# Currently-Available Highest-Quality Linear PCM Video?

Discussion in 'Electronic Basics' started by Radium, Oct 18, 2006.

Hi:

What are the sample rates and picture resolution [in pixels X pixels]
of the professional progressive [non-interlaced] linear-PCM video
format used today? I beleive linear-PCM video signals are used in
professional studios.

Thanks,

2. ### Guest

Start here

http://catalog2.panasonic.com/webap...toreId=11201&catalogId=13051&catGroupId=14594

and here

and here

http://www.thomsongrassvalley.com/products/film/spirit_4k/

There are MANY more.

GG

3. ### Bob MyersGuest

If you're talking about standard-definition television, and
the case of standard sampling formats for the conversion
and storage of "analog" video into digital form, the most
common international standard today is probably CCIR-601,
which uses a common 13.5 MHz sampling rate for both
525/60 and 625/50 video systems; this results in image
formats (not "resolutions," please) of 720 pixels x 480 lines
for the former and 720 x 576 for the latter. I don't know
what you mean by bringing "linear-PCM" into a question of
sampling rates and image formats.

Bob M.

What equation did you use to get the numbers 720, 480, and 576?
Linear PCM is uncompressed PCM. Thats what I am talking about. Sampling
rate must be at least twice the highest frequency in the signal. Due to
physical conditions, it is safe to makes the sample rate at least 2.5x
the highest frequency signal.

In NTSC, the horizontal frequency is 15.734 kHz, the vertical frequency
is 60 Hz, and the color subcarrier frequency is 3.579545 MHz. The means
that the horizontal sample rate must be 39.335 khz or higher, the
vertical sample rate must be at least 150 hz, and the color subcarrier
sample rate must be no less than 8.9488625 mhz.

What is the pixel X pixel resolution -- or "format" if you wish -- of
today's first-class video signal? Surely it would have to be more than
720 X 576. My monitor is displaying a pixel X pixel -- or "screen area"
-- of 1280 X 1024 with 32-bit color. I am not sure of the frequencies
of my monitor.

Also, is there supposed to be a special difference between the first
number and the second number [such as 1280 X 1024 or 720 X 576]?

Thanks,

http://en.wikipedia.org/wiki/CCIR_601

CCIR-601 uses interlaced -- not progressive -- signals. I like
progressive and dislike interlaced.

6. ### Guest

You really need to go look things up and not argue about things you
apparently know little to nothing about. Bob Myers told you exactly
right about SD TV. You can go look this up- the sample rate is 858 x h
rate = 13.5MHz for luma. The 720 is the active number of samples out of
the 858. That 'dead' time is to allow the CRT based monitors time to
retrace the Horizontal. The 480 line is computer talk for 525 line
system of which 483 are active. The remaining lines are to allow
Vertical retrace of a CRT monitor. If you think about it, LCD and
Plasma do not require retrace since there is no scanning, just counting
off samples. HDTV works in a similar fashion but the numbers are all
different. Try ATSC.ORG. The monitor pixel ratios are often 4:3 or 16:9
to make square pixels but none of that is etched in stone. Rectangular
pixels are often used and just re-map the square into rectangular. Now
go study.

GG

SDTV is not "first class"
What about the color subcarrier sample rate? The horizontal frequency
sample rate?
That is why plasma and LCD are better than CRT. In addition, plasma is
better than LCD.
IIRC, HDTV is a tad closer to "first class" than SDTV. Again, I could
be wrong.

8. ### Guest

You might be surprised if you saw a serial digital feed on a broadcast
monitor. Many folks would find it to be totally satisfactory -- if they
ever got to see it.
There is no subcarrier for component digital. You didn't look it up.
The subcarrier is introduced when the component is encoded into
composite. This does not happen with DVD or SD Digital TV. The
horizontal 'sample rate' is a continuous 13.5 MHz with a new line
starting on every 858th sample and the vertical 262.5 X 858 samples.
The 2 chroma difference channels are each sampled 429 x H for another
858 time periods for a total data rate of 27MHz sample rate. Bump those
numbers up to 150MHz for HD.
That has nothing to do with why LCD or Plasma is 'better' than a CRT.
You will find a good number people who think the CRT is a better image
even in HD. Much of what is 'bad' about the CRT is the support
electonics, mainly poor power supply regulation.
GG again

AFAIK, it maybe "satisfactory" but it is not the best quality currently
available.
Is HD is best quality currently available? How much is the pixel X
pixel screen area in HD? My LCD computer screen is 1280 X 1024 pixels.

