Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Charles Poynton » Thu Apr 08, 2010 10:46 am

Colleagues:

I'm happy to be working with SpectraCal to bring you Poynton’s Vector, planned for monthly release. Issue 1, "CONTRAST, BRIGHTNESS, and the naming of things," was distributed with the newsletter a few weeks ago. You're welcome to use this thread to discuss issues raised there.

Expect Issue 2, "COLOUR, TINT, and the naming of things," within the next week or so, at which time I'll create a thread for discussion of that one.

- Charles
Charles Poynton
Calibration Soldier
 
Posts: 27
Joined: Wed Apr 07, 2010 2:16 pm
Location: Toronto, Ontario, Canada

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Bob Walters » Fri Apr 09, 2010 4:58 am

Charles,
This forum & Spectracal "might" not realize how lucky we are to have you on board. I've been enjoying color science with all its ramifications i.e calibration, lighting et als since 1999 and can honestly say, reading your books, articles, etc., etc. I have ALWAYS learned something that, put in lay terms, explains a question I've had for a long long time!!! Thanks again for joining us!!!!!!!!!!!!!!!!!!!!!
Bob Walters
Calibration Hero
 
Posts: 417
Joined: Tue Aug 05, 2008 10:46 am

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby andrewfee » Fri Apr 09, 2010 5:41 am

I would also like to thank you for joining. I must admit, I skipped over reading your article as I would like to think that I know what I am doing when it comes to brightness/contrast these days, but having just read it now, it was actually rather interesting.

I would agree with you that there does not seem to be much need for the brightness (black level) control on modern displays with good sources - then again, I believe that most display controls could actually be eliminated with an all-digital chain (thinking ahead, it's perhaps something a newer revision of the HDMI standard could introduce) and that the traditional naming of these controls should be changed. Black level and white level are perhaps the most accurate choices, but black level and brightness would probably make more sense to the average consumer.

The problem with eliminating the black level control on displays though is when it comes to dealing with less than ideal sources. I suspect that even with Blu-ray there are probably some discs out there that have an elevated black level encoded to the disc, and there are certainly DVDs out there that do. Not that I watch it any more, but broadcast television can vary from programme to programme and from one channel to the next, so eliminating the control entirely may cause more problems than it fixes.


I look forward to reading your next article, and eagerly anticipate when you get around to discussing gamma. (something that should have been taken care of years ago and built into the standards in my opinion)
andrewfee
Calibration Hero
 
Posts: 274
Joined: Wed Jun 14, 2006 12:29 am

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Scott_R_K » Fri Apr 09, 2010 2:26 pm

A suggestion , would it be possible to place a Link , to each Article , in your first Post in each Thread ? This would make it easier for everyone to quickly "jump" back to the Article during any discussions .

Also , thanks for joining and participating on SpectraCAL .

Scott............... 8)
Scott_R_K
Calibration Hero
 
Posts: 152
Joined: Tue Aug 15, 2006 5:21 pm

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby cal87 » Sun Apr 11, 2010 1:34 pm

Scott_R_K wrote:A suggestion , would it be possible to place a Link , to each Article , in your first Post in each Thread ? This would make it easier for everyone to quickly "jump" back to the Article during any discussions .

Also , thanks for joining and participating on SpectraCAL .

Scott............... 8)


I must have deleted the newsletter with the link to the first article. Fortunately, I just received the latest newsletter and was able to find the links to both.

http://www.spectracal.com/poyntons_vector.html
cal87
Calibration Soldier
 
Posts: 31
Joined: Tue Oct 23, 2007 6:02 am

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Charles Poynton » Sun Apr 11, 2010 5:56 pm

Hey, thanks for the encouragement, guys.

andrewfee says,

“I suspect that even with Blu-ray there are probably some discs out there that have an elevated black level encoded to the disc.”

The second issue of Poynton's Vector states Poynton's Fourth Law (in the margin of the second page): Once a program is mastered, errors in mastering are indistinguishable from expressions of creative intent. It's a serious problem: How, in general, can you tell that the black level was mistakenly lifted as opposed to deliberately lifted? Producers don't put colourbars on the front of a Blu-ray transfer - nor should they. The onus should be on the production to get it right.

It's clear to me that a BLACK LEVEL control will always lurk under the covers somewhere, accessible to advanced users. My view is that 99.999% of consumers can't identify the functions of BRIGHTNESS and CONTRAST, and so I'm with Apple that if you remove one of the controls, the consumer can then figure out what the other one does!

Cheers,

- C
Charles Poynton
Calibration Soldier
 
Posts: 27
Joined: Wed Apr 07, 2010 2:16 pm
Location: Toronto, Ontario, Canada

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby andrewfee » Sun Apr 11, 2010 11:20 pm

I suspect that the chance of it happening with Blu-ray is slim, but certainly with broadcast material and some DVDs, there are cases where the letterboxing/pillarboxing of the content is elevated above video black. In that case at least, surely it is a mastering error.

