Currently-Available Highest-Quality Linear PCM Video?

[Bob Myers]

I am well aware of that. Most digital video uses -- MPEG-layer or some
other form of compression -- not linear PCM. I don't know why?


Why isn't linear-PCM used in video?

Why do you think it should be? What advantage would it bring?
What costs (either in actual $ costs or performance) would using
linear PCM require?

Are you sure that isn't that the bit-rate? There is a world of
difference between bit-rate and sample-rate.

How many bits do you think are contained within each
"pixel"? What do you think the word "pixel" means, and
how is it distinguished from "sample" in the context of
digital imaging?

For example, CD audio has a sample rate of 44,100 hz but a bit rate of
1,411,200 bps.

If you divide 1,411,200 by 44,100, what number do you
get? What is the significance of this number?

Linear-PCM doesn't have to be used but what harm is caused by using it?

Answer the above questions, and then you'll have your own
answer to this one. Why are you so hung up on "linear PCM"
as opposed to any of several hundred (at least) other possible
encoding schemes?

Bob M.

 

[Bob Myers]

1-bit CD-Audio Technology makes theoretically a better use of the real
existing Resolution on the CD, processing only what is really there.

You are confusing the manner in which the data
stored on the CD is processed in playback with the
format of the data itself, as it appears on the CD.

Bob M.

 

[Daniel Mandic]

What??? Then it isn't CD audio.

No it's a Hamburger.

there.

What????

Be quiet. Otherwise state your arguments.






Best Regards,

Daniel Mandic

 

[Daniel Mandic]

You are confusing the manner in which the data
stored on the CD is processed in playback with the
format of the data itself, as it appears on the CD.

Bob M.

But a CD is not the CD alone... It's also the Player , otherwise I
may not mention Bits at all.


Best Regards,

Daniel Mandic

 

[Radium]

Why do you think it should be?

The same reason audio CDs used linear-PCM video instead of compressed
MP3s. MP3 and other compressed formats are inferior in quality to the
uncompressed linear-PCM.

Just as linear-PCM audio is better than MP3s, linear-PCM video is
better than MPEG video, VC-1 -- or other compressed -- video.

IMHO, VC-1 is even worse than MPEG

What advantage would it bring?

It wouldn't have those nasty "jaggies" associated with VC-1 or other
compressed video

What costs (either in actual $ costs or performance) would using
linear PCM require?

I agree, linear-PCM video would be more expensive but it wouldn't have
those annoying artifacts that occur in compressed video formats.

The lasers required would have to be of shorter wavelength than those
used in the writing and reading of conventional DVDs. The wavelength
should be 400 nm since thats the "sweet spot" between the advantages of
short wavelengths [i.e. less physical space on disc required] and the
hazards of ionizing UV radiation. 400 nm is around the shortest
wavelength of non-ionizing UV light.

In addition, the disc would have to be somewhat bigger -- around the
size of 33-speed phonos -- to accomodate the large data size required
for linear-PCM video.

I really don't see this as a major hurdle. What makes it so difficult
for the digital video industries to design the above [i.e. 400 nm
recording/playback lasers, optical discs the size of 33-speed phonos,
and uncompressed linear-PCM video]?

How many bits do you think are contained within each
"pixel"? What do you think the word "pixel" means, and
how is it distinguished from "sample" in the context of
digital imaging?

A "pixel" is *definitely* digital. A "sample" can be analog or digital.

If you divide 1,411,200 by 44,100, what number do you
get?
32

What is the significance of this number?

The bit-rate divided by the sample-rate. What else?

Answer the above questions, and then you'll have your own
answer to this one.

Why are you so hung up on "linear PCM"
as opposed to any of several hundred (at least) other possible
encoding schemes?

Because -- as I said before -- I don't like the annoying artifacts that
occur in compressed video formats. Those nasty pixelations associated
with VC-1 -- and other compressed -- video just make me gag.

http://en.wikipedia.org/wiki/VC1

Bob M.

The only compressions I am okay with are WMA^ and the *real* WMV I
described in the following threads:

1.
http://groups.google.com/group/micr...aa56d?lnk=st&q=&rnum=4&hl=en#4416c0283edaa56d

2.
http://groups.google.com/group/comp...a2396?lnk=st&q=&rnum=3&hl=en#afc33183826a2396

The *real* WMV should have the sampling rates and progressive
resolution [pixel X pixel] of the best quality video signal currently
existing. In addition, the WMV's sample rates and pixelXpixel format
should be exactly the same as the linear-PCM signal it was prior to
compression. As for the color-depth [in "bit-resolution"], decrease it
all you want and I still won't mind. In fact I am interested in seeing
how a movie would look if the WMV's color-depth is reduced so much that
the file-size is just 1-bit [regardless of how long the movie is]. But
don't you dare decrease the pixel resolution or sample rate, do so and
you'll find my vomit all around the room.

