Since there was so much debate on this topic and his quotes were being referenced in places, I decided to email Charles Poynton. He has given me permission to post our correspondence here. I started to ask him what the thought about "%IRE," but thought better of his time...
Me: Dr. Poynton,
First and foremost, I would like to thank you for making publicly available the work on your website. It has proven to be a valuable resource for me as I study video systems as a hobby. Thank you.
My name is Casey and I am a member of a popular internet forum (you may have heard of it), www.avsforum.com
. Your work is often cited in the more academic discussions that take place on AVS, and is widely regarded as authoritative. Recently, a debate has been started regarding blacker-than-black and whiter-than-white signals. It seems that respected forum members are in disagreement on the topic of video material being present above digital 235 and below digital 16 on commercially mastered DVD's. Often times these debates are solved by quoting your work, but this particular debate finds the two sides interpreting the same quotes from you in two different ways. Both factions also seem to interpret the Rec. 601 and Rec. 709 specifications differently.
I won't go into the particulars of the debate as I'm sure you are very busy and have better things to do, but I was wondering if you would be so kind as to clarify a question for us:Should commercially mastered DVD's contain actual video material (not noise) below 16 or above 235 if they have been mastered according to standards?
This question is of importance to us so that we might calibrate our displays correctly. I realize that this question is quite likely to be trivial for you, and I apologize in advance for taking your time with it. I would understand if you do not have the time to respond; however, any light you could shed on the subject would be greatly appreciated.
Dr. Poynton's first response:
C. Poynton: Casey -
>Should commercially mastered DVD's contain actual video material (not noise) below 16 or above 235 if they have been mastered according to standards?
Studio standards do not, at the moment, completely define what is supposed to happen outside the interface code range 16 through 235. Studio standards are in the process of being revised to incorporate elements that are necessary in the flat-panel world, and improved documentation of signal treatment in the footroom and headroom regions is likely to be part of the work.
In 1984, at the dawn of digital video, provisions were made for footroom and headroom for several reasons. The main reason was - and remains - the accommodation of filter undershoots and overshoots. To maximize image quality, undershoots and overshoots should be preserved through the signal chain as long as possible even if they are eventually clipped.
Generally, transient (short duration) excursions into the footroom and headroom regions are allowed, and can be expected in studio mastered material. Consider "transient" to encompass 4 or 5 samples in the footroom region.
During the last 10 years, the headroom region has been increasingly used to accommodate specular highlights. In the headroom region, specular highlights may be conveyed using interface codes above 235, and their durations can be longer than 5 samples - but should not last longer than about 16 samples.
Ideally, the display transfer function should follow its inverse gamma curve from reference white (interface 235) up to peak white (interface 254). Inverse gamma should be a 2.4-power for studio viewing conditions (100 nit display, 64 lux ambient). A substantially brighter display of a substantially brighter ambient calls for a lower value of gamma; for example, for typical desktop computer environments the 2.2 value of sRGB is appropriate.
Details - perhaps more than you want! - are available in "Digital Video and HDTV Algorithms and Interfaces".
Feel free to copy this note to .
Just to make absolutely sure I was interpreting his email correctly, I asked for some clarification.
Me: Dr. Poynton,
Thanks so much for taking time to respond. We were ultimately trying to figure out if our digital displays should be calibrated to show WTW data through 254, or if we should try to force the display to clip at 235 (while still preserving headroom for processing until we reach the display). From your comments regarding gamma and the fact that recent trends are to place highlights above 235, I suppose the answer to our question is that we should not force our displays to clip at 235, and instead ensure that our displays follow D65 and a proper gamma response through 254. Please correct me if that conclusion is incorrect.
Again, thanks so much for taking time to respond. I'll be sure to check out your book. As an engineer myself, I never consider too many details a bad thing.
C. Poynton: Casey -
>... ensure that our displays follow D65 and a proper gamma response through 254
That's my recommendation, for best picture quality.
Be aware that display vendors are, today, stuck quoting "brightness" (luminance, in cd.m^-2, or "nits") at REFERENCE white, but if they follow the true power law up to 254, that's 20% brighter but they can't quote it. In other words, if they clip at 235, they can effectively quote a higher brightness value on the spec sheet. My conclusion is that the CEA should establish a policy that permits them to quote the PEAK luminance, but that's not the case at the moment.
Now I know that some of you are going to have a hard time with this. I would kindly ask that you not email or call Dr. Poynton and call him an idiot, or ask him if he is related to Chris. Let's not try to bend his comments around as they are very specific.
Hopefully this will shed some light on the subject...