FPGA Central

> Assume 400 lines horizontal resolution, that means a data
> rate of 30*490*400 about 6 Mhz. Assume 12 bits per pixel,
> that's a data rate of 9 Mhz.

Sorry, that's 9 MBytes/sec data rate.

[email protected] wrote:

> WAV is simply a specific container format for audio information.
> WAV files can hold linear PCM, non-linear PCM such as A-law or
> u-law, encodings such as MPEG, AC-3 and such.
>
> Thus, if you simply specify that something is in WAV format, you're
> only saying that it's a RIFF-format file and that it's internal
> format is unspecified. It could be linear PCM (wave format type 1),
> it could be MPEG (wave format type 80).
>
> Both linear PCM and MPEG ARE pulse-code modulations, meaning that
> at certain instants, the instantaneous amplitude of the is sampled
> and encoded. One of them simply has gone through some additional
> lossy encoding procedure.


I am confused now.

http://www.adobe.com/support/techdocs/329133.html

Here they show "Windows Pulse Code Modulation (PCM) data (.wav)" as
being diffrent from "Linear PCM"

On 1 Oct 2005 19:45:56 -0700, "Radium" <[email protected]> wrote:

>[email protected] wrote:
>
>> WAV is simply a specific container format for audio information.
>> WAV files can hold linear PCM, non-linear PCM such as A-law or
>> u-law, encodings such as MPEG, AC-3 and such.
>>
>> Thus, if you simply specify that something is in WAV format, you're
>> only saying that it's a RIFF-format file and that it's internal
>> format is unspecified. It could be linear PCM (wave format type 1),
>> it could be MPEG (wave format type 80).
>>
>> Both linear PCM and MPEG ARE pulse-code modulations, meaning that
>> at certain instants, the instantaneous amplitude of the is sampled
>> and encoded. One of them simply has gone through some additional
>> lossy encoding procedure.
>
>
>I am confused now.
>
>http://www.adobe.com/support/techdocs/329133.html
>
>Here they show "Windows Pulse Code Modulation (PCM) data (.wav)" as
>being diffrent from "Linear PCM"

I suspect they mean a "raw" data format for "Linear PCM" where the
file just starts out with words of PCM data in some
externally-agreed-upon format.
Windows .wav files have something like 56 bytes of header info (the
RIFF header thing) at the begining that tells what kind of data it is
(linear PCM, ADPCM, etc), sample rate (you wouldn't want it to play
back at the wrong speed, would you?), how many channels, how many
bytes per sample, number of samples in the file, and other things I
forget.
The .wav format most often contains linear PCM and so is often
equated to linear PCM, but it could contain something else.

Ben Bradley wrote:
> I suspect they mean a "raw" data format for "Linear PCM" where the
> file just starts out with words of PCM data in some
> externally-agreed-upon format.

No. If you read carefully, you'll find that the two raw formats [no
headers] are .pcm and .raw. There is also the "64-bit raw doubles"
which can be in .dbl or .raw format.

> The .wav format most often contains linear PCM and so is often
> equated to linear PCM, but it could contain something else.

What else does it contain?

"Radium" <[email protected]> wrote in message
news:[email protected] oups.com...
>
> Ben Bradley wrote:
>> I suspect they mean a "raw" data format for "Linear PCM" where the
>> file just starts out with words of PCM data in some
>> externally-agreed-upon format.
>
> No. If you read carefully, you'll find that the two raw formats [no
> headers] are .pcm and .raw. There is also the "64-bit raw doubles"
> which can be in .dbl or .raw format.
>
>> The .wav format most often contains linear PCM and so is often
>> equated to linear PCM, but it could contain something else.
>
> What else does it contain?

