an 26
US008369540B2
(12) United States Patent
Bhattacharya
(54)
(56)
AUDIO SIGNAL AMPLIFICATION
U.S. PATENT DOCUMENTS
5,677,962 A *
6,088,461 A *
(73) Assignee: NXP B.V., Eindhoven (NL)
Notice:
12/2009
PCT/IB2008/051896
§ 371 (0X1)’
(2), (4) Date:
Nov. 12, 2009
JP
JP
JP
07-240647 A
2001298336 A
2003036081 A
9/1995
10/2001
2/2003
OTHER PUBLICATIONS
International Search Report for Application No. PCT/IB2008/
051896 (Dec. 3, 2008) ddd.
* cited by examiner
PCT Pub. No.: WO2008/146189
Primary Examiner * Duc Nguyen
PCT Pub. Date: Dec. 4, 2008
Assistant Examiner * Kile Blair
(57)
Prior Publication Data
US 2010/0215194 A1
(52)
(58)
Garudadriet a1. .......... .. 381/104
FOREIGN PATENT DOCUMENTS
PCT No.:
ABSTRACT
A system for amplifying a digital audio signal comprises a
receiver 12 for receiving a digital audio signal, a level esti
mator 14 arranged to calculate the audio level of the digital
audio signal, a gain control 16 arranged to receive a gain level,
Aug. 26, 2010
Foreign Application Priority Data
the gain level de?ning the desired ampli?cation of the digital
audio signal, a logic circuit 18 arranged to calculate the head
space in the digital audio signal and to divide the gain level
into a scaling gain and an ampli?er gain, the scaling gain not
(EP) ................................... .. 07109248
exceeding the calculated headspace, a digital signal processor
20 arranged to amplify the digital audio signal With the scal
Int. Cl.
H03G 3/00
381/109
Seefeldt ...................... .. 381/107
2009/0323985 A1*
(86)
(51)
2/2012 Bjorn-Josefsen et a1.
6/2012
U.S.C. 154(b) by 737 days.
May 14, 2008
May 30, 2007
8,126,164 B2*
8,194,889 B2*
10/2001 Kon et a1.
8/2005 Anderson
(22) PCT Filed:
(30)
Harrison et a1. ............ .. 381/109
Linet a1. ..................... .. 381/104
2001/0026624 A1
2005/0175194 A1
12/599,814
(65)
10/1997
7/2000
Subject to any disclaimer, the term of this
patent is extended or adjusted under 35
(21) Appl. No.:
(87)
Feb. 5, 2013
References Cited
(75) Inventor: Puranj oy Bhattacharya, Bangalore (IN)
(*)
US 8,369,540 B2
(10) Patent N0.:
(45) Date of Patent:
(2006.01)
ing gain, a digital-to-analogue converter 22 arranged to con
US. Cl. ...................................... .. 381/104; 381/109
vert the ampli?ed digital audio signal into an analogue signal,
and an ampli?er 24 arranged to amplify the analogue audio
Field of Classi?cation Search ................ .. 381/104,
381/106, 107, 120, 109; 330/268, 274, 278,
signal With the ampli?er gain.
330/284
See application ?le for complete search history.
