Originally Posted by DLCPhoto
I'm hoping this is a typo, since I thought you had indicated that curves with a lower target will produce more brightness, useful for those of us with large screens and/or low gain, but the cost is reduced saturation.
From your first post in the JVC Autocal thread:
This stuff is new, and therefore somewhat confusing initially, to most of us, and I want to make sure I've fully understood what you've explained.
It is a typo, thanks for spotting it, I am getting tired and it was late at night
Asharma mentioned a 200-1200-4000, and he should be getting a bit more saturation than with the new 2000-1100-4000. I though in my tired head that he was using a 1000 curve. I'll correct my reply to him.
Originally Posted by stanger89
I just had a radical thought, that I want to write down while I'm thinking of it.
Can an Integral strip the HDR metadata from an HDR signal? And I mean that in the correct sense (not the Oppo sense
) of removing the AVI info frame that specifies HDR, such that an HDR signal would be sent to the projector, but the projector wouldn't be forced into "HDR mode"?
If so, we could use our custom gamma, wouldn't have to reselect it ever time something changes, and the DI wouldn't be disabled!
Does that sound plausible?
I had the same radical thought, as it would be exciting indeed, but the problem is that the fury does nothing, it only tells the player that the projector can't do HDR, which forces the player to send SDR. this is how we get the SDR BT2020 from the Panny, forcing the palyer to send SDR BT2020 without downconverting to rec-709, which it does by default if you switch HDR off in the player.
So as long as the HD Fury sends the HDR flag off, there is no way to get the player to send HDR.
However, I was thinking of asking Ken if they could do something for us, so that the HD Fury would report HDR to the player but would not report HDR upstream. That would solve our problem.
This would be the opposite of the SDR BT2020 trick:
Instead of reporting SDR to the player and HDR to the PJ, we want to report HDR to the player and SDR to the PJ. It should be possible. I'll report back when I've been able to have a chat with him.
Originally Posted by krichter1
Well... I'll tell ya what buddy... Great work
... your 1000-4000 gamma looks excellent!
on my gear (WOWeee)! Cannot believe what a difference it makes already let alone if you say these new ones might look even better on a larger unity gain... thank you (again!), my friend!
I targeted for exactly 1000 nits on the Oppo (+7/+5 & JVC 0/0) and even used the HDR color clipping pattern to bring red & green up to ~1000 nits (surprised the 6500K red gain ended up at -12 & green @ -8 and the Oppo saturation at -2 but looks really good so far (still playing tho).
Now if we can just get those iRule IP strings for custom gamma to work we'd be in great shape to quickly switch gamma profiles (but I tested and its a JVC thing; not that the command strings are bad)!
Glad it works for you, but the others are correct, if you get a 4000 curve with a roll-off, there is no need to adjust the clipping point to 1000nits with the contrast setting, you are destroying the curve doing this.
The whole point of the 4000nits curve (with a roll off) is that they resolve up to 4000nits without a penalty in brightness for the overall picture. That's why I call them universal curves: you use one curve for all titles and you resolve up to 4000nits when necessary.
If you really want to clip lower, which I don't recommend anymore, use a curve that ends in 1000 or 1100, but you have to change curve everytime you play a title mastered above that, which is a real hassle for little to no benefit (and some artifacts).
My advice today is to use a 4000 curve and use only that one. when setting contrast, always set contrast to the last number in the curve name, otherwise you are going against the intent of the curve.
Originally Posted by ckgolf
I believe if you use the 1000-4000 gamma you need to clip to 4000 nits, not 1000.
Originally Posted by krichter1
Which one did you end up importing Steve and what was your Ct/Br results on the Oppo? Just curious since you have an even bigger screen than my 110" AT.
Also for Manni... I just noticed you added another yesterday which slipped by me (would you think this is better than the universal one)? . . .
02-14-17: 140-1100-4000 V2. This is a revised version of the curve above of the same name. It should be marginally brighter and more accururate. This is my default curve now. It works fine in low lamp (140nits here) and high lamp (200nits here).
I have posted three more since that one, you're behind
To clarify again, all curves ending with 4000 are universal curves. They are the only curves I recommend using, and you should set contrast to resolve up to 4000nits or slightly below. Unless you are far from that, I would not adjust contrast. I would NEVER adjust contrast to clip lower than that. These curves should need almost no adjustments in contrast on well configured, good players like the Panny. I have contrast at zero on the PJ and at -1 on the Panny for these curves.
Originally Posted by arve
I pushed some scripts that can load custom gamma curves to github. These scripts allow more control points than the autocal software uses.
Sounds great but the link doesn't work for me... Also, as someone else asked, how do you re-import the curve into the JVC? The internal tables have more control point (the JVC Autocal use up to 33) but the custom gamma files are formatted for 11 control points. Can your tool re-import the optimised files back into the PJ without using the JVC Autocal software to import the text file back in?
