[Q]Are the ADPCM and the PCM lossy or lossless???
Baik, Seong-bok
say the ADPCM is lossy. Could anyone please tell me why?
Thanks in advance...
--
+-------------------
Baik, Seong-Bok.
---------------------+
Joseph S. Wisniewski
>
> Usually, the PCM seems to be thought as lossless, but most people
> say the ADPCM is lossy. Could anyone please tell me why?
>
> Thanks in advance...
PCM (Pulse Code Modulation) is loseless becuase it's not a data
compression method. It's just simply the storage or transmission of the
raw values (the Code for the Pulse) that come from an A/D converter.
ADPCM (Adaptive Differential PCM) takes the differential of a PCM signal
(the difference between successive samples). At this point, you haven't
actually lost any information yet, since the difference between two 16
bit samples is (technically) a 17 bit number. The "lossy" part comed
from the "adaptive" slgorithm. An adaptive scaling factor is applied to
these differences, to get them down to 4 bit numbers (well, some ADPCM
schemes use from 2 to 6 bit numbers, but 4 is most common). When the
scaling factor is small, you capture low level changes in the signal
well (and most any low frequency data) but high level, high speed
transients (large values of high frequency components) will distort
because you don't have enough dynamic range to track them, so you run
into a slew rate limit (clipping of the differential). When the adaptive
scaling factor is large, you capture large dynamice well, but small
variations are distorted, since the quantitization step size is too
large, and you get quantitization noise. So the signal you get back
almost always looses some information.
Hope this helps.
Joe
Hubert Crepy
> PCM (Pulse Code Modulation) is loseless becuase it's not a data
> compression method. It's just simply the storage or transmission of the
> raw values (the Code for the Pulse) that come from an A/D converter.
<nitpick warning>
While this is true once you're in the digital domain, it could be argued
that PCM as a representation of the analog signal is lossy, due to
sampling rate (Nyquist criterion) and quantization noise.
<end nitpick warning>
--
Hubert Crepy - IBM France Speech R&D - remove spam-blocker from address
to answer me
Paul Murrin
> ADPCM (Adaptive Differential PCM) takes the differential of a PCM signal
> (the difference between successive samples). At this point, you haven't
> actually lost any information yet, since the difference between two 16
> bit samples is (technically) a 17 bit number. <snip>
I was under the impression that ADPCM encoded the difference between the
current sample and a prediction of that sample rather than successive
samples. Are there different types of ADPCM?
Paul.
Erik de Castro Lopo
Yes about two dozen different kinds. Its a real can of worms.
Erik
--
+-------------------------------------------------+
Erik de Castro Lopo er...@zip.com.au
+-------------------------------------------------+
"... the industrial-capitalist mode of software production
was doomed to be outcompeted from the moment capitalism
began to create enough of a wealth surplus for many
programmers to live in a post-scarcity gift culture."
-- Eric S. Raymond
amitv...@gmail.com
> "Baik, Seong-bok" wrote:
> >
> > Usually, the PCM seems to be thought as lossless, but most people
> > say the ADPCM is lossy. Could anyone please tell me why?
> >
> > Thanks in advance...
>
> compression method. It's just simply the storage or transmission of the
> raw values (the Code for the Pulse) that come from an A/D converter.
>
> (the difference between successive samples). At this point, you haven't
> actually lost any information yet, since the difference between two 16
> from the "adaptive" slgorithm. An adaptive scaling factor is applied to
> these differences, to get them down to 4 bit numbers (well, some ADPCM
> schemes use from 2 to 6 bit numbers, but 4 is most common). When the
> scaling factor is small, you capture low level changes in the signal
> well (and most any low frequency data) but high level, high speed
> transients (large values of high frequency components) will distort
> because you don't have enough dynamic range to track them, so you run
> into a slew rate limit (clipping of the differential). When the adaptive
> scaling factor is large, you capture large dynamice well, but small
> variations are distorted, since the quantitization step size is too
> large, and you get quantitization noise. So the signal you get back
> almost always looses some information.
>
> Hope this helps.
>
> Joe