8 Claims, 4 Drawing Sheets
10
LEVEL
f
14
1s
ESTIMATOR
12
v
RX
f
LOGIC CIRCUIT
V
AUDIO
U
DAG
DSP
18
24
AUDIO
f
AMPLIFIER
22
an
26
US. Patent
Feb. 5,2013
Sheet 1 of4
US 8,369,540 B2
GAIN
SETTING
AUDIO
‘
DSP
'
DAC
‘
AUDIO
'
AMPLIFIER
V
Fig.1
GAIN
SETTING
AUDIO
‘
DSP
'
DAC
‘
AUDIO
'
AMPLIFIER
V
US. Patent
Feb. 5, 2013
Sheet 2 of4
US 8,369,540 B2
10\
14
‘
LEVEL
'
ESTIMATOR
16
f
/12
RX
f 18
LOGIC CIRCUIT
V
V
AUDIO
:
DSP
20
24
AUDIO
:
DAG
f
AMPLIFIER
22
Fig. 3
26
US. Patent
Feb. 5, 2013
Sheet 3 of4
US 8,369,540 B2
A
GAIN
(dB)
AMAX
.-'_ _"_"_
1'-
’
INPUT GAIN (dB)
Fig. 4
LEGEND
I
------ --—
—
---------
SIGNALOUTPUT
SCALING
GAIN
AMPLIFIER
TOTAL
GAIN GAIN
HS
US. Patent
Feb. 5, 2013
Sheet 4 of4
US 8,369,540 B2
RECEIVE DIGITAL
AUDIO SIGNAL
I
U) N
CALCULATE AUDIO
LEVEL OF SIGNAL
I
RECEIVE GAIN LEVEL
TO BE APPLIED
I
CALCULATE
HEADSPACE IN SIGNAL
I
DIVIDE GAIN BETWEEN
DSP AND AMPLIFIER
Fig. 5
U) 00
\\\\ \ S2
CD 01
I
AMPLIFY DIGITAL
AUDIO SIGNAL
I
U) N
CONVERT TO
ANALOGUE SIGNAL
I
AMPLIFY ANALOGUE
AUDIO SIGNAL
U) 00
US 8,369,540 B2
1
2
AUDIO SIGNAL AMPLIFICATION
level into a scaling gain and an ampli?er gain, said scaling
gain not exceeding the calculated headspace, a digital signal
processor arranged to amplify the digital audio signal With the
scaling gain, a digital-to-analogue converter arranged to con
vert the ampli?ed digital audio signal into an analogue signal,
and an ampli?er arranged to amplify the analogue audio
This invention relates to method of and a system for ampli
fying a digital audio signal. The invention provides signal
scaling assisted volume ampli?cation.
Volume control in multimedia systems is typically
achieved through a gain setting provided to the ampli?er, as
signal With the ampli?er gain.
shoWn schematically in FIG. 1. The gain in dB, provided by
OWing to the invention, it is possible to provide an ampli
high gain settings. Non-switching audio ampli?ers, for
fying system that solves the problem of achieving high gain
settings for audio signals having loW input level, Without
distortion, by using signal scaling assistance from a digital
example those of class A, AB, B, BC etc. utilise the linear
range of transistors, as a result of Which they produce distor
gain that the user Wishes to obtain is achieved through a
the ampli?er, can be both positive as Well as negative. Fre
quently though, ampli?ers produce distortions for very loW or
signal processor. For very high volume settings, the aggregate
tions at very high and loW settings. The distortion is especially
audible at high volume settings Where the ampli?er curve
conjunction of simultaneous gain setting for the ampli?er and
saturates, and the output audio levels do not increase substan
distortion. The respective gain settings are adaptively derived
tially With increased ampli?er settings. In recent years,
Class-D ampli?ers have been developed that provide distor
tions comparable to the non-sWitching audio ampli?ers, but
based on short-term signal level characteristics computed on
the signal processor. A digital level limiter on the DSP ensures
that signal clipping does not occur. One advantage of this
the maximum output levels are still relatively loW.
An alternate approach to volume control is to provide an
signal scaling on the DSP so as to minimiZe the overall
20
ampli?er With constant gain at the audio output, While scaling
the digital audio signal doWn to an appropriate level in a
digital signal processor, as shoWn schematically in FIG. 2.
Note that the gain setting in this case, provided by the digital
invention is that the effective gain settings achieved in the
process (for audio signals having loW levels) signi?cantly
25
extend beyond that of the ampli?er alone Without compro
mising on the quality of audio output. For class-B/C ampli
?ers, the technique can also be extended to provide good
signal quality at very loW volume levels.
signal processor (DSP) is alWays less than or equal to 0 dB
(unity gain). Under this set-up, the ampli?er uses up the same
amount of output poWer no matter What the volume setting by
Advantageously, the step of calculating the headspace in
the digital audio signal comprises calculating the headspace
the user, as the ampli?er is providing a constant gain, and the
DSP is providing a variable gain.
margin and La is the audio level of the digital audio signal.