A wave file could have ADPCM (adaptive differential PCM) or 8-bit u-law or a-low
companded audio. Maybe others too. See:
http://support.microsoft.com/kb/q89879/
http://www.epanorama.net/documents/t...ulaw_alaw.html

"Jon Harris" <[email protected]> wrote in message
news:4yJ%e.8271$UD6.7374@trnddc04...
> "Radium" <[email protected]> wrote in message
> news:[email protected] oups.com...
>>
>> Ben Bradley wrote:
>>> I suspect they mean a "raw" data format for "Linear PCM" where the
>>> file just starts out with words of PCM data in some
>>> externally-agreed-upon format.
>>
>> No. If you read carefully, you'll find that the two raw formats [no
>> headers] are .pcm and .raw. There is also the "64-bit raw doubles"
>> which can be in .dbl or .raw format.
>>
>>> The .wav format most often contains linear PCM and so is often
>>> equated to linear PCM, but it could contain something else.
>>
>> What else does it contain?
>
> A wave file could have ADPCM (adaptive differential PCM) or 8-bit u-law or
> a-low companded audio. Maybe others too. See:
> http://support.microsoft.com/kb/q89879/
> http://www.epanorama.net/documents/t...ulaw_alaw.html

P.S. I should have added that a wave file is just a "container" that can contain
a wide variety of audio formats, sample rates, bit widths, number of channels,
etc.. The most common is your basic linear PCM, but that's certainly not the
only possibility. People off use "wave file" to mean "uncompressed 44.1kHz
16-bit stereo audio file", but that is a pretty sloppy usage of the term.

Radium wrote:
> [email protected] wrote:
>
>
>>WAV is simply a specific container format for audio information.
>>WAV files can hold linear PCM, non-linear PCM such as A-law or
>>u-law, encodings such as MPEG, AC-3 and such.
>>
>>Thus, if you simply specify that something is in WAV format, you're
>>only saying that it's a RIFF-format file and that it's internal
>>format is unspecified. It could be linear PCM (wave format type 1),
>>it could be MPEG (wave format type 80).
>>
>>Both linear PCM and MPEG ARE pulse-code modulations, meaning that
>>at certain instants, the instantaneous amplitude of the is sampled
>>and encoded. One of them simply has gone through some additional
>>lossy encoding procedure.
>
>
>
> I am confused now.
>
> http://www.adobe.com/support/techdocs/329133.html
>
> Here they show "Windows Pulse Code Modulation (PCM) data (.wav)" as
> being diffrent from "Linear PCM"

As far as I can see, all standards on that page use PCM. The differences
are in (type of) compression, number of channels, bits per sample, and
sample rate, and header information. Transmitted ASCII is also PCM.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Jerry Avins wrote:

> Transmitted ASCII is also PCM.

So all digital signals [including text] are PCM?

Is MIDI also PCM?

Radium wrote:
> Jerry Avins wrote:
>
>
>>Transmitted ASCII is also PCM.
>
>
> So all digital signals [including text] are PCM?

No there is digital PWM, PPM, and probably others that don't come
immediately to mind. Strictly, NRZ codes (used on magnetic media) are
not PCM, but there's a simple translation.

> Is MIDI also PCM?

Yes.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

On Sat, 01 Oct 2005 12:26:37 -0400, Jerry Avins <[email protected]> wrote:

>Stewart Pinkerton wrote:
>> On Fri, 30 Sep 2005 17:24:43 -0400, Jerry Avins <[email protected]> wrote:
>>
>>
>>>Stewart Pinkerton wrote:
>
> ...
>
>>>>Ahem, PWM isn't necessarily digital - remember the wonderful Sinclair
>>>>X-10 amplifier from the mid '60s?
>>>
>>>No. please describe it.
>>
>>
>> Class D amp using germanium small-signal transistors, rated at 10
>> watts, actualy produced about 3 watts, rated at 0.1% THD at full power
>> - unfortunately, those were switching artifacts which stayed at the
>> same level regardless of output, so the whole thing sounded absolutely
>> dreadful! Lasted for about a year on the market, and was followed by a
>> whole slew of highly novel but ultimately trashy Sinclair products.
>>
>> Clive Sinclair (who didn't actually design the X-10) is a true
>> innovative genius - who couldn't nail together two pieces of wood,
>> never mind produce reliable electronic gear.
>
>I'm glad I never had first-hand experience with it! Still, one could
>make a case that a class-D amplifier is digital at least in some
>respects. The amplitude at the switch is certainly quantized, and pulse
>width also in many modern implementations. I will assert unconditionally
>that *some* PWM applications are fully digital, and that it is a useful
>digital encoding for some applications.

