oboe more present" ... composer
"not much difference. Strings a bit richer, smoother, better blended in B"..ree. eng.
"homs spot microphones-no delay? Didn 't mind the up fruntness" .. rec. eng.
"S has delay" ... recording engineer
"too much delay" ... composer
"oboe and strings stronger in S ..... rec. eng.
79
"too much delay" ... musician
"second may be slightly preferable to first" ... compoj.::r
5. comments: A ·main pair; B-woodwind, bass & percussion spot mies. added (calculaled
delays) .........90% (10%-?)
"b IS better" ... mUSIcmn
"b-woodwmds and horn~ stronger" rec eng .
.ob has no delay" . ree. eng
"woodwmd lmes clearer In b" ... ree. eng
"unsure" ... ree. eng
"perc. louder in b" .. composer
"b-brass, pere. and bas~ delayed" ... composer
"shghtly more detail In b" .. composer
6. comments: A-maIn pair, 8-s010 mlC add~d (no delay) ........... 60%
hear difference .. .30%; no difference .. 10%
"can't heardifference although ln B vin may be hIghlighted somewhaf' ,composer
"m A french hom (brass) better" ... composer
"a-very distant; b-closer" composer
"better solo vlohn sound In B, but s()und~ too present and up front, also too dry. Overai
better sound, though" ... ree. eng
"vlOhn spot not blended" .. rec. eng
"B-no delay; timbre IS affected" .. rec. eng.
"viohn closer" .. compo~er
"B-so10 stronger (different space); 2nd vlolins stronger; vlohn ~()la-mlddle" .. rec. eng
"more strIng sound" ... musIcmn
7. comments: A-mam paIT only. 10%
B-maIn paIr WIth spot microphones (calculated delay")
no dif..60%
"seems ta have less pre~ence". muslcmn
"no difference" ... rec eng
"little difference" ... composer
"no differenee" .rec. eng.
"?" .. rec. eng
"no difference " ..composer
"B more balanced" .. composer
NOTE: excerpt nO.8 mcorporates the mam pair and spot
.30~
mlcrophone~ ln
both A & BI
8. comments. A-main paIr wnh spot microphone~ (no delays) .. 35o/c
B-maIn paIr WlIh spot microphones (calculated delays) ... 35%
hear difference ... 30%
"A has more depth than B" .. composer
"B-instruments are more separated" ... composer
"solo inst seem not as Întrusive In B" .. composer
"A-could have used more piano; B-hked the harp ~!Jot" ... rec. eng
"B-close mics-no delay" ... rec. eng
"B-soU! js more distant but also more dramatIc" ... composer
"violin, harp, plana-stronger, c10',er" .. ,rec. ~ng.
"better strings in A" ... musician
,.
80
APPENDIX D
LIST OF GRAPHS
Page
Graph
1. Eventide Ultra-Harmonizer H3000 Frequency Response
.
82
2. Eventide Ultra-Hannonizer H3000 THD+N vs. Frequency ..................
83
3. Eventide Ultra-Hannonizer H3000 IMD vs. Frequency........................
84
4. Roland E-660 Digital Parametnc Equahzer Freq. Response . ....... .........
85
5. Roland E-660 DigItal Parametric Equahzer THO+N vs. Frequency .........
86
6. Roland E-660 Digital Parametric Equalizer IMD vs Frequency ..............
8-
7. Lexicon 480L Dlgltal Effects Processor Mach.A; Out-Left Freq. Response
8~)
8. Lexlcon 4XOL DIgnal Effects Prex:essor Mach.A; Out-Right Freq. Response
~~"
9. Lexicon 480L DigItal Effects Processor Mach.B; Out-Left Freq. Response
9,'
10. Lexlcon 480L D1gital Effects Processor Mach.B; Out-Right Freq. Response
9"
11. Lexlcon 480L DIgital Effects Processor THD+N vs. Frequency ...........
9::::
12. LexJcon 480L Digital Effects Processor IMD vs. Frequency .................
9:
'
81
--EVENTIDE ULTRA-HARMONIZER H3000
FREQ. RESPONSE
1--TT]-,I--' r- ----l- r--] -T'llT!