This simple formula for the calculation of the headspace
according to the formula HS:LC—La, Where LC is an error
30
Splitting of the signal ampli?cation betWeen tWo ampli?
ers is known from, for example, United States of America
Patent Application Publication US 2001/0026624, Which dis
closes a system in Which volume adjustment is performed in
small steps at a DSP and volume adjustment is performed in
Wide steps at electronic volume circuits. Adjustment only by
ensures that no audio artefacts occur in the outputted signal.
Preferably, the amplifying method further comprises
35
the dividing of the gain level into a scaling gain and an
the DSP is performed for a small volume range less than or
equal to a predetermined level. For a volume higher than or
ampli?er gain, Where the scaling gain does not exceed the
calculated headspace, comprises setting the ampli?er gain
equal to the predetermined level, ?ne adjustment by the DSP
at the transient period of volume adjustment is combined to
reduce the increment of variation so that the volume adjust
ment is performed gradually. HoWever this system does not
40
equal to the ampli?er gain limit and setting the scaling gain
equal to the gain level minus the ampli?er gain limit.
Volume gain in multimedia systems is typically achieved
through a gain setting provided to the ampli?er, or through
digital audio signal scaling. Both techniques assume the
45
audio signal to be full-scale, and transfer the onus of audio
solve the problem of reducing the distortion of the input
signal at high and loW levels of ampli?cation.
It is therefore an object of the invention to improve upon the
level enhancement to the ampli?er. Most ampli?ers produce
distortions for high gain settings. And most real-life signals
knoWn art.
According to a ?rst aspect of the present invention, there is
provided a method of amplifying a digital audio signal com
prising receiving a digital audio signal, calculating the audio
level of the digital audio signal, receiving a gain level, said
gain level de?ning the desired ampli?cation of the digital
audio signal, calculating the headspace in the digital audio
signal, dividing the gain level into a scaling gain and an
ampli?er gain, said scaling gain not exceeding the calculated
headspace, amplifying the digital audio signal at a digital
signal processor With the scaling gain, converting the ampli
?ed digital audio signal into an analogue signal, and ampli
fying the analogue audio signal at an ampli?er With the ampli
?er gain.
According to a second aspect of the present invention, there
is provided a system for amplifying a digital audio signal
comprising a receiver for receiving a digital audio signal, a
50
cover only a fraction of the available dynamic range.
All of the knoWn techniques for audio gain control assume
that the audio signal is full scale i.e. covers the entire digital
dynamic range. In reality, the input signal level is often much
55
loWer, and it is to cater to the needs of listening to such audio
signals that audio systems are stretched to their maximum
levels. Even normal audio signals are typically expected to
cover only about 70% of the audio dynamic range. The
present invention seeks to use the available dynamic range
headspace for the received audio signal to provide a higher
gain setting than possible through use of the ampli?er alone.
This is especially valuable When the input signal level is loW.
60
The present invention operates by dividing the aggregate
gain set by the user into tWo parts, ?rstly a gain setting on the
ampli?er, and secondly a gain setting on signal scalar in the
level estimator arranged to calculate the audio level of the
digital audio signal, a gain control arranged to receive a gain
level, said gain level de?ning the desired ampli?cation of the
digital audio signal, a logic circuit arranged to calculate the
headspace in the digital audio signal and to divide the gain
accessing an ampli?er gain limit for the ampli?er. During the
amplifying of the analogue audio signal at an ampli?er With
the ampli?er gain, the maximum ampli?cation applied by the
ampli?er does not exceed the ampli?er gain limit. Similarly,
digital signal processor that can assume a value greater than 1 .
65
In this Way, the range of user-settable gain effectively
increases beyond What the ampli?er can provide in isolation,
at a loW cost. The alternative for this to enable handling of
input signals With loW levels Would be to have ampli?ers With
US 8,369,540 B2
3
4
larger gain ranges, which would be wasteful in terms of
system cost and power consumption.
audio ampli?er 24 is never driven into gain levels that would
cause the audio ampli?er 24 to distort the received digital
audio signal.
Embodiments of the present invention will now be
The audio level estimator 14 takes the received digital
audio signal as its input and generates an estimate of the audio
described, by way of example only, with reference to the
accompanying drawings, in which:
level in dB, assuming 0 dB setting for the full-scale signal.