But it's *not* a digital encoding, it is a *modulation* which happens
to involve the use of a switch to raise efficiency. The same is true
of Pulse Density Modulation, aka BitStream. OTOH, Pulse *Code*
Modulation, and in particular Cross Interleaved Reed-Solomon Code, is
definitely digital.

>All that aside, the OP's assumption that compressed signals aren't (or
>can't) be encoded as PCM is nonsense.

Agreed.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

On 1 Oct 2005 09:39:28 -0700, "Radium" <[email protected]> wrote:

>Jerry Avins wrote:
>
>> I'm glad I never had first-hand experience with it! Still, one could
>> make a case that a class-D amplifier is digital at least in some
>> respects. The amplitude at the switch is certainly quantized, and pulse
>> width also in many modern implementations. I will assert unconditionally
>> that *some* PWM applications are fully digital, and that it is a useful
>> digital encoding for some applications.
>
>PWM is not digital. PWM is in pulses but that does not make it digital.
>
>> All that aside, the OP's assumption that compressed signals aren't (or
>> can't) be encoded as PCM is nonsense.
>
>Sorry. I was stating that PCM is better than MPEG. Compressed signals
>can be in PCM. I was comparing WAVE files to MP3 files. In my
>experience WAVs are better than MP3s.

True by definition, but IME audibly indistinguishable once you get to
256kbits/sec.

>Makes me wonder what PCM video would look like if compared to the
>crappy MPEG stored on DVDs.

Clean, but short..................

> DVD video is certianly better than the
>video you get in those VHS cassettes. However, if only DVD used the
>visual equivalent of .WAV audio, it would be much better in picture and
>color resolution. DVDs do use PCM audio, which is fine, but it is time
>they start implenting their video in PCM and stop their useless MPEG
>video.

Clearly, you have access to much larger storage media than we do, or
you only like 5-minute films! :-)

MPEG2 is very far from useless, and MPEG4 is even better.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

On 1 Oct 2005 11:51:53 -0700, "Radium" <[email protected]> wrote:

>Richard Crowley wrote:
>> MPEG *IS* "PCM".
>
>Thats like saying MP3 is the same as WAV. Which is wrong. No offense.

WAV is a container format, doesn't tell you anything about the
contents. Perhaps we should refer to LPCM on the one hand, and
compressed digital on the other.

--

Stewart Pinkerton | Music is Art - Audio is Engineering

Radium wrote:
> Jerry Avins wrote:
>
> > Transmitted ASCII is also PCM.
>
> So all digital signals [including text] are PCM?

No, we are defining signals in this context as a quantity
whose value is a function of time. In which case, text is not
a signal in this context.

Signals are NOT PCM. PCM is merely one way of representing
signals. There are many ways: variations in the strength
and orientation of a magnetic field coated on the surface of
a flexible backing, displacement of a groove to one side or
another on a vinyl disk, and so on.

PCM is a REPRESENTATION of a signal. It is one of many.

> Is MIDI also PCM?

No, in most cases it is a series of commands. The execution of
those commands could result in a signal, to be sure and as such
is a representation of a signal.

Radium wrote:
> > The .wav format most often contains linear PCM and so is often
> > equated to linear PCM, but it could contain something else.
>
> What else does it contain?

The question is not what it does contain, but what it could.

First of all, WAV files MUST contain two parts, or "chunks" to
be valid, and the contents of those chunks determine what is
contained.

fmt - format chunk
Contains information about how the data is encoded,
including sample size, number of channels, sample
rate and, relevant to the current discussion, the
format of the data. The layout of the minimum format
chunk is:

wFormatTag - format of audio data
nChannel - number of channels
nSamplesPerSec - sample rate
nAvgBytesPerSec - average data rate
nBlockAlign - block alignment in bytes of the data
chunk
wBitsPerSample - number of bits in each sample

The fmt chunk can be extended to include more information
especially regarding formats becide linear PCM.

