Poynton’s Vector 3 - Gamma – 2.0, 2.2, or 2.4?

Poynton’s Vector 3 - Gamma – 2.0, 2.2, or 2.4?

Postby Charles Poynton » Tue May 04, 2010 2:52 pm

Colleagues:

It seems that a few copies of the forthcoming Issue 3, Gamma – 2.0, 2.2, or 2.4? have leaked out prior to the official release (within hours, I hope). When it's posted – or even before! – please discuss it here!

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

Re: Poynton’s Vector 3 - Gamma – 2.0, 2.2, or 2.4?

Postby JohnAd » Wed May 05, 2010 1:31 am

:D Don't think it was a pre-release, It was linked from

http://www.spectracal.com/poyntons_vector.html

yesterday.

I'm interested in your comments that studios are approving masters in very dark conditions, to some extent this contradicts your description of current practice in the linked open letter. Could you give a bit more flavour about this difference and what values you would suggest using for white luminance, veiling glare, surround luminace as inputs to a CAM for calculating a user's gamma curve if the aim is replicating the perceieved image under approval conditions.

Thanks

John
JohnAd
New User
 
Posts: 5
Joined: Sat Apr 24, 2010 12:02 am

Re: Poynton’s Vector 3 - Gamma – 2.0, 2.2, or 2.4?

Postby Charles Poynton » Wed May 05, 2010 11:40 am

JohnAd writes,

... studios are approving masters in very dark conditions, to some extent this contradicts your description of current practice in the linked open letter ...


I have made some measurements in mastering rooms recently, and I am in the course of updating my old-fashioned views on the topic. The semi-published “open letter” on my web site http://www.poynton.com/notes/PU-PR-IS/index.html hasn’t yet been fully updated to reflect current practice. At some point perhaps the measurement results will be explained and published.

In short: HD mastering today for long-form content (e.g., for Blu-ray) is done on BVM-D24 or D32 displays (or displays approximating those). Gamma (as a pure power function) is about 2.4; reference white is around 80 nt. The viewing environment is very dark – maybe 1 lx ambient illuminance and 1% surround.

If you have a home theatre situation approximating those conditions, then you should use 2.4 gamma like the BVM. If the display and/or viewing environment is brighter, then gamma should be reduced somewhat, to say 2.3 or maybe even 2.2. Exactly how much? Sorry, that's still a research topic, as far as I can tell.

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

Re: Poynton’s Vector 3 - Gamma – 2.0, 2.2, or 2.4?

Postby JohnAd » Wed May 05, 2010 1:57 pm

Charles Poynton wrote:In short: HD mastering today for long-form content (e.g., for Blu-ray) is done on BVM-D24 or D32 displays (or displays approximating those). Gamma (as a pure power function) is about 2.4; reference white is around 80 nt. The viewing environment is very dark – maybe 1 lx ambient illuminance and 1% surround.

Great, thanks a lot. I look forward the next issue.

John
JohnAd
New User
 
Posts: 5
Joined: Sat Apr 24, 2010 12:02 am

Re: Poynton’s Vector 3 - Gamma – 2.0, 2.2, or 2.4?

Postby vyogee » Fri May 07, 2010 5:48 am

Charles Poynton wrote:JohnAd writes,

... studios are approving masters in very dark conditions, to some extent this contradicts your description of current practice in the linked open letter ...


I have made some measurements in mastering rooms recently, and I am in the course of updating my old-fashioned views on the topic. The semi-published “open letter” on my web site http://www.poynton.com/notes/PU-PR-IS/index.html hasn’t yet been fully updated to reflect current practice. At some point perhaps the measurement results will be explained and published.

In short: HD mastering today for long-form content (e.g., for Blu-ray) is done on BVM-D24 or D32 displays (or displays approximating those). Gamma (as a pure power function) is about 2.4; reference white is around 80 nt. The viewing environment is very dark – maybe 1 lx ambient illuminance and 1% surround.