I prefer "first class" uncompressed linear-PCM video signal viewed
through a "first class" plasma screen.
Both LCD and plasma are more resistant to EMI/RFI than CRTs. Plamsa
offers better clarity than LCD or CRT.

10. ### Guest

Quite simply put, you're not going to get it now or in the near future.
The best you can get today is MPEG2 HDTV over the air with an ATSC
receiver delivering 19.34 mega BITS/second. That's compressed between
75 and 80:1 and that's the BEST available. Remember HDTV pushes out 150
Mbytes/second and those are not 8 bit but 10 bit so your 1.5Gbits gets
dropped to 19.34Mbits. But, I just got done watching NCIS in HD and it
really looked quite good. So go spend some money on a 1920x1080 native
resolution set with ATSC, put up an antenna and have at it. I've been
watching HD for nearly 3 years and have 2 computers to record HD OTA
feeds. They work very well and you don't need the latest whiz-bang
computer to do it. A Sempron 2500 will do just fine.
More of the support electronics issues but I agree that the CRT is not
what I want. Non-CRT sets all suffer from math rounding error noise
when converting the digital values to PWM to get variable brightness.
CRTs with all their faults do not have that issue.

<snip>

GG

11. ### Bob MyersGuest

No equation; that's what the standard defines. The 720
is the number of active samples per line; if you divide the period
of one horizontal scan line up using a 13.5 MHz sampling rate,
you will get 858 samples per line (in the case of the standard
"NTSC" line rate, 15.73426 kHz). Of these, 720 are "active"
(contain information corresponding to the actual displayed
image). The 525/60 scan format also has about 484 lines per
frame of active video (i.e., what's left - 525 minus 484 lines -
is used in vertical blanking), and CCIR-601 decided to use a
standard format of 480 lines just to round it to a convenient
number. Similarly, there are about 576 active lines in the
625/50 scan formats.
Which has absolutely nothing to do with how the data is
encoded, which is what "PCM" is all about - hence I have no
idea why you are bringing that term up in a question regarding
sampling rates. Note that the 13.5 MHz rate easily meets the
Nyquist requirement of being >2X the bandwidth (technically,
it's bandwidth, not "the highest frequency in the signal) of the
video (note that U.S.-standard video occupies roughly
5 MHz of a 6 MHz channel).
Standard definition video - i.e., anything other than high-definition
TV, which has different standard formats - is never any better
than about 720 x 480 pixels/lines for the U.S. 525/60 standard
format, or 720 x 576 for the "European" 625/50 format. The
actual effective resolution, in the proper sense of that term (the
degree to which detail can actually be resolved in the displayed
image) is generally a good deal less than this. Yes, computer
formats contain considerably more pixels than anything SDTV is
capable of.

Standard HDTV formats are 1280 x 720 pixels and 1920 x 1080
pixels.
The way it's normally stated in PC usage, the first number is the
number of pixels per line, while the second is the number of lines
per frame. Traditional TV practice has been just the opposite -
give the number of lines first, then the number of "pixels" - but
actually, until the advent of digital television there really wasn't any
such thing as "pixels" in the TV engineer's vocabulary.

Bob M.

12. ### Bob MyersGuest

In that case, what do you think "first class" means? You asked
about video as it appears in professional studios - unless you're
Those phrases are nonsense. There can be different sample rates
used for the "chrominance" components vs. the luminance (i.e.,
go look up "4:2:2" and "4:2:0" sampling as opposed to "4:4:4"),
but you would never refer to this as the "color subcarrier sample
rate."
Actually, this has very little to do with the relative performance of
these technologies. And before we start talking about which is
"better," you'd have to ask what "better" means. "Better" in what
aspect?

Since you have yet to tell anyone just what you are thinking of
when you say "first class," it's impossible to respond any further.
Do you mean the digital cinema standards?

Bob M.

13. ### Bob MyersGuest

Yes, it IS, in terms of entertainment video. Any "television"
type programming you are going to encounter is either SDTV
or HDTV; the vast majority is SDTV, and you would be
surprised at just how good a 720 x 480 image can look on
the right display.
You still don't seem to have any clue at all how "linear PCM"
plays into all this...