^The WMA should be monoaural and its sample rate should at least 44.1
khz which should also be the same sample rate of the signal when it was
in linear-PCM prior to compression. I don't mind if the WMA's
bit-resolution is compressed so much that the file size falls to just
1-bit [no matter how long the audio is]. Just keep it monoaural and
don't change the sample rate.

 

[Radium]

The bit-rate divided by the sample-rate. What else?

32 is also the bit-resolution [16-bit] times the number of channels [2,
because CD-audio is stereo, and has 2 channels]

16 [i.e. the bit-resolution] X 2 [i.e. stereo channel] = 32

 

[Bob Myers]

The same reason audio CDs used linear-PCM video instead of compressed
MP3s. MP3 and other compressed formats are inferior in quality to the
uncompressed linear-PCM.

Why linear PCM specifically, though? Why not any of
the other uncompressed digital formats? Do you know
why linear PCM is used when it IS used?

What storage capacity would be required to store one
hour of uncompressed 1920 x 1080 video at 24 bits per
pixels and using a 60 Hz progressive-scan format? What
bandwidth would be required to broadcast such a signal
(or carry it over a cable system), and using what modulation
method?

It wouldn't have those nasty "jaggies" associated with VC-1 or other
compressed video


I agree, linear-PCM video would be more expensive but it wouldn't have
those annoying artifacts that occur in compressed video formats.

I didn't exactly ask whether or not it would be more expensive
- I was asking you to specifically identify, and hopefully
quantify, the added expense. Can you do that?

The lasers required would have to be of shorter wavelength than those
used in the writing and reading of conventional DVDs. The wavelength
should be 400 nm since thats the "sweet spot" between the advantages of
short wavelengths [i.e. less physical space on disc required] and the
hazards of ionizing UV radiation. 400 nm is around the shortest
wavelength of non-ionizing UV light.

In addition, the disc would have to be somewhat bigger -- around the
size of 33-speed phonos -- to accomodate the large data size required
for linear-PCM video.

How did you arrive at that size? What would the storage
capacity of such a system be for any of the standard television
formats?

A "pixel" is *definitely* digital. A "sample" can be analog or digital.

Again, you didn't answer the question, and no, a "pixel" is not
"definitely digital." "Pixel" is derived from "picture element" -
in other words, a *sample* of an image. In imaging (and esp.
in video systems) "pixel" and "sample" (at least when used in the
context of the image data rather than the display device) are
precisely identical. Even then, your answer ("a 'pixel' is definitely
digital") has nothing to do with the question "how many bits do
you think are contained within each pixel?". In fact, since you seem
(based on your answer) to be hung up on the idea that "pixels"
ARE "digital," it makes that question all the more relevant.
Please answer it, and then you will have cleared up your confusion
regarding "bit rate" and "pixel rate."

The bit-rate divided by the sample-rate. What else?

Which means what? Do you think this might have something
to do with the number of bits in each sample? Do you now
see how that relates to your confusion regarding pixel rate
and bit rate?

Because -- as I said before -- I don't like the annoying artifacts that
occur in compressed video formats. Those nasty pixelations associated
with VC-1 -- and other compressed -- video just make me gag.

And yet you seem to believe that "linear PCM" is the only
possible format which would not exhibit those artifacts, without
expressing the slightest hint that you understand why this should
be so (were it true in the first place, which it isn't). Once again,
you have latched onto a buzzword or phrase without the slightest
understanding of what it actually implies. Do you enjoy looking
this foolish?

The *real* WMV should have the sampling rates and progressive
resolution [pixel X pixel] of the best quality video signal currently
existing. In addition, the WMV's sample rates and pixelXpixel format
should be exactly the same as the linear-PCM signal it was prior to
compression. As for the color-depth [in "bit-resolution"], decrease it
all you want and I still won't mind.

Really? What do you think "color depth" or "bit resolution"
equates to in terms of the image quality? What will a reduction
in this parameter change in the image?

Bob M.

 

[Bob Myers]

The bit-rate divided by the sample-rate. What else?

32 is also the bit-resolution [16-bit] times the number of channels [2,
because CD-audio is stereo, and has 2 channels]

So now you SHOULD be able to clear up your own confusion
regarding the difference between pixel rate and bit rate, no?

Bob M.

 

[Daniel Mandic]

What storage capacity would be required to store one
hour of uncompressed 1920 x 1080 video at 24 bits per
pixels and using a 60 Hz progressive-scan format? What
bandwidth would be required to broadcast such a signal
(or carry it over a cable system), and using what modulation
method?

Who is 1920? What is progressive-scan... is it not scanrate!? And what
hell-Display can show 24 bit/pixel??