FIG. 1 is a schematic diagram of volume gain setting into
an ampli?er,
FIG. 2 is a schematic diagram of the use of signal scaling
The level estimator 14 is short-term, with a smoothing time
for gain setting,
various possible implementations of the audio level estimator
constant of 5-30 ms, and may be block-based. There are
FIG. 3 is a schematic diagram of a system for amplifying a
digital audio signal,
14, such as a recti?er followed by a low-pass ?lter and con
version of the obtained output to dB.
FIG. 4 is a graph showing an example of how an input gain
to the system of FIG. 3 can be split between a scaling gain and
The gain setting logic 18 has two different inputs, the user
gain setting in dB (GIU), from the gain control 16, and the
an ampli?er gain, and
audio level in dB (LA), from the level estimator 14. The
second setting allows the logic circuit 18 to calculate the
FIG. 5 is a ?ow diagram of a method of amplifying a digital
headspace available (HS) for audio signal scaling by using:
audio signal.
FIG. 3 shows an example of a system 10 according to the
invention, for amplifying a digital audio signal. The system
10 comprises a receiver 12 for receiving the digital audio
signal, a level estimator 14 arranged to calculate the audio
level in the digital audio signal, a gain control 16 arranged to
receive a gain level set by a user, the gain level de?ning the
desired ampli?cation of the digital audio signal, and a logic
circuit 18, which is arranged to calculate the headspace in the
audio signal (from the calculated audio level), and to divide
the gain level into a scaling gain and an ampli?er gain, the
HSIL C—LA
20
limit for the ampli?er setting. The output scaling (G05) and
ampli?er (GOA) gains are then obtained as follows:
25
scaling gain not exceeding the calculated headspace.
Additionally, the system 10 comprises a digital signal pro
cessor 20 arranged to amplify the digital audio signal with the
scaling gain, a digital-to-analogue converter 22 arranged to
convert the ampli?ed digital audio signal into an analogue
signal, an ampli?er 24 arranged to amplify the analogue audio
signal with the ampli?er gain, and a loudspeaker 26, which
outputs the ampli?ed audio signal.
The system 10 provides signal scaling assisted volume
ampli?cation. Digital audio systems typically comprise a
GOA =
GIU
if GIU i AMAX
AMAX otherwise
30
The preferred signal scaling required under this setting is:
GSP : GIU- GOA
35
where GSP is the amount of gain that would be thrust on the
signal scalar (the DSP 20) if it were not constrained by the
availability of scaling headspace HS. Typically, this is the
DSP for scaling the gain of the digital audio signal, and an
amount of signal scaling that we would like to provide assum
audio ampli?er that is used to drive the speakers once the
ing that the input audio signal provides equivalent headspace.
digital signal has been converted into analogue form. Digital
audio volume control in multimedia systems is typically
where LC is a small margin, for example, —0.3 dB, left for
operational convenience. The constant AMAXis the maximum
ampli?er level in dB that can be achieved by the ampli?er 24,
with negligible audio distortion. This then forms the logical
In the event that this assumption is not valid, level-saturation
40
achieved through controlling the gain of the audio ampli?er.
One of the drawbacks with this method of gain control is that
logic is used to prevent audio distortion through signal clip
ping. The output scaling gain is thus obtained as follows:
the ampli?ed signal often becomes distorted when the ampli
?er is set to very low or very high gain. The digital audio
signals often only utilise 70% of the possible dynamic range
of the digital signal format, and sometime much less.
The system 10 is operated so that the gain of the digital
signal (by the DSP 20) is increased above a factor of 1, so that
the ampli?er 24 does not need to be driven to such a high gain
in order to give an equivalent output volume level. This
reduces the distortion added by the ampli?er 24, and means
that a lower-gain ampli?er 24 can be used, reducing the cost
45
The above design provides for signal scaling only after the
50
the gain GIU at any given point of time is partitioned into
non-trivial gain values GOA and G05 respectively. It is also
of the ampli?er 24, the power consumption, and the heat-sink
requirements etc.
The gains of the DSP 20 and of the ampli?er 24 are con
trolled together, so that the gain of the DSP 20 is controlled to
ampli?er 24 has reached the limit of its low-distortion range.