811-ill--H-~+~l
--- -!--,r ,-:-+tH-----J1--1---j--'-ll-i -:-"''-1
lit i IJ 1
Il! ;,
r-" --rTTn i---- -r--l- -1lTi j i-- ----1- -r-î TITI Tr---
10
r--j--·r-r-i--r-m
1
-
Iii' :
1
6
+R-rr,-------r
Il
l
1
-rr------l-----l-
t----t--t--I- -r--i-j-.] -;- -- ---- : --- i-1-,
2
_
(Xl
N
QI
~
0':
l ,'
1
t-I
l"
il!, !
r--- 1I--;-T1-r~--1
l,
-8 1
1
1
1
-10 L ____ 1
20
1
! ~
1 j
1I
1
-I-j -(! --- -- -- !
11 : '
i 1j
l
1 1':
1
\
- -
:
1'
1
-- -
-rJ
1
l "
1
l
1
1
-
1
11
~-
--
1
i
1 -
-,'--rr r ------;
Il!
il1- -- ----i1
j-
!
1 -
t t
i
!
1
1
1
1
l'
,
:!
Il
: ,
1
1
1
l
i
,
1
1
1
i i
:
1
l,
1
1
:
1
l
"
1
:
!;
,
il;:
'"
i
1i
, !
1
1
J
!
1
-----J
i ; :-; . ~- ---- ~
-:~-~-rr:-----1
\ i i
li i i , ' 1 il;:;
l' ; I i i
1
l L_LJ- j 1______ 1 __ -1_.-1 l 1 .1.1.l L - - - - - - L_.-1. -- _L-l _.4 Li - -----~
!
1
1
_
i
-li ---1--1
--1- -rr!";---- --:--T-I-r n--iTf-----I--r-t- i 1- -- ---- -i --- -t - r-,-t-r-j -rî ----:
11!
!
l'
1; Il!'
---t--i-m'
""7-,-~l----~-----:-~---i-,--t-t1-t--
11- ~--: -; -,- -- -._;
,IL
1 \
1
t
'
-+-,1 r i : - -
1
I l !' 1
r-r--l-'--fi-','
r--r--r-- ---- -,! -l--1'-~I'
1
-6
i l ,
I ll
L'
'
i
1
1
-4
1
Iii, :
\
1
-2
il
1: ;
1
50
100
200
500
F~Guency
lk
;
2k
1
5k
Il
lOk
20k
,
EVENTIDE ULTRA-HARMONI7ER H3000
THD+N vs FREQ.
'--'-1 1 !~ i
+-8
~
l
1
'
r
A,
"
500m
A-
i
1
~V-~- ~-r
I~~
1
!
III
1
!
1
1
Î-I~
l' i ,
i
1
l
~
i-
r-t--
1
-/
L
IOœ
~O
1
~
Il
1r
;
1
i
12J~L
i i ii
j
'TIHL --
,'II
t
III
100
1
200
1
___
500
Frequency
lk
1
--
1
1
1
l
1
1
- - !
1
'1
50
le
"
1
t---tÎ'-~-t-i+'l----l
! i
1 1 l ' :1
1
1
l
TIl ,-1
1
1
1
,1 i 1
i
i
I..J
t==~h
.- [---1 --+-+
~-+-----
•
1
20mr
Iii 1
1
1
"r
I
Il
I-!
,il
ry-
8
1
1
1
1 l
~ ~ lOOm~_i-=l±">- §f
'Il
!:
1
1
1
---'--!