Which means what? Do you think this might have something
to do with the number of bits in each sample? Do you now
see how that relates to your confusion regarding pixel rate
and bit rate?

Audio CD is not a sample. MP3 is a sample.



Best Regards,

Daniel Mandic

 

[Michael A. Terrell]

Who is 1920? What is progressive-scan... is it not scanrate!? And what
hell-Display can show 24 bit/pixel??

SVGA, with three, 8 bit channels.



--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 

[Radium]

If you divide 1,411,200 by 44,100, what number do you
get?

32

What is the significance of this number?

The bit-rate divided by the sample-rate. What else?

32 is also the bit-resolution [16-bit] times the number of channels [2,
because CD-audio is stereo, and has 2 channels]

So now you SHOULD be able to clear up your own confusion
regarding the difference between pixel rate and bit rate, no?

Bob M.

I never used the term "pixel rate"

 

[Ron Capik]

< ...snip... >


Really? What do you think "color depth" or "bit resolution"
equates to in terms of the image quality? What will a reduction
in this parameter change in the image?

Bob M.

And at some point you may want to look in to the works of
Bela Julesz. Things along the lines of textons, the elements
of texture perception ...and such.


Later...
--

 

[Daniel Mandic]

SVGA, with three, 8 bit channels.

Hi Michael Terrell!


I thought you meant S-Video





Best Regards,

Daniel Mandic

 

[Radium]

Why not any of
the other uncompressed digital formats?

What uncompressed digital formats -- other than linear PCM -- exist?

Do you know
why linear PCM is used when it IS used?

It offers the best quality. Imagine what wavetable synth would sound
like if it used mp3s in place of Wavs.

Read the thread where some imbecile asks about an "mp3table":

http://groups.google.com/group/comp...k=st&q=mp3table&rnum=1&hl=en#e080ba93f082e11e

Please answer it, and then you will have cleared up your confusion
regarding "bit rate" and "pixel rate."

I never used the term "pixel rate". Pixel is the video equivalent of
channel. Stereo audio has two channels. Mono has only one.

I did use the term "sample rate". Sample rate is the rate at which an
analog signal is sampled via Pulse-Amplitude Modulation. For CDs, this
rate is 44.1 khz. The PAM carrier signal is at a frequency of 44.1 khz.

Which means what? Do you think this might have something
to do with the number of bits in each sample?

CD audio has two channels because it is stereo. Each channels has a bit
resolution of 16-bit. So in each channel there are 16 bits per sample.
Combine the two channels are there are 32 bits per sample.

Do you now
see how that relates to your confusion regarding pixel rate
and bit rate?

Once again, I never used the term "pixel rate". I did use the terms
"sample rate" and "bit rate".

"Pixel rate" in video = "channel rate" in audio

Both "channel rate" and "pixel rate" and totally meaningless.

And yet you seem to believe that "linear PCM" is the only
possible format which would not exhibit those artifacts,

Nope. Linear-PCM video can be just as bad as MPEG -- and possibly even
worse -- if the image-resolution [in pixel X pixel], the color-depth
[bit-resolution], and/or sample rate [in hz] are too low.

What do you think "color depth" or "bit resolution"
equates to in terms of the image quality?

Color-depth or "bit-resolution" determines the number of colors that
can be displayed. 8-bit video can display a maximum of 2^16 colors [or
256 colors]. IOW, a color-resolution of n-bit means that a maximum of
2^n colors. In audio, the bit-resolution determines how many levels of
loudness can be handled [too loud and you get clipping, too soft and
you get quantization error].

What will a reduction
in this parameter change in the image?

A decrease the amount of available colors. However, compression of this
parameter is done via my "real WMV", the artifacts won't be as ugly as
it would be in linear-PCM or any compression other than my "real WMV".
Much in the same way the artifacts in a 120 kbps WMA file go
un-noticed. In a Wave file, 120 kbps would sound horrible -- this is
due to the absence of the perceptual encoding used in WMA. MP3s, and
other [non-WMA] compressed audio sound terrible to me. I also hate
aliasing so the sample rate for audio must be at least 44,100 hz.

 

[Radium]

Color-depth or "bit-resolution" determines the number of colors that
can be displayed. 8-bit video can display a maximum of 2^16 colors [or
256 colors].

Oops that should be 2^8 colors.

F--king
typos!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

 

[Ron Capik]

Color-depth or "bit-resolution" determines the number of colors that
can be displayed. 8-bit video can display a maximum of 2^16 colors [or
256 colors].

Oops that should be 2^8 colors.

F--king
typos!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

I wouldn't call that a typo, F--king or other wise.