I do agree though that even if there is no detail there, sometimes the black level may be raised for artistic reasons, but I would only expect that to affect the actual content and not the letter/pillarboxing.


Apple is able to eliminate the brightness (and every other) control because everything is set with ICC profiles using the video card LUT. (at least I don't think they have a higher precision internal LUT) When hooked up to a computer this is almost certainly a non-issue, but it could be a problem with other sources. Being a computer monitor with no controls, I suspect it is set only for PC levels and if you had a Video-level source you would not be able to display it correctly without a brightness control for example.

I would love to see it disappear, but I do think a black level control is still necessary on a display, if only for use with old/bad sources.


Actually, thinking back a year or so, I had a Pioneer KRP-500M. When brightness was set to the point that black was at its lowest level (which should have been the correct setting, as it is with most modern displays) shadow details that I confirmed to be above video black were not being shown. The only option to see all the detail near black was to raise the black level with the brightness control; the multi-point gamma correction did not help. Ideally the display would be set so that there was no need for a brightness control and you had the deep black level in addition to good shadow detail but that was not the case. In the end that was one of the reasons I got rid of the display; what is the point in paying a premium for a set with good black levels if you lose that advantage to see all the detail in a source?
andrewfee
Calibration Hero
 
Posts: 274
Joined: Wed Jun 14, 2006 12:29 am

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Angryht » Tue Apr 13, 2010 1:55 pm

What about setting white level (contrast). Should it be set to show all levels up to 254 or clipped at 235?
Angryht
New User
 
Posts: 6
Joined: Fri Apr 27, 2007 4:43 pm
Location: Omaha, NE

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Charles Poynton » Wed Apr 28, 2010 7:42 pm

Angryht wrote:What about setting white level (contrast). Should it be set to show all levels up to 254 or clipped at 235?


A display ought to follow a reasonable gamma-curve all the way up to (8-bit) code 254. See the post over in the YCbCr v RGB, Which color space is bible? thread http://www.spectracal.com/forum/viewtopic.php?f=51&t=1962&start=0.

- Charles
Charles Poynton
Calibration Soldier
 
Posts: 27
Joined: Wed Apr 07, 2010 2:16 pm
Location: Toronto, Ontario, Canada

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Angryht » Fri Apr 30, 2010 7:54 am

Thanks for the reply and the reference. I appreciate it.
Angryht
New User
 
Posts: 6
Joined: Fri Apr 27, 2007 4:43 pm
Location: Omaha, NE

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Angryht » Fri Apr 30, 2010 11:20 am

I guess I'm still a bit confused. In the referenced thread you have this quote:
Charles Poynton wrote:Best to think of reference black and reference white codes originating with R’G’B’ abstract values 0 and 1 respectively. Upon coding into 8-bit integers across the interface, the reference luma levels (instead of “valid” levels) are 16 and 235. I recommend that studio equipment and consumer equipment allow excursions down into the footroom and up into the headroom regions, that is, down to 1 and up to 254 in 8-bit space. (10-bit video should run 4 ... 1019 inclusive, with luma reference codes 64 and 940.) Chroma reference codes are 16 and 240 in 8-bit space (that is, 64 and 960 in 10-bit space).

I’ll write about clipping in June's Vector issue, but the short story is that I have consulted a dozen of my studio colleagues and all agree that no clipping should take place - transients into footroom and headroom should be preserved as long as possible (as ChrisWiggles has posted here). BUT - many post-production facilities are nervous that not-completely-clueful QC operations will reject any material that has excursions into these regions, and rather than fight, the post houses cave, and clip, knowing that they shouldn’t.

Stay tuned, though - wide-gamut is coming. When an appreciable number of studios take interest, then clipping has to stop.

- Charles

Maybe I'm missing something but are you advocating setting up white and black levels to allow 1 through 254 to be visible? I always thought along the lines of what Chris discusses below. Like I said, I think I'm missing something.
ChrisWiggles wrote:As we've moved to digital displays that don't really float around in black level as a CRT does, there is less need for anything below 16 in my opinion, because for most black level alignments in a proper environment, you will align the display's black point right at 16. There is still some minimal impact on any video processing in the chain, but I think it would be fair to characterize those impacts as pretty insignificant to the viewer, even a very critical viewer.


I am looking forward to reading the June Vector issue.
Angryht
New User
 
Posts: 6
Joined: Fri Apr 27, 2007 4:43 pm
Location: Omaha, NE

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Joel Barsotti » Fri Apr 30, 2010 1:33 pm

I don't mean to put words in Charles' mouth, but I believe what he's saying is the BTB and WTW should both be passed to the display.