1
1
T~'~l",.-Â-"
Il
14
Iii
r
~ 8~--~ 4--+ 1tfi ; 11-==1
_1 _-+!
ji!illl
200mF
i
1
1
l
,
"
i
~llJ
2k
5k
1111
1 1
1
1
10k
1
20k
T
1
CJ
C)
1
CJ
en
:r:
_t_~~~-J
1
1
_-_+---f---_=_= l
i
jl
o-+--4----4--H-+---+--
0::::
----4
w
a
-
N
~
:z
•
«
:c
LL
<
>
CJ (3
:E W
0.::: 0::::
1
0::
(J)
1
t
t-++-t-tt-t--t---+---~---+++-J...-+-.+--+- --f-----_f_ -
J- 0
-.J :E
:::J .........
w
Cl
.........
J"c-
W
>
W
-I-t--+---+---+--~~~+_
- - f - - ~---
r-
1
tt .
1
~(ï-~--
1
f-rt -f----f--i--J lL_~-L-l
1 l ,
1
1
1 1
1
1
-
1
1
1
;l
&
a
&1
(%)
84
-
31dHS
mu
1
-----1
----4 ~
~
'\
ROLAND E-660 DIGITAL PARAMETRIC EQUALIZER
FREQ. RESPONSE
la
8
[TTTT1Trr-----r-1i-mrn--T--T-rm,
-m---l
"1-- ~-T-+ I--~-1-~-r-ï/~----l
-----t----,, - ,
1
1
6
1
i
:
Il!
VI
~
~
a I-!
1
1
-2
l
,
-10
1
1
1
l'
1 1
1
i, I" l l
1
1
li
1
1
l
1
1
,.'
.-~ -------1----.--1-;--
TrT'
r--t--------i--t-r---'r-r:Î
l'
1 Iii
I---I--Ll-~-- ---
,
; !i
l.
,"
1
1
---
l
1
f
1
11
1
,
!
.
1 1 1
1
1
l
•
1
1---1 --~_ --'
l
'
,
1
:J
'1
l
l
1 ..
-L .
1
j -
1
'l'"
,1
l' 1
t
1
------l---t- Il!
r+i-h-h----- - ï---1---I--t--h -;!T--
iI
ï T-
t
-!
•
1 1 1 1
1
----- - -- --J--+--l ; !-- - ---- ---
!\
I---L, ---i.-i----i-+-!-LJ-_---l
' i
fT
l
i i 1
'1
-L---L--l
1
1
l
1 111
1
l,
1 i
l,,!
1 1 l
'
-T!--~~--------!-;--J---+-j-~-~-.LL----~1
, • il
1
1 1 1 Il
j-I---r-rnl"-------i-rrr
T:-r ,---d-r-r--i-r-rTrfT -.-'
L_
11..._______
1J_.1
l-1-111il ___J
1
-8
!
'li
1
: 1 1
~-r '----f--! -; ;
f--
1
-6
l
!
1
!- - t -
l
-4
:
Il!
Il
-r-1TrIT ------r--r--r- -TTiTi-----Til : i I-nl--
'- -1
l
'
1 1
--T--
00
l '1 '
l
' l , 1j 1; !\
Il! !
IIi
i--~-'-Ifll-1ft --- - i -1 -litiTn--
ri-r- r:T ! Ilr~----I'-t
i TT r-nr---u---t- ,-O--!-T' t-rl'-H------Il
1
20
!
1
! , I i1 i
I
I!
1
1
J
l
1__
J_L.1_
50
'I!
I l !
100
200
1
!
I l !
1 l' 1 i
Iii
_-1
500
Frequency
,
1
1________ 1
lk
2k
i
1
1
i
1
1
1 1
fit
1
1
1 !
1
1
__
5k
lOk
20k
0:::
W
N
~
-.J
«
:::J
CI
W
U
~
0:::
~
W
•
2.:0
<w
>..
0::: 0::::
< LL
0...
o
'C
~
(f)
-.J
>
<
~ >?