--

 

[Bob Myers]

I never used the term "pixel rate"

No, you used the term "sample rate." But since in
an imaging context (unless subsampling of the chroma
components is employed) a pixel IS a sample, the
two are equivalent.

Bob M.

 

[Bob Myers]

Who is 1920? What is progressive-scan... is it not scanrate!? And what
hell-Display can show 24 bit/pixel??

1920 refers to the number of pixels per line in one
of the standard HDTV formats (the other one having
720 lines of 1280 pixels each). "Progressive scan"
refers to a scanning format in which all of the lines in
a frame are scanned in order, rather than being divided
into separate fields (such as typical 2:1 interlacing, which
divides the frame into "odd" and "even" fields, the lines
of which must be interleaved to recover the original).
Finally, many displays can shows 24 bits/pixel; in color
displays, that's the term typically used to refer to an
RGB system with 8 bits (256 levels) per color.

Audio CD is not a sample. MP3 is a sample.

I have no idea what this is supposed to mean, but
both audio CDs and MP3-format audio files are based
on samples of the original audio signal(s).

Bob M.

 

[Bob Myers]

What uncompressed digital formats -- other than linear PCM -- exist?

Transmission or storage? Straight 24 bit/pixel RGB is
the most obvious example of an uncompressed digital
format, and it certain;y doesn't have to be encoded in
a PCM manner.

It offers the best quality.

So your answer is "no, I really don't even know what
PCM is," right?

I never used the term "pixel rate". Pixel is the video equivalent of
channel. Stereo audio has two channels. Mono has only one.

No, the term "pixel" is used in a digital video context
interchangeably with "spatial sample." A "channel" is
something altogether different. Again, you clearly are
completely unfamiliar with the field in which you are trolling.

I did use the term "sample rate". Sample rate is the rate at which an
analog signal is sampled via Pulse-Amplitude Modulation.

A "sample rate" is a completely independent question from
the modulation system used to encode or transmit information.
Sampled systems are no more required to use PAM than your
other new favorite buzzword, PCM.

"Pixel rate" in video = "channel rate" in audio

Incorrect. The pixel rate in video is, if anything, most closely
related to the more general term "symbol rate" (or "baud
rate"), at least to the extent that until you define the number of
bits per pixel, the data rate (in bits/second or whatever equivalent
unit you like) cannot be determined. Similarly, "sample rate"
by itself does not determine the data rate.

Both "channel rate" and "pixel rate" and totally meaningless.

That would come as a rather sizable surprise to a large number
of professionals (including myself) who deal with digital video
interface, etc., issues.

The rest of your post indicates that you should have been
able to answer the very questions you were trolling with. I
therefore fail to see any point to continuing this discussion.

Bob M.

 

[Radium]

Transmission or storage?
Both

Straight 24 bit/pixel RGB is
the most obvious example of an uncompressed digital
format, and it certain;y doesn't have to be encoded in
a PCM manner.

What is this RGB encoded? Or is RGB a format by itself. AFAIK, RGB
simply stands for Red, Green, and Blue. IIRC, RGB can be analog or
digital. So I doubt it is a digital format.

Most video -- whether analog or digital -- contains a red signal, a
green signal, and a blue signal. When all three colors are lit to full,
the result appears white. When none are lit, the result is black.

What is the file extension for the format you are talking about? Is it
..rgb? I doubt it

Just like audio files can have .wav, .pcm, and .raw extension, video
files should have an equivalent? What are the video equivalents of
those extensions?

So your answer is "no, I really don't even know what
PCM is," right?

PCM is digital representation of an analog signal [such as video or
audio]

No, the term "pixel" is used in a digital video context
interchangeably with "spatial sample." A "channel" is
something altogether different. Again, you clearly are
completely unfamiliar with the field in which you are trolling.

I am simply trying to find out the video equivalents of the terms used
in digital audio [e.g. "sample rate"]

A "sample rate" is a completely independent question from
the modulation system used to encode or transmit information.
Sampled systems are no more required to use PAM than your
other new favorite buzzword, PCM.

If audio uses it, then what is a visual equivalent of it?

Incorrect. The pixel rate in video is, if anything, most closely
related to the more general term "symbol rate" (or "baud
rate"), at least to the extent that until you define the number of
bits per pixel, the data rate (in bits/second or whatever equivalent
unit you like) cannot be determined. Similarly, "sample rate"
by itself does not determine the data rate.

I am well aware that sample-rate does not determine data-rate on its
own. Sample rate, bit-resolution, and number of channels, put together,
are what determine the data rate.

Data rate = sample rate X bit-resolution X number of channels

CD audio data rate: 1,411,200 bits per second
CD audio bit resolution: 16-bit
CD audio sample rate = 44,100 hz
# of channels in CD audio = 2

1,411,200 = 44,100 X 16 X 2