| Forums | Polls | Register | Reply | Search | Statistics |
Display Calibration Light Illusion Forums / Display Calibration /   
 

Feature Request: Intelligent Display Profiling

 
Author simonjaquemet
ZRO
#1 | Posted: 6 Oct 2020 19:26 
Hi all, hi Steve

We have been long time Lightspace customers. We didn't use it much for full display calibration recently as we mainly work with DCI projectors, which are very spot on and only require to sometimes measure the primaries, adjust and verify.

But now we're trying to calibrate a ASUS PA32UCX-PK.

We are using Resolve as a patch generator. The Lightspace laptop is connected via wired ethernet. Probe is cr100
Still I'd say in around 60% of cases we get bad readings during profiling. This could be due to network issues, some background process - we don't know.
Quick profiles mostly work. Longer ones always have some bad patches, which destroys the whole calibration.
The result is: we are spending days to find out a viable workflow and settings to profile the display and troubleshoot what could be the reason for the patch generation sometimes failing or losing sync.

Wouldn't this be a quite simple thing to solve in software, to make the whole profiling process much easier and save countless hours?

With my limited software development experience I imagine it not too difficult to detect bad patches, which show extreme delta error values and just re-read them.
Possibly adding some extra delay while re-reading which would solve the tedious process to find the right but shortest possible extra delay time - especially with a dynamic backlight display.
Or to add a filter, which removes readings above a definable delta error threshold from profiles.

If Lightspace already has this functionality somewhere hidden, let me know. Or if it is planned to be implemented in ColourSpace, would be a reasong to upgrade.

Author Steve

INF
Male
#2 | Posted: 6 Oct 2020 19:57 
The is almost definitely a Resolve or network issue.
Easy to prove, but using the built-in TPG in the ASUS, or direct HDMI from the LightSpace laptop.
However, it is not a simple thing to 'develop out', or manage in software, as there is no feedback loop from the TPG (Resolve) to let LightSpace know when it has failed.

But, with ColourSpace you can just select the patch that failed, and re-measure it.
Unfortunately, that is not something we can retrospectively add to LightSpace.
You can see this capability in the Display Profiling video guide on the website: https://www.lightillusion.com/video_guides.html
And the advanced graphs in ColourSpace make it very easy to see, and select, the patches to re-measure.

Steve
Steve Shaw
Mob Boss at Light Illusion

Author stuart_p
ZRO
#3 | Posted: 8 Oct 2020 23:16 
I've had this very same issue and I'm not so sure about it being totally a Resolve or network issue. I've had 17^3 profiles where you get 20-30 bad reads in say about the 4,000 patch mark where every 2nd patch is wrong. For instance patch 4001 is OK, 4002 bad, 4003 OK, 4004 bad etc. It's always the luminance that is wrong, xy seems to be OK but luminance reads basically 25% of what it should. My feeling it is to do with the CR100 and sync frequency. I generally shy away from the auto sync and use the frame rate of the patch generator i.e 24fps or 25fps. I also use delay 0.3sec as a minimum. I have profiled the PA32 OK with the CR100 in both SDR and HDR. Never had this issue with the CA210.
I haven't had this issue with Colourspace like I had with Lightspace using the CR100. The ability to reread bad patches is a big improvement, but I would like there to be a function where you can batch reread, shift select patches and reread. The point info in Colourspace was the way I could finally see what was happening as it shows target xyY against read xyY. That is incredibaly useful. Lightspace only shows you the read value. For seeing what is actually going on with issues Colourspace is really superior to Lightspace.

Author Steve

INF
Male
#4 | Posted: 8 Oct 2020 23:32 
I have the ASUS on my desk, and use the CR-100 all the time, using either the ASUS in-built TPG, or the ColourSpace patch window via extended desktop.
(Or a BoxIO.)
Never have any issues - ever.

Resolve is known to have issues, if the PC it is running on 'stutters'.
Usually due to interrupts, network issues, or CPU too busy.
It can even just 'delay' posting a patch, so the Extra Delay value in use becomes too short.
(We never use Resolve as a TPG...)
And with a good TPG you will not need any Extra Delay with ColourSpace.
(It is very different to LightSpace.)

As for bad readings, you could just copy the patch RGB values from the bad reads, and build a 'Preset' Patch list, and run that 'over' the original profile.
That would replace all the patches in the list automatically.
You can use the 'Tangent' lines to see which patches are out of the normal range, using the 3D graphs to see xyY errors.

Steve
Steve Shaw
Mob Boss at Light Illusion

Author stuart_p
ZRO
#5 | Posted: 9 Oct 2020 00:31 
I agree with not using Resolve on a network, I gave up that a long time ago. But I found manually setting the sync to tpg frequency certainly improves things a lot, especially with RGB Oleds like PVM or FSI DM, but it can vary from display tech to display tech. I also agree with the HDMI out, so long as you know you your vcgt are null and you have signal range correct.
Interesting point about extra delay, I will give it a good test.

I think I like the Shift Select, reread much better

Display Calibration Light Illusion Forums / Display Calibration /
 Feature Request: Intelligent Display Profiling

Your Reply Click this icon to move up to the quoted message

 

 (?)
Only registered users are allowed to post here. Please, enter your username/password details when posting a message, or register first.

 

 
Online now: Guests - 2
Members - 0
Max. ever online: 264 [6 Jun 2020 12:02]
Guests - 264 / Members - 0