+
- --
........ +
Cl
........ :::r:
C)
-1
g- - ----- lU')
8 Ll:
- --- -----i
---l
.~Il -r-- -Î~
D~
a
<D
CD
1
f·
-
--
W
o
:z:
<
-.J
C)
cr:
1
--t--
f-I--:;· }-1 -.. (. .-------~ ~
i -!
j- .-- -. --- ---1
.-~ - ~- T--n h-I-·
-1-.
t
::r-t-- ,-- ----1
i
t i____ _ L. ___1______ _
-
L -
-
LLL
B
~
ri
~
Ei
§
(%)N+OHl
86
Ei
~
B
~
~o
S"J
~
l-----rrrT:-:
~ ~
rr: ,Tri
1i
l'
'1'
n::
w
1
1
~
~
<
1
LJ
~
•
. .c a
0::: w
-< 0:::
D- li-
1
1
f
'i
1
1
1
i
1
1
1
!
i
Il! i l :
1
1
li
l"
1
1
lIt
1
1
1
1
il ,'
1
1
1
l
1l ' :
1
+1.
(f)
1
1
--!------LL--t
1
1
+-1 :-t-l-~----~
Il1
IlIl
1-7
III
!
II!
i
!
i
~
!
i
t
Il!I_I~I
1
l
Il1 1
'
l
'
rt~ , -, r--- r----t------1
1
l)
,.
1
1
l
l\
'\
1
1\
o
-,7
.c_
<
-.J
C)
0:::
l
"
l
1 1
l '\1
.
1
fi,
l
l, ' ,
'I! 1 l
l
~~LliJ_
~
B
\
'
1
III
1
!l,
,:
1
!
1 1
1
1
1
1 •
l'
1
1
' :
\ ,
1
~
'I! ~
ttt+,- '-I-j--t-~ l
Il Il
1
T
1
1
.....-.
C)
(0
1
1
1
;,
CJ
W
1
1
1
........... 0
CJ ::E
co
l '
Il
rL++-r-t----r---
« >
r~
Iii
:
1
~­
---.J
1
Il
'1;
1
---
0::::
:2;-
f
t i 8
IT:-n-I--t-T--t--1tt!T~ "tf-+- r-i
fq : 1-i -r 1-ffiT1-· --- t- -----1
rtf-:,:-'Ii-- L/: +T-~l:--~-f+ +j'-j_··--t--·--î
,il,
'Il
W
1
1 .
1
Il
:=J
G
l'
1J
1
Il
N
LLJ
!1 Il
\
!
1
'
-l~
~
il:
cS
c3
Cl
a
N
-
tU____~_------II.x
(%) 31dHS OHI
87
1
Il
l'i
~
~
N
~
_
* *
:Ir
THE
CE N T R E
8 ANF F
TEST No.
l ~ li
l
** S T
i
~.
RAT l 0
'*
t
5 N ti
c) :\ -
*' .. ~
~ N : 9 (.1 1
LExrCON 480L DIGITI\L EFFEi...·rs PROCESSÙR
MACHINE A OUT-LEFT
Time
l n PLI t
:6 No,
10:55
90
A
ARe f
fi' M S
1 [) • S
H 1. Q h F' as:;
0 e <:.
l • ,) <\ l
d 8U
0FF
L
(j
,
li
1
1
c.:
1: ù r-
w i= ~ s:; 0 F F i 'i 0 t
i
1
<2
,1 { ::
j
!
1
1
1 If
1 i
1 1 1 1 1 Il
1 1 Il III
1
i1 i
1
Wt=
,,
l. .J
t
i
r, t '= ,j
1
1 Tl
1 1 11111
1
1
1 1 11111
1 1
I~
1
l', j
,
1,
ii
1
Il
1
1
1
Il
III
1
1 1
,
1
1\
'II
1
1
1
1
1
1
1
q dd
+2 dB
I,·ttttt
I
l
1-(-[
1
1
Ill'II1
1
Il
1
da
1
-
1
10
1 1 1 l'
1
---!
