Jump to content

Rec. 709

ISF Calibration Photos

Recommended Posts

Moderator
23 hours ago, Tweaky said:

If you go back to page 6 of this thread and read my post from August 12, which is basically the same question, but a bit more technical, you can read REC 709's reply directly after it on August 16

if you look at Tonys posts above will see the HDR10 uhd test patterns from ray masciola, i too use these to setup my jvc for HDR nd have the same uhd disc it comes on. you can also down load the patterns for a charge from ray's site.

Share this post


Link to post
Share on other sites

Actually, since I've recently bought a 4K HDR TV [which still hasn't been delivered :( ] , I've been reading up on the calibration process at Rec 2020 reference gamut.

To say there is confusion about the process, and about the standard itself, would be a understatement.

And I'm not talking about just end users, I'm taking about seasoned ISF techs, and the guys that write the calibration software.

This basically all boils down to the High range of calibration in the 70% - 100%, where the HDR High Brightness starts kicking in.

 

Seems the start of the problem is the lack of standardization of just how many nits any particular movie is mastered to, be it for commercial release, or for transfer to UHD Bluray disc, some are mastered to 1000 nits, others 4000, and reference White for the HDR-10 spec is 10,000 nits.

No consumer visual playback device can output 4000, nits, and there is no display on the planet that can do 10,000 nit's, so the whole spec is ridiculous.

So what displays are doing to tame this brightness level so whatever the UHDR, is basically the same as a Iris of a projector, by stopping down to gradually reducing this high brightness to a reproducible level, this is being called Tone Mapping, and how hard it kicks in, at what range, for how long is all up to the individual manufacturer, th's why some UHDR TV's appear brighter than others.

This Tone Mapping switching in and out also effects the way any meter used for calibration reads the scene, and is compounded by the calibration software not knowing the ultimate brightness [Nit's ] level the display it is trying to calibrate, can go.

 

It seems they are working on it, or at least the guys writing the calibration software are discussing what can be done to at least help lessen the problem.

You can read what's going on at the AVS forum > Display Calibration > The Offical ChromaPure 3 thread.

 

To put the problem into some perspective, here is a post [#1244] I've copied and pasted from that thread, by Tom Huffman on the 19th Oct, who writes the software for ChromaPure calibration software.

 

Let me tell you a story.

When HDR was first introduced and I got the formula working, I realized right away that I really don't know how to implement it. The problem is that--unlike other gamma standards that define output at every video level by reference to 100% white, which is unspecified--HDR10 specified 100% white as 10,000 nits. No display can come even close to this. The OLEDs don't even reach 1/10 of this. Projectors are doing well if they reach 1% of this! As I had done before I reached out to a few industry insiders asking how this seemingly impossible standard was supposed to be implemented in the real world. I didn't receive any answer to my queries. This was strange. I had asked questions like this in the past and always got some input, but this time nothing. The only advice I received was to calibrate the best you could and just clip signals above the display's capabilities, which in most cases is about 70% video. This is not an ideal solution, but it worked.

As time went on I begin to hear a lot about tone mapping. In short, this is an attempt to bend the PQ curve to minimize clipping. The thing is, there is no standard for tone mapping. Its implementation is left up to every vendor and manufacturer. There is another phrase for tone mapping. It is called "Making sh*t up." It then occurred to me that the reason no one had answered my earlier query is that NO ONE KNEW.

After speaking to some industry folks recently, I decided to experiment with tone mapping. Based on one suggestion, I tried a flat 2.4 power law gamma. This looked shockingly good. It is counterintuitive at first glance. I mean what is the point of high dynamic range if you are not using high dynamic range? However, if you look at what HDR is doing, almost all of the increase in dynamic range is at the high end. There is some increase at the low end, but it is fairly small. By using a power law gamma you now cover the entire luminance range of the display without clipping. Since the display is in HDR mode you can take advantage of whatever high-end luminance that the display offers, so you increase the dynamic range by 500-600%. The only downside I see is that the low end is considerably elevated from what the PQ curve suggests. For example, the PQ curve specifies 0.32 nits output @ 10% video. A 2.4 power law gamma specifies 2.39 nits @ 10% video assuming 100% video is 600 nits. I am also going to try a hybrid gamma of PQ up to 50% video and then 2.4 power law above that. Since there is no standard, I am free to try literally anything. I'll use whatever looks the best by eyeballing, and I'll offer several options to users to pick whatever they think looks best.

This will all be released in the next version along with auto-cal for the Radiance Pro.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


  • Recently Browsing   0 members

    No registered users viewing this page.

×