If you have a home theatre situation approximating those conditions, then you should use 2.4 gamma like the BVM. If the display and/or viewing environment is brighter, then gamma should be reduced somewhat, to say 2.3 or maybe even 2.2. Exactly how much? Sorry, that's still a research topic, as far as I can tell.

- C


Mr. Poynton,
Thanks for the interesting subject of gamma. It was a good read but it brought up some questions.

In a home theater environment, for fixed pixel displays (if I got the term wrong, I actually meant, Plasma/LCDs), would you recommend a bias lighting behind the display. There was some mention that SMPTE specs call for a bias lighting behind a display.

This gamma stuff is intriguing. When I calibrate my display, I find that even at a gamma hovering around 2.35 in dark room is good enough to reveal the necessary shadow details (this is even with a bias lighting behind the display and the windowed 100% white of about 35ft/L). For a brighter image of say 100% window white of 48ft/L I normally target a gamma curve of ~ 2.2.

Not sure if my reasoning makes sense.
vyogee
Calibration Hero
 
Posts: 204
Joined: Mon May 19, 2008 7:53 pm
Location: Malaysia

Re: Poynton’s Vector 3 - Gamma – 2.0, 2.2, or 2.4?

Postby Malmedahl » Mon May 10, 2010 6:27 am

Welcome to the forums Charles! Very interesting to see you here. I bought you'r book Digital Video and HDTV a while ago and its great reading. Highly recommended in-depth reading.

Going to go about and read your articles.
ISF Certified
Malmedahl
New User
 
Posts: 5
Joined: Sun Sep 20, 2009 1:01 am

Re: Poynton’s Vector 3 - Gamma – 2.0, 2.2, or 2.4?

Postby SuperBilleN » Wed Jun 09, 2010 3:18 pm

vyogee wrote:
Charles Poynton wrote:JohnAd writes,

... studios are approving masters in very dark conditions, to some extent this contradicts your description of current practice in the linked open letter ...


I have made some measurements in mastering rooms recently, and I am in the course of updating my old-fashioned views on the topic. The semi-published “open letter” on my web site http://www.poynton.com/notes/PU-PR-IS/index.html hasn’t yet been fully updated to reflect current practice. At some point perhaps the measurement results will be explained and published.

In short: HD mastering today for long-form content (e.g., for Blu-ray) is done on BVM-D24 or D32 displays (or displays approximating those). Gamma (as a pure power function) is about 2.4; reference white is around 80 nt. The viewing environment is very dark – maybe 1 lx ambient illuminance and 1% surround.

If you have a home theatre situation approximating those conditions, then you should use 2.4 gamma like the BVM. If the display and/or viewing environment is brighter, then gamma should be reduced somewhat, to say 2.3 or maybe even 2.2. Exactly how much? Sorry, that's still a research topic, as far as I can tell.

- C



Mr. Poynton,
Thanks for the interesting subject of gamma. It was a good read but it brought up some questions.

In a home theater environment, for fixed pixel displays (if I got the term wrong, I actually meant, Plasma/LCDs), would you recommend a bias lighting behind the display. There was some mention that SMPTE specs call for a bias lighting behind a display.

This gamma stuff is intriguing. When I calibrate my display, I find that even at a gamma hovering around 2.35 in dark room is good enough to reveal the necessary shadow details (this is even with a bias lighting behind the display and the windowed 100% white of about 35ft/L). For a brighter image of say 100% window white of 48ft/L I normally target a gamma curve of ~ 2.2.

Not sure if my reasoning makes sense.


I´ve got the same queston. my 100% windows is 48ft/L and a bias behind the display. would you say that the target should be 2.4 for dim viewing at night and 2.2 for day viewing?
I´ve got an EU Pioneer lx5090 and would say that 2.2 mayby a bit to bright in night
SuperBilleN
New User
 
Posts: 6
Joined: Sun Apr 04, 2010 2:47 am

Re: Poynton’s Vector 3 - Gamma – 2.0, 2.2, or 2.4?

Postby Charles Poynton » Fri Jul 30, 2010 8:34 pm

vyogee writes,
In a home theater environment, for fixed pixel displays (if I got the term wrong, I actually meant, Plasma/LCDs), would you recommend a bias lighting behind the display. There was some mention that SMPTE specs call for a bias lighting behind a display.