It is possible to have variations over the above scheme where
55
feasible to extend this gain-sharing scheme for very low gain
settings, at which the ampli?er 24 may again have a limit
beyond which distortion sets in. The logic circuit 18 is simply
splitting the total gain required by the gain control 1 6 between
keep the gain of the ampli?er 24 within its non-distorting
the two gain circuits 20 and 24, according to one or more
range. This can be achieved either by increasing the gain of
the DSP 20 to achieve higher output volume levels after the
audio ampli?er 24 has been set to its maximum non-distorting
gain, or by reducing the gain of the DSP 20 to achieve lower
output volume levels after the audio ampli?er 24 has been set
algorithms.
60
audio-video) playback system to deliver requisite output
to its minimum non-distorting gain.
audio level without a more expensive and power-consuming
ampli?er. The system 10 can be deployed across the entire
In use, the user will set the overall required gain required
from the system 10, through the gain control 16, and then the
system 10, through the logic circuit 18 will split the gain
between the DSP 20 and the audio ampli?er 24 such that the
The system 10 effectively increases the range of user
settable gain of an audio ampli?er beyond what it can provide
in isolation, at a low cost. By this, it allows an audio (or
65
range of digital audio (or audio-video) systems that use digi
tal signal processing in addition to having an audio ampli?er
at the output stage.
US 8,369,540 B2
6
5
The applicant hereby gives notice that neW claims may be
FIG. 3 illustrates graphically one embodiment of the split
ting of the gain betWeen the DSP 20 and the ampli?er 24. On
the x-axis is the input gain level set by the user via the gain
formulated to such features and/or combinations of such fea
control 16, and on the y-axis is the output gain delivered by
both amplifying devices, the DSP 20 and analogue ampli?er
tures during the prosecution of the present application or of
any further application derived therefrom.
For the sake of completeness it is also stated that the term
24. The logic circuit 18 divides the total gain betWeen the DSP
20 and analogue ampli?er 24. In the algorithm embodied in
the graph of FIG. 4, the ampli?er 24 is operated to provide the
term “a” or “an” does not exclude a plurality, a single proces
sor or other unit may ful?l the functions of several means
“comprising” does not exclude other elements or steps, the
gain of the system 10, until the level AMAXis reached.
The logic circuit 18 is arranged so that signal scaling in the
audio DSP 20 is only used once the ampli?ers gain limit
(AMAX) for high ?delity output has been reached. Thereafter,
recited in the claims and reference signs in the claims shall not
be construed as limiting the scope of the claims.
The invention claimed is:
any residual gain is preferentially provided by the signal
1. A method of amplifying a digital audio signal compris
scalar (the DSP 20). The gain delivered by the DSP 20 is
limited by the available headspace (HS) of the input audio
signal; if this headspace is exceeded, the signal Will clip or
Wrap-around, leading to serious distortions. Therefore, the
scaling gain that is imposed on the audio signal by the DSP 20
ing
receiving the digital audio signal,
calculating an audio level of the digital audio signal,
receiving a gain level, said gain level de?ning a desired
is the minimum of the residual gain and the available scaling
headspace (HS) of the input audio signal.
20
ampli?cation of the digital audio signal,
calculating a headspace HS in the digital audio signal
The system 10 takes advantage of the fact that for the vast
according to a formula HS:LC—La, Where LC is an error
majority of real-life audio signals, the signal only spans a part
of the available dynamic range. This leaves headspace for
signal scaling that is then used to assist the ampli?er 24 in
increasing the outputted audio levels. The method imple
mented by the system 10 therefore provides for a dynamic
temporal measurement of the audio headspace (HS), and the
margin and La is the audio level of the digital audio
signal,
dividing the gain level into a scaling gain and an ampli?er
25
gain, said scaling gain not exceeding the calculated
headspace,
amplifying the digital audio signal at a digital signal pro
logic circuit 18 uses this to assist volume scaling to levels that,
cessor With the scaling gain to obtain an ampli?ed digital
in a conventional system, Would have caused severe audio
distortion if implemented through a loW-cost ampli?er alone,
or Would have required an ampli?er With substantially higher
30
amplifying the analogue audio signal at an ampli?er With
The method of amplifying the digital audio signal is sum
marised in FIG. 5. The method comprises the steps 51 receiv
the ampli?er gain.