1
1
1
1
III
o dB
-2 dB
1
1 1
1 1
-4 dB
-6 dB
1
-B d9
'3
ï
-10 db
10 Hl
lQO Hl
RATIO
-
-
-
-
-
-
-
-
1. 09
1 kHz
10 kHz
FREQUENCY
dB
1.01 dB
0.51 dB
0.41 dB
0.21 dB
0.02 dB
0.09 dB
0.09 dB
0.09 dB
0.03 dB
0.01 dB
0.00 dB
0.(11 dB
0.01 d8
0.07 dB
20.004 I<Hz
~O.OO4
16.004
12.504
10.002
8.002
4.(101
2.001
1.0(10
~H::
\1Hz
.Hz
I<H::
kHz
~ Hz
~H::
~Hz
500 H::
250
125
63.0
31.5
31.5
88
Hz
H:
HL:
H:
H.::
lolO kHz
***
TEST No.
Time
In pu t
A
*** ST
*** RAT l a ***
THE BANFF CENTRE
:
2
10J-AM29u1
tt*
LEXICON 480L DIGIT.\L EFFEC''rS PROCESSOR
MACHINE A OUT-RIGHT
26 Nov
10:56
91)
A
R t'1 S
o.? tee t
1.001 ~ H:::
10.8 d8U
Ret
SN#
Hiqh Pass OFF
Law Pass OFF
r
Not
.0 dB
Il
1
1
1111
1
1
t.J
1
1,1
1
q
dB
1
t2 dB
1
1
Il
1
Il
1
1
1
1
l'l'
1 U"I
" "
t
1
1
o J&
1
1
rI!
-2 dB
-4 dB
-b dB
1
-6 dB
s
ï
-10 db
1
1.
10 Hz
100 Hz
RATIO
-
-
-
-
-
-
-
1 kHz
10
FREQUENCY
0.99
0.46
0.::4
0.12
0.06
dB
dB
dB
20.004
16.004
~H;:
1~.504
kH;:
dB
10.002
~
0.(1)
dB
dB
(1.01 dB
0.01 dB
(1.03
dB
0.05
0.05
0.05
0.(14
0.02
dB
dB
dB
dB
dB
~H2
H4
8.002 " H=
4.001 "H::
2.001 ~ H ;:
1.000 "H;:
500
250
1:25
6'3.1
31.5
31.5
89
H:::
H;:
Hz
Hz
H::
H::
~Hl
100 kHz
.... *
THE
BAN F FeE N T R E
TEST
No.
1
***
Tlme
:
11:~5
~6
InPLtt
A Ret
H1Qh
RAT 1 0
~
*.. 5 T SN" 11) "3 - A r'l ~ 9 1 **
'* ** LEXICON 480L DIGITAL EFFECTS PROCESSOR
MACHINE B OUT-LEFT
A
r..: 1'1 SOt? t
10.8 d8U
Pass
OFF
f.
(1
1.0U1
02 C.
t e..
t-
~H.:
Law Pass OFF
Not
~~e l Q
h t eJ
1
1
!Il! 1
III
1
1
1 Il
·ô
d&
+-J
d&
1
a
+2 dB
a
a a
1
Il
1
Î
1
Il
Il
,
III
o dP
-2 dB
1
1
1'1
-4 dB
-6 dB
1
rtl
s
ï
10 Hz
100 Hz
RATIO
0.68 dB
1.21 dB
1.34 dB
1 .54 dB
1 • 72 dB
1.65 dB
1.65 dB
1 .66 dB
1 .71 dB
1 .74 dB
t.75 dB
1.77 dB
1 • 19 dB
1
kHz
10 kHz
FREOUENCY
20 • (Ill ~ Hz
16.008
12.507
10.006
8 • O.) 5
4.003
2. (101
1 • (101
500
250
125
63. (1
31.5
90
kH:
" H;:
~
H:
~
Hz
kHz
~
Hz
kH:
Hz
Hz
H::::
H::::
Hz
-8 dB
-10 db
100 tHz
i
***
THE
TEST No.