The term “fixed pixel” is fine. “Bias” lighting, or what I would call surround lighting, is fine. Historical practice in mastering was to have surround at 5 or maybe 10 percent of reference white, but as I mentioned, current practice is very dark, just 1% or so, and that’s what’s appropriate for faithful presentation. (Perhaps Oprah and Y & R are mastered in a lighter surround, I’d LOVE to know!)
This gamma stuff is intriguing. When I calibrate my display, I find that even at a gamma hovering around 2.35 in dark room is good enough to reveal the necessary shadow details (this is even with a bias lighting behind the display and the windowed 100% white of about 35ft/L). For a brighter image of say 100% window white of 48ft/L I normally target a gamma curve of ~ 2.2.

All of that is in accordance with what I would expect (although I have a preference for nits = cd·m-2 instead of fL).

- C
Last edited by Charles Poynton on Wed Aug 18, 2010 5:16 pm, edited 1 time in total.
Charles Poynton
Calibration Soldier
 
Posts: 27
Joined: Wed Apr 07, 2010 2:16 pm
Location: Toronto, Ontario, Canada

Re: Poynton’s Vector 3 - Gamma – 2.0, 2.2, or 2.4?

Postby Bob Walters » Mon Aug 02, 2010 6:32 am

FYI,
80 nt = 23.349 fL
Last edited by Bob Walters on Thu Aug 19, 2010 4:28 am, edited 1 time in total.
Bob Walters
Calibration Hero
 
Posts: 417
Joined: Tue Aug 05, 2008 10:46 am

Re: Poynton’s Vector 3 - Gamma – 2.0, 2.2, or 2.4?

Postby Glen Carter » Wed Aug 18, 2010 2:26 pm

Mr. Poynton, your contributions are greatly appreciated...... Thanks!

How much emphasis is being put into the actual black level performance ,of any given display, when considering gamma settings? My concern has been the elevated threshold of black with the newer displays (post CRT) and digital front projectors with low native contrast ratios.
If we have a front projector with low native Contrast, the actual black level can be higher than the 2.4 function calculation for 5%-10%...
Example: if we have 100%=50nt(14.6fL), then, with Gamma=2.4, 10%=.199nt(.058fL). If the 0% of the projector, on the screen, is .15nit, then, IMO, we need to just our gamma calculations to represent the range of .15nt to 100nt, where 10%=.349nt

This may be nit picking, however, understanding the environment and gamma are important. The "correct" gamma may be, what we are given, the best of a couple choices or custom corrected gamma. I feel, choosing the "correct" gamma really comes down to which looks best in the given environment to the trained eye.
Glen Carter
ISF Certified Calibrator
Home Theater Calibration
Glen Carter
New User
 
Posts: 1
Joined: Wed Aug 18, 2010 1:30 pm

Re: Poynton’s Vector 3 - Gamma – 2.0, 2.2, or 2.4?

Postby Charles Poynton » Thu Nov 17, 2011 8:51 pm

Glen -

You raise an important point. Some modern displays - LCDs with static backlights, for example - have higher off-state luminance than a properly adjusted CRT, so their performance isbn't as good in the dark. However, typical LCDs have a much lower diffuse faceplate reflectance than a CRT - say 1% compared to 6 or 10% - so LCD performance in a light room can be better than a CRT. I spent Saturday with the Sony BVM-E250 AMOLED (see Issue 11), and I remain convinced, it's the best of both worlds! (But at ~25 k$ for 24.5 inch diagonal.)

The rest of your question is really a research topic, the goal being to determine the "best" gamma curve for given display and viewing conditions. I have some ideas, they'll be in my PhD thesis at some point; meanwhile ...

I feel, choosing the "correct" gamma really comes down to which looks best in the given environment to the trained eye.


Seems reasonable to me assuming you have the experience!

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


Return to Poynton's Vector

Who is online

Users browsing this forum: No registered users and 1 guest

cron

x