Originally Posted by lovingdvd
1. With my 140" wide, 2.37 AR 0.95 gain screen - when I am zoomed to fill the screen when watching a 2.40 4K movie with the WCG filter in place and iris fully open on high lamp from around mid to mid-long throw I get 13 ftL, if I recall. That works out to around 45 ftL, I think. Do I recall correctly that nits = ftL? So when you say I need 100 nits minimum to get benefit from HDR, you are saying that I need 100 ftL? If that's the case, then I need a little more than double the current max lumens I can get out of my unit?
Heck, I don't think even the new JVC laser gets me there, or if it does, then just barely?
2. As we've discussed, with some titles HDR does look nicer, even with my lumens issue. With that in mind, are you saying that I might prefer your new custom gamma over gamma D? If so I'll give it a try. Once you import a custom gamma, is it only active once you set the correction value to Imported? IOW, if I want the JVC to ignore the custom imported value (such as if I don't like the result), I just take it out of Imported for the correction value and then its like it doesn't exist?
3. Currently for HDR I use gamma D, with contrast at +12 and picture tone set accordingly (can't recall OTTOMH). Leave the brightness at 0 and set the panny settings as described where you posted the gamma import data?
The custom gamma curve should give better results, but you have less than 50nits available (that's 50cd/m2, around 14fL). There is NO POINT for you to use HDR. You are compressing everything and you would lose dynamic range and contrast playing HDR content in SDR on such a brightness deprived setup. You need at least 100nits (twice as much) to get some kind of benefit from HDR. Below that, use SDR BTR2020, get your DI back and be VERY happy. You're not missing out on anything
You need to read all the posts linked in the first post of the JVC Autocal thread on the subject, I really don't have the time to do more support on this, sorry. I'm sure others will help, all the usuals have caught up and are doing a brilliant work taking over (thanks to them by the way!).
Originally Posted by stevenjw
I'm confused as I thought the Import function for gamma in the JVC is expecting 11 points of R,G,B. How are you importing something else into a Custom gamma?
I have the same question. Their tools would have to deal with the import. The PJ supports more than 11 points but the import function expects files formatted with 11 control points, so I don't see how this could work unless the tools communicates directly with the PJ and loads the improved tables.
Frankly, I don't think we really need more than 11 control points, at least when working on gamma white exclusively, which is what I recommend to do for these curves, after having run an autocal and setting 100% white to D65 so that the RGB balance is already taken care of for the most part.
Originally Posted by Wookii
I just thought I'd take the time to say a big thank you for all your efforts in coming up with these new curves, they really have turned the corner for HDR viewing on these projectors.
I ran an autocal last night and imported your 140-1100-4000 V2 curve and it is excellent, a massive step up over gamma D. On/off CR in my set-up went up from 23k:1 to 55k:1 versus standard gamma D settings.
If it matters in terms of a data point, I have a 110" wide (120" diagonal) 2.35 HP 2.4 screen. I'm getting around 55FtL (188 nits) in high lamp iris open with the filter engaged.
My next step is to try and follow your pointers and use Calman to create a custom curve, as I'd just like to increase brightness in the mid-tones slightly over your V2 curve, it just looks a smidge too dark in my set-up, probably because I'm not quiet up to the 200 nits you can achieve.
One question I had though (and apologies if I have missed this, this thread has been moving fast in the last few days); once you have created your custom gamma curve in Calman by applying the screen offset. Are you able to push on with Calman and use that gamma curve to complete a Calman Autocal greyscale calibration and/or create a 3D LUT? I appreciate you may not have physically done this, as you may not have 4K HDR content on your HTPC, but I just wondered if this was possible?
If it is, it will enable me to complete a 3D LUT for my Radiance Pro. The final step for me will then be see if I can get the Radiance to output the signal with the HDR flag removed so that it stops the PJ switching to gamma D, and more importantly allows me to engage the DI on HDR content.
Glad it works for you too
As far as I know, Calman doesn't support the new JVCs for greyscale autocal, but it's not necessary, the JVC Autocal does a much better job at that with a reasonably accurate Spyder. Calman doesn't support the shaping LUT of the Radiance Pro yet, you would have to check with Jim and the Spectracal team. I don't think it can yet, but I'm not following this closely. Yes the main advantage at this stage for us would be to be able to re-enable the DI. We're working on this...
Originally Posted by LTD_666
Manni, first, I would like to congratulate you about your great work making gamma curves that permit viewing HDR contents almost directly without any modifying. I've was being testing it and I've viewed they are great. Much better than others based in gamma D that I tried before. Anyway I would like to build my own specific gamma. The most difficult for me is not the calibration process using Autocal. The most difficult is,before to start a calibration, to generate a reference gamma to compare with. I would appreciate if, when you feel yourself no so tired and you have time enough, you explain us the detailed steps to build a reference gamma. That is, how to introduce parameters as are the target and content clipping, roll-off value ... . I am not sure the specific steps with Calman (I think it is the software you are currently using) although, really, I would prefer another tool because Calman license is too expensive for me. The software that I would prefer to use is Chromapure or much better HCFR.
Thanks in advance and greetings from Spain
You're welcome, I'm glad this works for you too.
I have explained all this in a post recently. This post is linked in the first post of the JVC Autocal thread. Please read all of the HDR/Gamma section in that post, and all the posts it links to. I can't provide support on this, but many others in this thread will do.