|
|
Posted by Jukka Aho on 11/10/06 02:08
Phil M wrote:
> Hi all. I'm trying to figure out if there's any visual (to the eye)
> difference between 4:1:1 and 4:2:2 footage?
>
> I recorded footage from Betacam SX camera master via the SDI port to
> DVCAM and upon playback, I couldn't visually see any difference in
> colors.
Comparing the differences might be easier on a computer. For example,
with a suitable program, you could view the color channels alone
(without the luminance information), and even make a synchronized
side-by-side comparison of the two color formats.
Can you capture raw SDI data to a Windows computer? And if you can, did
the SDI capture/interface card come with a DirectShow filter for viewing
and processing that captured SDI video data in ordinary Windows
programs?
(For example, the leftmost card on this page would appear to come with
such filter and capability:
<http://www.computermodules.com/broadcast/smpte259M-sdi-PCI.shtml>)
> I understand the theory that in 4:2:2 the chroma components are
> sampled at half the sample rate of luma, and in 4:1:1 the horizontal
> color resolution is quartered. But visually, is there anything more
> that the eye can see when NTSC video plays at full motion? Or the
> difference shows up only on color measuring equipment or when one
> grabs a screenshot
> still and magnifies it?
Not exactly what you want, but see here anyway... for some ad-hoc
comparisons of 4:1:1 to 4:2:0:
<http://www.adamwilt.com/pix-sampling.html>
You're apparently asking for some kind of subjective assessment of the
differences between 4:2:2 and 4:1:1, but that's of course... subjective.
I guess you would have to arrange a double-blind test (with some
well-prepared test clips) to find out what percentage of people will
actually notice and object to the difference - if given the chance.
Note that these days it is entirely possible that "NTSC video" (digital
video with NTSC resolution) is not any more displayed in its "native"
resolution where a single pixel row of digital data would neatly
correspond a single scanline on the CRT screen, or a single row of LCD
or plasma pixels. Instead, with all the modern innovations - PC TV tuner
cards, HD resolution LCD and plasma screens, etc. - it is becoming
increasingly common that video is first deinterlaced (sometimes with
crummy algorithms that leave a lot to be desired) and then shown on a
progressive display device, interpolated to the native resolution of
that device, and to some huge physical size. That kind of complicated
digital processing could also make the differences between the various
color subsampling formats much more visible than what you might think
when you're only viewing the video on a computer screen, in a tiny
window with a 1:1 pixel mapping, or on a small video monitor.
Then again, there's often little you can do about subsampling formats,
anyway, if your video ends up being broadcast, or on a DVD (4:2:0).
(Which subsampling format(s) do they use for the ATSC broadcasts?)
--
znark
Navigation:
[Reply to this message]
|