No, those technologies are more resistant to MAGNETIC
interference than the CRT; this is not "EMI" or "RFI." It is
also impossible that either plasma or the LCD would be superior
vs. the other in terms of "clarity," if by that you mean delivered
image resolution. Both are fixed-format technologies, so when
driven at their native format, each pixel of the image is as clear
as it's ever going to get.

Bob M.

14. ### Bob MyersGuest

Correct, but not relevant to the question you asked, which
concerned standard sampling rates and the resulting image
formats. If you're talking about sampling video, you are talking
about converting an analog video signal to digital form - and
ALL analog standard-definition broadcast video standards use
interlaced scanning.

There are similar standards for progressive-scanned video,
studio practice, they were also not relevant to the question.

Since you seem to be able to find Wikipedia, why aren't you

Bob M.

15. ### Michael A. TerrellGuest

He can't troll the wiki.

--
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

I asked a question about linear PCM video 3 years ago. Here was my
response:

"There is uncompressed PCM for video. The data rate is 270 Mbit/sec for

standard definition, 525 line, 60 field interlaced. The computer folks
refer to this as 480i. For Hi def, the data rate bumps up to 1.5
Gbit/sec for 1920x1080 interlaced. Looks danged fine,too. This stuff
is only seen in studios and post production facilities. It makes going
to work fun. "

My question is what are the sample rate and color-resolution for the
"Hi def" mentioned above with 1920 X 1080? And why is it interlaced? I
want it progressive.

Linear-PCM is uncompressed digital info. If linear-pcm is used for
audio [e.g. WAV files], then why not for video??

Here is my insane type of premium video:

Linear PCM video [with at least 320-bit color resolution, at least
109,200x100,800 pixel progressive [non-interlaced] picture resolution,
with a sample rate of at least 1,350 THz sampling rate viewed via a
plasma screen with at least 1,000x the capabilities of the
aforementioned values.

17. ### Bob MyersGuest

Note that the sentence "there is uncompressed PCM for video"
does not equate to "all digital video is PCM" or "PCM is required
for high-quality digital video." PCM is simply one possible encoding
and transmission scheme, nothing more - and it is not the one used
in most digital video systems.
The sample rate can be determined from the pixel format (in
this case, 1920 x 1080), the frame or field rate (60 Hz, typically,
in the U.S.), and the amount of overhead time required for
horizontal and vertical blanking. If you didn't need ANY
blanking time, then the minimum sample rate is simply

1920 pixels/line x 1080 lines/frame x 30 frames/sec (it's interlaced)

Note that if you work the units out as well, it comes out in terms
of pixels/second, exactly as it should.

It can be; it simply doesn't HAVE to be, so there is no need
to be dragging that question in at this point.
You've got that right - it's insane. Here's why:
OK, now I see what you mean by "color resolution" - you
are talking about what's more commonly referred to as
"color depth" or "dynamic range." 320 bits/pixel (presumably,
something over 100 bits/color) is absurd; there is no display
device that can provide this range, nor can the human eye
deal with it. Somewhere around 10-12 bits/color, properly
encoded, is about the maximum required, and most systems
reduce the effective data rate by limiting the spatial resolution
in the chroma channel (i.e., you don't really get as many bits
PER PIXEL for color as you think you need).
Again, absurd numbers. The eye cannot resolve detail
beyond a certain point (approx. 60 black/white cycles per
visual degree is a good rule of thumb for the maximum),
and anything above that is a waste. But this means that you
can't just concern yourself with the number of pixels in
the image - the image size as displayed to the viewer and the
expected viewing distance also come into play.
Sample rate is never a system requirement, except in
terms of a maximum permissible sample rate to fit within
system bandwidth constraints. The sample rate REQUIRED
for a given pixel format and frame rate is driven by those
parameters, and then you just see whether or not it's going to
fit in the available bandwidth.

As usual, it seems you haven't learned anything at all about
the field you're trolling in before making up absurdities.

Bob M.

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

For example, CD audio has a sample rate of 44,100 hz but a bit rate of
1,411,200 bps.
Linear-PCM doesn't have to be used but what harm is caused by using it?

19. ### Daniel MandicGuest

Depends to the CD.... not every Compact Disc Audio is full 16bit.

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

Best Regards,

Daniel Mandic

P.S.: But 1,411,200 bps can happen. ;-)

20. ### Guest

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