Examples of some of the audio formats that can be represented
in WAV files are:

Linear PCM
ADPCM
u-Law
A-Law
MPEG
Dolby AC-2

There are a number of prorietary formats as well.

data
contains the actual audio, ecoded as described in the fmt chunk

Radium wrote:
> [email protected] wrote:
> > Both linear PCM and MPEG ARE pulse-code modulations, meaning that
> > at certain instants, the instantaneous amplitude of the is sampled
> > and encoded. One of them simply has gone through some additional
> > lossy encoding procedure.
> I am confused now.

Yes, everyone knows that.

> http://www.adobe.com/support/techdocs/329133.html
>
> Here they show "Windows Pulse Code Modulation (PCM) data (.wav)" as
> being diffrent from "Linear PCM"

Now, it does NOT say that, not anywhere in that document. Why
do you conclude so?

It says that a software package supports a variety of input and
output formats. Noweher does it say ANYTHING of what you are
suggesting.

[email protected] wrote:
> Radium wrote:
>
>>Jerry Avins wrote:
>>
>>
>>>Transmitted ASCII is also PCM.
>>
>>So all digital signals [including text] are PCM?
>
>
> No, we are defining signals in this context as a quantity
> whose value is a function of time. In which case, text is not
> a signal in this context.

When characters are being received by a modem, the signal that
represents them varies with time. I think you make an artificial
distinction here.

> Signals are NOT PCM. PCM is merely one way of representing
> signals. There are many ways: variations in the strength
> and orientation of a magnetic field coated on the surface of
> a flexible backing, displacement of a groove to one side or
> another on a vinyl disk, and so on.
>
> PCM is a REPRESENTATION of a signal. It is one of many.
>
>
>>Is MIDI also PCM?
>
>
> No, in most cases it is a series of commands. The execution of
> those commands could result in a signal, to be sure and as such
> is a representation of a signal.

Be clear about what you mean by representation. A typical MIDI command
consists of letters in ASCII and numbers either as ASCII or binary.
These all consist of strings of bits, sent one at a time i.e., serially.
Each bit is represented by a pulse (on) or a place where a pulse could
have been (off). Such a representation is an encoding, and imposing that
encoding on a carrier of sorts (current loop, RTTY via FSK, +/-12V as in
RS-232, and more) constitutes pulse-code modulation of that carrier.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

Stewart Pinkerton wrote:
> On Sat, 01 Oct 2005 12:26:37 -0400, Jerry Avins <[email protected]> wrote:

...

>> I will assert unconditionally
>>that *some* PWM applications are fully digital, and that it is a useful
>>digital encoding for some applications.
>
>
> But it's *not* a digital encoding, it is a *modulation* which happens
> to involve the use of a switch to raise efficiency. The same is true
> of Pulse Density Modulation, aka BitStream. OTOH, Pulse *Code*
> Modulation, and in particular Cross Interleaved Reed-Solomon Code, is
> definitely digital.

On one occasion, I built a system to send 8-bit numbers. Efficient use
of bandwidth was not an issue -- it was a wire link -- and only signal
power was involved. The transmitter sent a pulse whose width was
proportional to the number, while the receiver measured the width of the
pulse. The crystals at two ends were more than well enough matched to
distinguish 254 from 255. If that's not PWM, what is it?

I did it that way to because it was fast enough and used the fewest
resources of an already heavily loaded system. (Such a link could send
ASCII, too.)

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

What we've come down to is exactly what is the definition of
"PCM", "Pulse Code Moduation"?

It is *not* a modulation scheme (like AM or FM or PM or FSK,
or PSK, etc.) One of *those* modulation schemes is required to
actually convey "PCM" signals.

It seems to me that it is an obsolete term invented back pre WW2
to distinguish conventional analog telephony from the new-fangled
"digital" (which word was not in common use yet).

Is there a significant definition of "PCM" that doesn't just mean
what we call generically "digital" today? For example, if you
go and read the definition at Wikipedia, you will see that it is
just a synonym for generic "digital".

I further posit that the only reason it is still used today is that
it sounds more "exotic" and "***y" to advertising gerbs and
unsuspecting consumers. Whenever I see the acronyn "PCM",
I automatically activate my BS filter.