Then at the display you use a Pluge pattern to align the displays black level to 16.
Then for white you setup the display to disclose up to 254.

So the net result is your display is reproducing 16-254.
SpectraCal Developer
User avatar
Joel Barsotti
Site Admin
 
Posts: 3707
Joined: Thu Dec 03, 2009 12:11 pm
Location: Seattle, WA

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Charles Poynton » Mon May 03, 2010 6:07 pm

Joel said,

I don't mean to put words in Charles' mouth ...


... but, well said! 8-bit codes 0 and 255 are prohibited in encoding and across an SDI interface, but processing should not mess with any other codes. Joel is correct that you would expect a properly set display to clip (or at least render into invisibility) codes below 16, but all else up to 254 should display distinctly.

By the way, where's issue 3? (Just kidding: Expect it within hours ...)

- C
Charles Poynton
Calibration Soldier
 
Posts: 27
Joined: Wed Apr 07, 2010 2:16 pm
Location: Toronto, Ontario, Canada

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Angryht » Wed May 05, 2010 8:17 am

Thanks to both of you for your answers.
Angryht
New User
 
Posts: 6
Joined: Fri Apr 27, 2007 4:43 pm
Location: Omaha, NE

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Mr.D » Wed Jun 02, 2010 4:33 am

I used to be in the camp that said disclose all the way up to 254 on display.
However after conducting various tests with a range of consumer video formats I've come to the conclusion that the vast majority of the range above 235 is mostly noise generated by chroma upsampling and scaling stages in the production chain.

We struggled to find any imagery that consequently looks to be missing real picture information when clipped at 235 ( or display white mapped to this point).

Targeting 240 as a nominal peak white level on display seem to provide a good level of contrast with a bit of additional wiggle room for any mastering that has overshot 235.

Obviously individual setups come into play: some displays will not clip regardless of how hard the contrast control is set , however some people are able to target the 235ish white point with offboard CMS systems.

Generally speaking targeting the 240 white ref gives an improvement in contrast without unnecessary clipping in specular highlight detail and affords improvement in end display appearance.

This is with reference to end display in RGB.

I found no real reason to target 254 on the display despite investigating more esoteric signalling issues and subsequent gamma modification.

My own setup targets 254 on the display hardware as it will not clip ( although overly high contrast induces gamma bumps and run out in blue) and I use a PC based CMS to target a white point around 240.

Properly mastered material does seem to rigidly target 235 (level above seems to be ringing/noise) , broadcast footage not afforded the luxury on non-realtime final grading can excurse higher but clipping lower than 254 again does not seem to compromise white detail to any appreciable extent even with this material: the footage that exhibits detail higher than 235 seems to be so blown out anyway that there is little if any detail actually restored ( and I mean literally a couple of pixels that loosk suspiciously like ringing anyway)

As always try it and see but I do feel that mapping peak white lower than 254 down to 240 has few if any disadvantages and visible improvements to offer in contrast. ( and possibly end display precision but I haven't reached any concrete enough conclusions with that).
Mr.D
New User
 
Posts: 8
Joined: Mon Aug 28, 2006 8:44 am

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby andrewfee » Thu Jun 03, 2010 8:37 am

I would agree with your findings that there doesn’t seem to be much, if any, content that has actual picture information above 235.
My recommendation would be to show 16–254 where possible, but that it is acceptable to use 16–235/240 if you need the extra light output.

On most direct-view displays where you have more light output than you need I would stick to using 16–254 (I follow EBU guidelines with my displays, so video white is at 80 nits, peak white at 100 nits) but if I were using a projector and could not reach my target brightness using 16–254, I would have no problem using 16–235/240 for the extra light output because there seems to be very little, if any, consequence for doing so.

I would be interested in hearing what you conclude regarding end display precision. Consumer displays are generally awful in this regard and could use any help they can get—that’s why I am stuck using CRT again and have currently removed my VideoEQ from the chain.
andrewfee
Calibration Hero
 
Posts: 274
Joined: Wed Jun 14, 2006 12:29 am

Re: Poynton’s Vector 1 - CONTRAST, BRIGHTNESS, ...

Postby Mr.D » Thu Jun 03, 2010 1:11 pm

On an admittedly oldish plasma that is far from transparent to 8bit I do seem to see an improvement in banding targeting 235uish rather than 254, however there are a couple of stages in the chain that would fog the issue somewhat. Its going to depend on what sort of processing is going on between source and display and whether any of those stages can benefit from freeing up some precision.

As ever try it and see . I think there is enough evidence out there now to say that mapping lower than 254 does not entail any compromise in actual picture information so its something people should be trying without feeling they are violating some meaningful edict.
Mr.D
New User
 
Posts: 8
Joined: Mon Aug 28, 2006 8:44 am


Return to Poynton's Vector

Who is online

Users browsing this forum: No registered users and 1 guest

cron

x