Re: The color of QuickTime 5 (Gary Adcock's response to Don's comments)
Re: The color of QuickTime 5 (Gary Adcock's response to Don's comments)
- Subject: Re: The color of QuickTime 5 (Gary Adcock's response to Don's comments)
- From: Don Hutcheson <email@hidden>
- Date: Fri, 05 Oct 2001 15:23:39 -0400
Gary,
I agree completely. You neatly point out why the Mac is the only viable
platform for color management today, and why 'color managing' video is still
impractical, at least in a real-time playback sense.
Half a century ago when the NTSC had the foresight to define the assumed
receiver color space. If displays adhere reasonably closely to the NTSC
standard, and especially to the gamma value, the need for color management
in play environments is small to zero.
On the creation side the best (and "legal") solution is to edit on a good
NTSC-adherent display and then assume the recipient will view in the same
color space. Joe couch-potato still won't see what we see, because we can't
really measure or control his display. But if the guy is a videophile and/or
intelligent enough to "profile" his display in an ICC sense, then he should
also be able to adjust the few hardware controls most displays come with to
achieve a best approximation of the NTSC standard.
If we ever want to solve the TV-store window effect (20 different colors on
20 different sets) then the best path is better displays with less tendency
to drift, or perhaps some type of auto-calibration, for which the goals were
clearly stated 50-plus years ago by the much-maligned but still (IMHO)
brilliant NTSC.
It's also wise to remember that, like Dolby NR, color management actually
adds distortion to an image. We get away with this in still imaging because
the output process is so noisy that most of the distortion is invisible or
an overall improvement. But I suspect that in video any serious effort to
alter the video on the fly in a matrix or 3-D way may lead to nasties like
quantization (e.g. banding in blue skys,) dead flesh, etc.
What I'd like to know is this. What real-world video problems would dynamic
3-D ICC-based color management solve that couldn't be 90% fixed on a monitor
with NTSC-ish RGB coordinates by simple display adjustments like white
point, gamma and gray tracking?
*************************************
Don Hutcheson
Hutcheson Consulting
(Color Management Solutions)
11 Turnburry Rd
Washington, NJ 07882
Phone: (908) 689 7403
Fax: (908) 689 5305
Mobile: (908) 500 0341
email@hidden
*************************************
>
> At playtime, QT would extract the gamma and white point from a profile
>
> (attached to the clip or specified independently as a QT default) and use
>
> those data to build the LUTs. Any 3-D LUT and/or matrix in the profile would
>
> probably have to be skipped due to the huge processing burden, meaning
>
> color-dependent functions like saturation and hue correction would not be
>
> applied. But gamma and white point are usually the most significant part of
>
> any profile and for video I'd say honoring the intended Gamma would be a
>
> great first step towards QT color management.
>
>
>
Ok Don
>
>
as a newbie on this list, but "nominally" versed in color --under the
>
video standard this is a great idea on a computer but how is this
>
going to be implemented on NTSC displays or worse yet on standard
>
tv's ?
>
While the ref monitors used in TV studios may be able to
>
handle the modified color/gamma data via play back via an RT video
>
card with component out (digital voodoo/media100/ cinewave )
>
How is it going to work from an avid (as proprietary as avid
>
is but still regarded as the defining video standard just as
>
photoshop is for print ) wintel systems or on the millions of HD sets
>
yet to be delivered and the myriad of types connectors that will be
>
used.
>
How is is this info transferred via a composite video (single
>
cable ie. rca or bnc) signal instead a of component ( YUV )or optical
>
connector being used between other motion video devices.
>
How is the gamma and white point going to be handled on sets
>
with dynamic contrast (gamma) adjustment based on ambient room light
>
as this is standard on all high end NTSC sets shipping today, and I
>
have not even added PAL with it's larger base color space and wider
>
dynamic range that's used in the rest of the world.
>
>
ga
>
--
>
gary adcock
>
studio37
>
chicago,il 60612
>
312-243-3244
>
email@hidden