Film-Tech Forum ARCHIVE: Preferred screen brightness

This is topic Preferred screen brightness in forum Digital Cinema Forum at Film-Tech Forum ARCHIVE.

To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=16;t=003645

Posted by Kevin Fairchild (Member # 4985) on 04-05-2019, 11:38 AM:

I know 14fL is spec, but I'm curious what brightness each of you think looks best.

Posted by Harold Hallikainen (Member # 5405) on 04-05-2019, 12:42 PM:

Since the studio mastered it for playback at 14 fL (48 cd/m^2), supposedly that would look best. Lower levels can cause darker scenes or portions of scenes to become invisible. Higher levels can make flicker more visible. Both higher and lower can affect the perceived color.

Harold

Posted by Steve Guttag (Member # 268) on 04-05-2019, 04:07 PM:

+1 on that.

Posted by Marco Giustini (Member # 4544) on 04-05-2019, 05:38 PM:

+1 too with a *

* - as long as 14fL is measured on an AVERAGE on screen. I found that on a high gain screen a centre value of around 23fL gives an average of 14fL (measured on 9 positions).
Way too many times just the hotspot is measured, giving a dull picture on those high gain(even worse: silver) screens mounted on flat frames on short throw auditoria.

Posted by Steve Guttag (Member # 268) on 04-05-2019, 10:45 PM:

Well yeah. You have to follow the whole standard and not just cherry pick the convenient one.

Posted by Marcel Birgelen (Member # 6801) on 04-06-2019, 01:39 AM:

Something that "looks best" is a very subjective thing. I personally might consider a particular movie to be overly dark, unfitting to the tone of the movie, so I may think it "looks best" with a bit more brightness...

But, as a tech, a projectionist or exhibitor, that's not really my task. My task would be to present it at the brightness it was intended to be screened. So, like others put it, you should present it at 14fl.

Posted by Kevin Fairchild (Member # 4985) on 04-06-2019, 04:19 AM:

Thanks everyone.

quote: Marcel Birgelen
But, as a tech, a projectionist or exhibitor, that's not really my task. My task would be to present it at the brightness it was intended to be screened. So, like others put it, you should present it at 14fl

I understand what you’re saying but I don’t entirely agree. In the case of audio, if I were to present the feature at the intended 7.0 volume, I would be bombarded with complaints that the volume is too loud. As an exhibitor, I believe your loyalty should be towards the customers’ preference, not the creator’s.

I’d like to edit my original question:
If you had your own screening room for your own personal viewing, where would you set your brightness to? Personally, I like the colors to be slightly on the vivid side so I tend to turn it up a little.

Posted by Marcel Birgelen (Member # 6801) on 04-06-2019, 08:15 AM:

We do have our own screening room. Still, I'm not able to answer your question, since it's highly dependent on the movie itself. I guess that for most movies, the overal brightness is just about fine. So, generally, I don't mess with the brightness, unless it's getting into such extremes, that it starts to irritate.

I guess it's the same with the sound. I try to play it at "7.0", but some mixes are so hot, they get turned down a notch, because they're simply unbearable at 7.0. Maybe the director intended to blow my eardrums, but that's where I draw a line.

Some mixes, I'd also consider defective. I have helped them a hand in the past, I must admit... (Interstellar, I'm looking at you now)

Still, the general rule that you present the stuff it should be presented should apply. Customer complaints can be factored in, but I personally won't factor in the singular customer complaint, I'd rather try to educate them.

I did the same back when I worked at a cinema, back when we still had actual film to run [Wink]

. If individual customers came by complaining the movie was too loud, too silent, too dark, too light, too black & white, I'd tell them that we always aim for a presentation the way it was intended to be presented and that, in this case, it was the director's choice it worked out that way. If we had hordes complaining about something being too loud or the other way around, then we generally tried to "fix" it, although even then, we generally tried to inform the public that we turned the volume down or up before the show started.

Posted by Daniel Schulz (Member # 1881) on 04-06-2019, 11:21 AM:

I like 14fL for regular movies that were graded for 14FL, but I've become very fond of the 21fL that IMAX has settled on. Their new projectors have enough contrast ratio to look really good at that light level, as they can still go dark when the material calls for it. As projectors get better and better I find myself wishing for SMPTE/DCI to revisit the luminescence spec.