On Sun, 2 Oct 2005 13:15:27 -0700, Richard Crowley wrote:

> What we've come down to is exactly what is the definition of
> "PCM", "Pulse Code Moduation"?
>
> It is *not* a modulation scheme (like AM or FM or PM or FSK,
> or PSK, etc.) One of *those* modulation schemes is required to
> actually convey "PCM" signals.
>
> It seems to me that it is an obsolete term invented back pre WW2
> to distinguish conventional analog telephony from the new-fangled
> "digital" (which word was not in common use yet).
>
> Is there a significant definition of "PCM" that doesn't just mean
> what we call generically "digital" today? For example, if you
> go and read the definition at Wikipedia, you will see that it is
> just a synonym for generic "digital".
>
> I further posit that the only reason it is still used today is that
> it sounds more "exotic" and "***y" to advertising gerbs and
> unsuspecting consumers. Whenever I see the acronyn "PCM",
> I automatically activate my BS filter.

Essentially with PCM there is a 1 to 1 mapping of a digital word value to a
corresponding analogue voltage level.

With other trickier schemes like MP3 this is not so, and you must do some
processing to arrive at the PCM signal, which is what the DAC demands in
order to produce the analogue output.

d

Don Pearce wrote:

> With other trickier schemes like MP3 this is not so, and you must do some
> processing to arrive at the PCM signal

This is exactly what I've been trying to say. MP3 and WAVE are so
different, even though they are both digital. There is a fatal amount
of difference. The problem for me is that I no clue at all as to *what*
is different!

"Don Pearce" wrote ...
> Essentially with PCM there is a 1 to 1 mapping of a digital word value
> to a
> corresponding analogue voltage level.
>
> With other trickier schemes like MP3 this is not so, and you must do
> some
> processing to arrive at the PCM signal, which is what the DAC demands
> in
> order to produce the analogue output.

I've never seen that definition published anywhere.
Reference?

Don Pearce wrote:

...

> Essentially with PCM there is a 1 to 1 mapping of a digital word value to a
> corresponding analogue voltage level.
>
> With other trickier schemes like MP3 this is not so, and you must do some
> processing to arrive at the PCM signal, which is what the DAC demands in
> order to produce the analogue output.

PCM is a one-to-one mapping of bits in a word to states in a modulated
signal stream. The bits usually represent numbers (binary, signed
binary, offset binary, a Gray code, twos-complement binary, more) and
the numbers can represent anything.

Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯ ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯

On 2 Oct 2005 15:56:04 -0700, "Radium" <[email protected]> wrote:

>Don Pearce wrote:
>
>> With other trickier schemes like MP3 this is not so, and you must do some
>> processing to arrive at the PCM signal
>
>This is exactly what I've been trying to say. MP3 and WAVE are so
>different, even though they are both digital. There is a fatal amount
>of difference. The problem for me is that I no clue at all as to *what*
>is different!

Unfortunately, you don't seem to understand that a .wav file can *be*
an MP3. If the .wav file is linear PCM, then your lack of
understanding is due to sheer ignorance, as the MP3 standard is very
well defined.

--

Stewart Pinkerton | Music is Art - Audio is Engineering

On Sun, 2 Oct 2005 19:09:51 -0700, Richard Crowley wrote:

> "Don Pearce" wrote ...
>> Essentially with PCM there is a 1 to 1 mapping of a digital word value
>> to a
>> corresponding analogue voltage level.
>>
>> With other trickier schemes like MP3 this is not so, and you must do
>> some
>> processing to arrive at the PCM signal, which is what the DAC demands
>> in
>> order to produce the analogue output.
>
> I've never seen that definition published anywhere.
> Reference?

You're reading it.

d

<[email protected]> wrote in message
news:[email protected] ups.com
>> Assume 400 lines horizontal resolution, that means a data
>> rate of 30*490*400 about 6 Mhz. Assume 12 bits per pixel,
>> that's a data rate of 9 Mhz.
>
> Sorry, that's 9 MBytes/sec data rate.

I can still remember when storing data at this rate was
actually quite a feat. Today, a few $$100s...

Now, storting uncompressed video data is merely wasteful,
not difficult or impossible.