What level of cd/m2 ( nits ) do you calibrate the monitor to for Premiere Pro using rec. 709 ?

[Letty2019]

I'm confused with all the info out there on the net, thought I'd try here:

Calibrating my LED monitor (w/spyderXelite)  to:

rec. 709 

gamma 2.4  (tv/film broadcast standard)

now.... for brightness....

100 cd/m2?  140? 150?

Watcha think?

Thanks!

Letty

ps:  When I set it to rec.709, it's only letting me do gamma 2.22 and the white point says  x0.313 y0.329  (not 6500K)  😞    is that about right?

[salvo34]

It's really hard to do what you want but it's good that you are trying it.

It might help people to help you if you say what monitor ( exactly ), what computer ( exactly ) and what your monitor it hooked up to ( discreet video card via DVI for example ?? ) and exact info on what card it is.... 

Even with that info it will be a slow process to do what you want cause those numbers are only pertaining to your particular 'outcome' for that particular monitor.

Anyway, the best thing is to use the color bars ( can generate within PPro ) and get them as close to real colors and tones as possible. Then broadcast and see if it's close to correct.

I have mine set to what you want ( eizo monitor LCD ) but also use a pro video monitor out of an SDI card. Ultimately I use the video monitor as the 'correct' monitor to do final work.. ( reference monitor ).

Why so complicated ? Cause monitors are all different and each will have some controls ( brightness, contrast, etc ). Most DON'T have good color controls.

Use the bars and see how close you can get without spyder. Then use spyder and let it do the stuff automatically ( do you have a usb cable from monitor to your desktop so you can save the color profile ? ).

[salvo34]

P.S.

If you have a public cable access broadcast station near you ( within 50 miles ? ) you can call and make an appointment to visit them and tell them you'd like to see what the engineer works with ( the actual broadcast part of the place ). If the engineer is nice and has time he'll give you a rough sketch of what is happening the the signal from ingest to output etc. You won't understand it all any more than I did but you'll get to see with your own eyes the basics of the technology.

good luck !

A post house would be good to visit too, but I doubt you'd be welcome to visit as they are usually busy as bees working.

[chrisw44157881]

Technically you'll want 100 cd/m^2 per spec
white point 0.3127, 3290 6500k
One could argue that nits is ambient dependant(which is somewhat true) but nits and gamma are related due to our eye's non-linear response curve), so you'd want a controlled grading environment for true quality control.
This is also why HDR at low nits doesn't really make any sense.

[R Neil Haugen · Accepted solution]

I'm a contributing author at MixingLight.com, a pro colorist's teaching website. So I am around dicsussions of color management issues at length every day over Premiere, Resolve, Avid, Mistika and other apps on various gear with exports from broadacast/OTT to web/socials to dark-room theatrical release.

The full Rec.709 stats suggest 100 nits for monitors while working in a semi-darkened room, and this is really the better option if you are also working with a 2.4 gamma.As noted in a comment above, the gamma and nits are "blended" by our eyes, so actually working 2.4/100 nits in a semi-darkened room then displaying in a brighter room works out just fine.

If you can't get your calibration software to work with a 2.4, you can work with the 2.2, but do so in a room with a bit more light. Not quite a full-on office light, but not too subdued either. You might also go to say 120 nits, but I wouldn't go above it.

And note, this is assuming you're more interested in web/socials than broadcast.

Neil

[chrisw44157881]

and ignoring perceptual quantizing for a moment, if anyone's curious, the nits relationship is the dynamic range ratio of blacks to white. 0.05 nits / 100 is 1:2000 contrast ratio. If you can keep that number constant, you'd have a relatively more accurate translation into 2.2 gamma 120 nits or whatever in between as a brighter room would raise the black level on your monitor's reflectance.

The whole idea behind all of this is matching the final output playback device's dynamic range because dynamic range shifts all perceptual sub-elements like saturation, local contrast, etc. 

if you wanted to check this with math:
Monitor luminance = Contrast × (Input Gamma) + Brightness
dynamic range = black/white(max nits)

I have maintained that variable gamma does exist and HLG has adopted it, but its formula is too complicated for calibrating monitors so as of right now so we are sticking to SMTPE standards. Although I have made a super-unofficial gamma curve in excel for rough approximiation of relative gamma.
https://www.liftgammagain.com/forum/index.php?attachments/gamma-jpg.5976/

PQ is useless in an uncontrollable environment as the gamma cannot be adjusted to match the ambient light unlike HLG which supports variable gamma.
In the future, most smart devices will require an ambient light sensor to adjust for raised blacks and washed out nit levels from living room windows.

[Letty2019]

Thanks again.

[Letty2019]

Thanks again,

[R Neil Haugen]

The comments from most colorists and the color managmement experts of the major calibrabrion software firms regarding HLG versus PQ are rather direct. As something to work within, HLG is seen as a plain mess in many ways.

So ... they see PQ as the only sensible working environment. HLG is acceptable in end-user device use sort of maybe ... but not in the working environment of those prepping the material.

It's a Wild Wild West out there now.

Neil

[chrisw44157881]

There's already heavy discussions to tame to PQ metadata and adjust gamma on lower end devices, so obviously they didn't think this through. PQ in direct sunlight with low nits output just ain't gonna work in the real world. And exactly who's living room is 5 nits? You'd fall and hit the coffee table, lol.

[Letty2019]

not sure why, when I reply to the post above you, it goes to your section??? 

[R Neil Haugen]

Because of the 'nesting' nature of this stupid software, when you reply to someone in a grayed-section, it adds to the bottom of the grayed section. I normally use the blue Reply button at the top of the thread to simply keep my posts in chronological order through the thread.

Neil

[salvo34]

I feel bad for Letty... she just wants a simple calibration to get close to rec 709 2.4.

All the responses are gonna be all over the map.. due to compexity, but the final decision is what it looks like when she exports to delivery. Could be major motion picture or episodic TV or web streaming, nobody knows and nobody can see her monitor live and visit her dark room etc.

She needs to put up the color bars and do it manually on her primary monitor to start.

Since PPro shoves everything into rec 709 in program monitor by default, it's impossible to say what she calibrates and what Adobe does with it.

Simple stuff...

look at color bars and do your best to adjust to what looks close to SMPTE.

And, yes, prettty soon nobody in this hobby world will care about the SMPTE standards.... cause you can't manage color in PPro no matter what you try.

[chrisw44157881]

as an addenum to the OP's initial question(which I believe was answered both technically and perceptually)

1. adobe has a free 2.4 to 1.96 gamma lut you can use for youtube delivery.

2. their DCP encoder takes 2.2 input videoo and auto coverts it to 2.6 XYZ DCP.

3. there is a new HLG timeline you can use and even export to PQ delivery.

color bars aren't going to fix anything once the file is played back on an external player. All you can hope for is to reach the largest percentage people possible for that delivery type. This is why I choose chome, as chrome is the #1 used browser and so 1.96 gamma has a higher probability of working than any other combination. As soon as you watch something on your iphone, TV, it's gonna look washed out. That's life until they create adaptive perceptual contrast / gamma in the end user devices.

[Letty2019]

Wow, great reading all these posts.  I think I understood about 43.5% of it all. 😉

If I was to do a DCP, then 2.2 gamma is the way to go overall?

Thank you all again.