35
modi?cations Will be apparent to the skilled person. Such
variations and modi?cations may involve equivalent and
other features Which are already knoWn in the art, and Which
may be used instead of, or in addition to, features already
described herein.
Although the appended claims are directed to particular
combinations of features, it should be understood that the
scope of the disclosure of the present invention also includes
45
space, comprises setting the ampli?er gain at least approxi
mately equal to the ampli?er gain limit and setting the scaling
gain at least approximately equal to the gain level minus the
ampli?er gain limit.
5. A system for amplifying a digital audio signal compris
ing
50
a receiver for receiving the digital audio signal,
a level estimator arranged to calculate an audio level of the
digital audio signal,
a gain control arranged to receive a gain level, said gain
level de?ning a desired ampli?cation of the digital audio
55
signal,
a logic circuit arranged to calculate a headspace HS in the
digital audio signal, according to a formula HSILJLG,
Where LC is an error margin and La is the audio level of
the digital audio signal, and to divide the gain level into
closed herein either explicitly or implicitly or any generali
60
a scaling gain and an ampli?er gain, said scaling gain not
exceeding the calculated headspace,
a digital signal processor arranged to amplify the digital
present invention.
Features Which are described in the context of separate
embodiments may also be provided in combination in a single
embodiment. Conversely, various features Which are, for
brevity, described in the context of a single embodiment, may
also be provided separately or in any suitable sub combina
tion.
ampli?er does not exceed the ampli?er gain limit.
4. A method according to claim 2, Wherein the step of
dividing the gain level into a scaling gain and an ampli?er
gain, said scaling gain not exceeding the calculated head
any novel feature or any novel combination of features dis
sation thereof, Whether or not it relates to the same invention
as presently claimed in any claim and Whether or not it miti
gates any or all of the same technical problems as does the
2. A method according to claim 1, further comprising
accessing an ampli?er gain limit for the ampli?er.
3. A method according to claim 2, Wherein, during the step
of amplifying the analogue audio signal at an ampli?er With
the ampli?er gain, the maximum ampli?cation applied by the
40
the scaling gain, step S7 converting the ampli?ed digital
audio signal into an analogue signal at the DAC 22, and ?nally
step S8 amplifying the analogue audio signal at the ampli?er
24 With the ampli?er gain.
From reading the present disclosure, other variations and
converting the ampli?ed digital audio signal into an ana
logue signal, and
capabilities.
ing the digital audio signal, step S2 calculating the audio level
of the digital audio signal, step S3 receiving a gain level from
the gain control 16, the gain level de?ning the desired ampli
?cation of the digital audio signal, step S4 calculating the
headspace in the digital audio signal, step S5 dividing the gain
level into a scaling gain and an ampli?er gain (the scaling gain
not exceeding the calculated headspace), step S6 amplifying
the digital audio signal at the digital signal processor 20 With
audio signal,
audio signal With the scaling gain to obtain an ampli?ed
digital audio signal,
65
a digital-to-analogue converter arranged to convert the
ampli?ed digital audio signal into an analogue signal,
and
US 8,369,540 B2
7
8
an ampli?er arranged to amplify the analogue audio signal
With the ampli?er gain.
6. A system according to claim 5, Wherein the logic circuit
8. A system according to claim 6, Wherein the logic circuit
is further arranged, When dividing the gain level into a scaling
is further arranged to access an ampli?er gain limit for the
calculated headspace, to set the ampli?er gain at least
approximately equal to the ampli?er gain limit and to set the
ampli?er.
7. A system according to claim 6, Wherein the logic circuit
is further arranged to ensure that, during the step of amplify
ing the analogue audio signal at an ampli?er With the ampli
?er gain, the maximum ampli?cation applied by the ampli?er
does not exceed the ampli?er gain limit.
gain and an ampli?er gain, said scaling gain not exceeding the
scaling gain at least approximately equal to the gain level
minus the ampli?er gain limit.
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