Posted by Marco Giustini (Member # 4544) on 04-06-2019, 04:36 PM:

If the projector has enough contrast ratio and it was my personal screening room, I’d set it to 20fL.

Posted by Carsten Kurz (Member # 5396) on 04-06-2019, 04:48 PM:

Yup. You don't like an overly bright setting on e.g. an S2K machine, but with a better contrast, going higher doesn't hurt.

- Carsten

Posted by Marco Giustini (Member # 4544) on 04-06-2019, 09:06 PM:

quote: Steve Guttag
Well yeah. You have to follow the whole standard and not just cherry pick the convenient one.

In my 20 years in the cinema industry I’ve never encountered anybody following that part of the standard.

Posted by Frederick Lanoy (Member # 5416) on 04-07-2019, 02:53 AM:

One day on a Barco DP4K40LHC (Laser High Contrast), I did a mistake and forgot to assign the correct laser dimming to Flat. So I used the same value for Flat and Scope which means Flat was too bright as it was a scope screen.

The movie was "Shape of water" (1.85) and during my test, I notice how good the image looked. I was actually fooled by the High Contrast on the projector (both native and ANSI). When I measured once again the brightness on Flat corrected white (XYZ), I was close to 22 ft/l and it looked great on this specific example. I was almost sad to dim to reach 14 ft/l.

Otherwise, target is 14 ft/l unless the content is graded for higher value. In France, we have Eclair Color that needs 30 ft/l. It can look good if the content is good.

Posted by Marco Giustini (Member # 4544) on 04-07-2019, 05:34 AM:

Frederick

Was that shown on a high gain/silver screen?

Posted by Harold Hallikainen (Member # 5405) on 04-08-2019, 01:04 PM:

SMPTE 431-1:2006 specifies luminance and chromaticity tolerances for theaters and review rooms. The reference value for the center of the screen is 48 cd/M^2 or 14 fL. For theaters, the tolerance is +/- 10.2 cd/M^2 or +- 3 fL.

The reference value for the sides of the screen is 85% of center with a theater allowed 75 to to 90% of center. It's interesting that 100% of center is not allowed. There was considerable discussion of this at ISDCF regarding emissive screens which, I expect, would be pretty even. Further, if headroom is available, it should be possible with digital projectors to adjust the luminance of the sides up to 100% (or adjust the center down so it is the same as the sides). There was considerable discussion as to why the 85% value was there. Is it because that's what projectors were capable of, or is it indeed desired that the sides be dimmer than the center. There was discussion of a visual effect where the side suddenly drops to black (masking), and that a slightly dimmer side was desirable. I do not remember the name of this visual effect. For emissive displayes, it was finally decided that the radiation pattern of the LEDs (beamwidth) resulted in the sides being a bit dimmer than the center, though I have not seen measurements showing that they meet this specification. Further, if you were to measure from a non-center location (such as the left end of a row 2/3 back), the side of the image would be brighter than the center at this location. SMPTE is working on a recommended practice for screen luminance measurement that includes the center of the seating area, the left and right sides of the center, and the left and right sides of the front row. If the theater has a balcony, additional measurements are in the balcony. The goal is documenting the luminance at the ideal location (center of audience) and at several non-ideal locations. How is the image quality at these locations? Should they be within the tolerances specified by ST 431-1?

On higher luminance, note that the draft at https://www.dcimovies.com/drafts/DCI-DRAFT-HDR-D-Cinema-Addendum_v09_2018-1116.pdf for high dynamic range puts peak white at 500 cd/M^2, or about 10 times the current peak white.

https://www.dcimovies.com/drafts/DCI-DRAFT-Direct-View-Display-D-Cinema-Addendum_v09_2018-1116.pdf is a draft document on emissive screens. They are to support standard dynamic range (48 cd/m^2) and may optionally support HDR (500 cd/m^2). There is considerable discussion of off-axis performance.

Finally, DCI release https://www.dcimovies.com/announcements/DCI-Technology-Evolution-Statement_2019-0329.pdf which describes the two above linked documents and also tells about viewer preferences.

Harold

Posted by Steve Guttag (Member # 268) on 04-08-2019, 01:53 PM:

Whereas this was a discussion in an SMPTE meeting some years ago, I'm not able to present that information here.

Even with Film projectors of yesteryear and decent lenses, and a matte-white screen, it was certainly possible to exceed the Standard.

I also never favored lumping corners in with sides...they are completely different in terms of lighting.

Posted by Emerson Caitano (Member # 9759) on 04-08-2019, 02:40 PM:

I prefer to work with 21 FT, similar to IMAX

Posted by Harold Hallikainen (Member # 5405) on 04-08-2019, 02:45 PM:

The SMPTE standard does not "lump in corners with side." There is a separate specification for corners and sides. However, for theaters, the tolerance on corners is open. It is specified for sides.

On preferring higher luminance than the movie was created for, it seems like it may be preferable to some individuals, but not the movie creator. The movie creator may have wanted to do 21 fL for peak, but there's no standard for that (but I guess IMAX uses that throughout). The HDR draft from DCI allows for a peak white of over 140 fL though it is not necessary to use that. It's up to the creator.

Harold

Posted by Steve Guttag (Member # 268) on 04-08-2019, 04:37 PM:

You're getting into semantics. For review rooms it states "sides and corners" with a single percentage reading applied to both. Though it is "at least" type number (can be higher) the realities are, if the corners are included, that is where the darkest area will be, presuming even light distribution.

I'd have to check but I think the Theatre Luminance limits used to be similarly worded as the Review Room Luminance limits, if one went back far enough.

Posted by Bobby Henderson (Member # 840) on 04-08-2019, 08:50 PM:

How does foot lambert levels convert over into nits? The reason I ask is because of LED boards and how their brightness levels are rated. For example, Samsung's new LED signage & cinema screen products list one or two nit levels in the product specs. The boards for The Wall claim nit levels of 1600 (peak) and 500 (max). The boards in the Onyx product line are 400 nits. Both products are for indoor-only use. FYI, outdoor LED boards must be much brighter to compete with ambient sunlight. The Daktronics LED displays our company sells can have max brightness levels up to 12,000 nits.

Posted by Harold Hallikainen (Member # 5405) on 04-08-2019, 09:04 PM:

A nit is 1 candella per square meter. 1 fL is 3.4262591 nit. Multiply fL by 3.4262591 to get nit.

Harold

Posted by Marcel Birgelen (Member # 6801) on 04-09-2019, 05:39 AM:

As a reference, Dolby Vision content can be mastered to up to 10k nits per pixel, there is currently no projection technology that comes close to this though. Some top of the bill OLED screens are seemingly capable of producing up to 8k nits, but I'm sure you're going to burn them out rather quickly if you keep pushing them at those extremes.

Posted by Steve Guttag (Member # 268) on 04-09-2019, 08:09 AM:

Note:

Checking some documents...S196 (film) lumps sides/corners as a single specification in review rooms and specifies just sides for theatres.

431-1 (D-Cinema) does indeed separate them out in separate categories but to the same effect (same specification for both in review rooms and "not specified" for Theater when it comes to corners.

In short, the result is very similar but I think the 431-1 is more clear in its presentation (chart). S196 is more brief in its presentation (just text).

S196 doesn't put an upper limit on sides/corners like 431-1 does.

Posted by Harold Hallikainen (Member # 5405) on 04-09-2019, 11:46 AM:

Thanks for checking on the standards! Also, it's interesting that for film, "peak white" is 16 fL since it is "gate open" with no film. It is assumed you lose about 2 fL due reflection and attenuation by clear film.

Harold

Posted by Steve Guttag (Member # 268) on 04-09-2019, 01:24 PM:

It was more sophisticated than that. The goal was to simulate lowest density of processed film. If you were to photograph the sun, lets say and process that...what would the throughput be. I can't say for sure if or who did this but I know that was the goal in matching the two. Unfortunately a great resource from those days is no longer with us (John Pytlak).