Time:
~
***
:6
11:::;6
InPLlt H
A Ref
10.8 dDU
Hl.qh Pc:<\ss OFF
p
RAT 1
*** ST
a ***
Nav
qo
Law
RMS
De t:,?,_ tor
1 • <)1) 1 ~ H;::
Pas:; OFF
not
CENTRE
BANFF
a
SNtt
10}-Ar-t:901
LEXICON 480L 0 IG l'1',1\L EFr'ECrS PROCESSOR
MACHINE B OUT-R IGII'1'
Wel.::) h 1: t? d
fffiPTl
tttfti li rn
"
+6
1
Il
~I
i"l~1--'r I~~
liT
J i lfil
l l' 1"Il
1
!
III
1
1
1
1
dB
+4 dB
1
.
***
I~
\1
11
+2 dB
o
~B
- 2 dB
1
1
1
-4 dB
-b dB
13
-8 dB
ï
-10 db
10 Hz
100 Hz
RATIO
0.19
0.91
1.12
1.39
1.62
1 .62
1.65
1.66
1 .71
1.73
1.75
1.75
1.79
dB
dB
dB
dB
dB
dB
dB
dB
dB
1 kHz
10 kHz
FREOUENCY
20.009 ~ Hz
16.008 ~ Hz
12.507 kHz
10.1)1)6
-
8.1)1)5
4 • 1)0:'
2.001
1 • (II) 1
500 H=
dB
dB
::;50
dB
6:::.1
31.5
dB
" Hz
Hz
'. H:
~ H;::
~ H;::
~
H;:
125 Hz
91
H::
H.:
100 kHz
Lexicon 480L Digital Effects System (Twin Delay Program):
TI-ID measurement (&) versus Frequency
FREQ.
Mach.A-Left
Mach.A-Right
Mach.B-Left
Mach.B-Right
30Hz
60Hz
120Hz
250Hz
500Hz
lK
2K
4K
8K
16K
20K
.021%
.021
.013
.013
.013
.013
.013
.013
.013
.013
.012
.013
.61
.016
.016
.016
.016
.016
.016
.015
.015
.014
.015
.61
.008
.008
.008
.008
.008
.008
.008
.008
.008
.009
.58
.021
.021
.021
.021
.020
.020
.018
.020
.70
92
***
THE BANFF CENTRE
TEST No.
1
1l:ZQ
Tll1\~
Inp'.lt
~,
.:..
H'lO
0.0217
No.
T~I1l""
.'
A
l ND
(1.0480
TEST No. 4
Time
I,1pLlt
11:30
A LEFT
Ü ~ "(. ?:
t ,,:\ r
LE',:E'_
l,:, • 3 d31J
'.
*
:t :At
IMD :I::t.
2,::'
N·:) ,
t MACHINE A RIGHT
'? (1
(;ïiS
D~ te c: ta rLE\!EL
10.8 dBU
%
;I(~*
IMD
t ~
26
Nov
9(1
11:::9
103-AM:901
N·::!v 'i 'J
A
In pu t
SN~
*.: MACHINE
"/
11: :9
In pu t
ST
RI1S
i~ 1:) •
Tl!l1o:!
TEST
:::!6
IMD
A
l i'1 D
0.(1597
TESr
***
***
*MACHINE
B LEFT
[j ete -= t
RrlS
LE·. . ·EL
l2 • 7 .j 8U
,.
.~
***
IMD
* *MACHINE
:t
'::3 r-
B RIGR"j,'
26 N·::!v 90
f;' t1 S
A
Detector
LE'./EL
IMO
0.0152 %
1:::.6 dBU
93
***
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement