or Connect
AVS › AVS Forum › Display Devices › Display Calibration › HCFR - Open source projector and display calibration software
New Posts  All Forums:Forum Nav:

HCFR - Open source projector and display calibration software - Page 78

post #2311 of 3446
Thread Starter 
Menu buttons aren't in english in the screencap above so something is screwy with your set-up. Try reinstalling the program or just delete the ColorHCFR.ini file in the installation directory and choose english when you start it up again. I'm working on updating the other region files but can't debug any of this until next week.
post #2312 of 3446
OK! I see.
I de-instal the program, cclean the registesr base with Ccleaner, reboot the computer and re_instal HCFR.
Buttons are still in French and Sat still wrong.
I delete the ini file and restart HCFR, same results.
I will try to delete all older occurences of HCFR in other directory .
post #2313 of 3446
Over the weekend I touched up the calibration of my TV. When I first calibrated last year the notebook computer I was using did not have an HDMI port. This year my notebook computer does have a HDMI port. So I took measurements using both the view images within ColorHCFR and also using the AVSDH 709 calibration disk in my PS3. I received significantly different results:

Using AVSHD 709



Using ColorHCFR Images


The big difference is the blue level. I do not know if this is being cause by the setup on my computer, if there are issues with the ColorHCFR images or what. Can someone suggest further testing that might resolve the issue?
post #2314 of 3446
Thread Starter 
Quote:
Originally Posted by Kerlucun View Post

OK! I see.
I de-instal the program, cclean the registesr base with Ccleaner, reboot the computer and re_instal HCFR.
Buttons are still in French and Sat still wrong.
I delete the ini file and restart HCFR, same results.
I will try to delete all older occurences of HCFR in other directory .

nevermind, the language used on pop-up buttons is determined by your locale setting for the PC so that will always be French. I will have to debug this later when I have a meter available because I'm out of town.
post #2315 of 3446
Thread Starter 
Quote:
Originally Posted by dschlic1 View Post


The big difference is the blue level. I do not know if this is being cause by the setup on my computer, if there are issues with the ColorHCFR images or what. Can someone suggest further testing that might resolve the issue?

You will only get the same (DVD vs. computer generated) results if your video card does not touch the RGB levels, meaning it has to have linear gamma tables loaded, no ICC color profile etc. It looks like yours is set to produce a color temperature shift to blue.
post #2316 of 3446
Quote:
Originally Posted by zoyd View Post

No, that's probably because your display (what model?) can't produce black lower than 0.05 cd/m^2.

Also, at what point do you play with the gamma setting on the tv? with that calibration it was default 0. Though it says it's only for middle level luminance but I noticed the blacks became darker on The Dark Knight png screencaps I used for quick checks. Before viewing Blu-Ray's.
Quote:
Originally Posted by zoyd View Post

For the 10pt control to work properly make sure you have contrast above 90 (95 would be good).

Ok, thanks. I'll give it another try tonight. Though from what I've read too high a contrast hurts blacks? default Movie contrast on this tv is 95. With that calibration it was 78 for night viewing with a dimmer uplighter.
post #2317 of 3446
Thread Starter 
Quote:
Originally Posted by xvfx View Post

Also, at what point do you play with the gamma setting on the tv? with that calibration it was default 0. Though it says it's only for middle level luminance but I noticed the blacks became darker on The Dark Knight png screencaps I used for quick checks. Before viewing Blu-Ray's.
Ok, thanks. I'll give it another try tonight. Though from what I've read too high a contrast hurts blacks? default Movie contrast on this tv is 95. With that calibration it was 78 for night viewing with a dimmer uplighter.

set gamma to give you close to the gamma target you want (2.2, 2.3, etc.) and then fine-tune with 10pt control.

no, contrast sets your peak white, brightness sets your black level and you should always do that first using a clipping pattern. check out some of the other threads for general calibration questions and help with your specific display.
post #2318 of 3446
Thread Starter 
@Kerlucun

Did you run a primaries+white @100% sequence first?
post #2319 of 3446
Quote:
Originally Posted by zoyd View Post

set gamma to give you close to the gamma target you want (2.2, 2.3, etc.) and then fine-tune with 10pt control.

no, contrast sets your peak white, brightness sets your black level and you should always do that first using a clipping pattern. check out some of the other threads for general calibration questions and help with your specific display.

It seems 2.4 looks best. I'm still seeing the same black detail but the blacks are a lot deeper. I used to think the tunnel scene in The Dark Knight Rises on the Bat-Pod 45 minutes in always looked grungy, green-y/blue like it was crushed on the concrete when the guy runs away from the bike. Now this looks good.

I think this is the best I could dial in the 10pt for ITU-R BT.1886. I can't get exact accuracy since the tv can't fully saturate the RGB points.

The black isn't perfect but it's deeper than before. Tron Legacy blacks are coming out from scene to scene like charcoal black and some oily blacks. Dependant on what it is. I also can't get over how glowy the film is now. It always felt a bit... cloudy but not too cloudy and lacked contrast.

0gdg.jpg
bh0s.jpg
f367.jpg
yc3s.jpg
djhq.jpg
u69s.jpg

_______________



Couple of days ago once I started, never took a screenshot at the time but this was roughly how the colour space looked on this tv.

Before doing this, for the past 10 months films always felt rather saturated in movie mode. The blue bleed was terrible through Underworld Awakening that I had to set the B Gain to 0 to try and rectify it. The colour space was so bad. Auto wasn't any better but it was quite faint and Native is like florescent blue. Native's blue is heading towards neon purplish blue.

ood3.jpg

_______________


For so long now, I'm no longer seeing this saturation bleed that it feels weird but great and because of that the viewing angle has changed greatly. In the past if I wasn't looking directly, a little bit off and you'd see a heavy blue filter through the darker colours as well as cloudy artifact blacks. The blue bleed was also terrible. It was so heavy you would swear one was wearing shades but clear on the one direct focal point like looking through a hole.

In Tron Legacy and such where Sam is getting kitted out, all over the dark grey-ish corners of the floor you could see this sea of blue saturation that appeared to be dancing or glowing like dark blue anti-freeze. It was partly rectified by setting the B Gain to 0 months ago via the AVS 709 patterns doing it by eye but you could still see some of it faintly. Now I don't see it and it's a nice mix of dark matt/shiny parts on that floor.

Most scenes with mahogany wood such as GhostBusters 2, the wood almost appeared glowing that same sort of dancing pixel effect like noise. Same with Big Trouble in Little China in most of the dark scenes since it was terribly saturated in the reds.

In The Dark Knight Rises when Catwoman was kneeling down working away at the safe, the yellow was terribly saturated and was a lot worse when she started beating up the guys in the yellow/orange lit room. Trying to rectify it by eye in the past was impossible. All that bleed has gone. It was so bad it originally looked like dark glowing orange/red around the fingers and heel.

Days ago at first I thought these colours are starting to look messed up as the points were coming into line since I was so used to seeing heavily saturated patterns. Well, it appeared normal to the eye. The films no longer had this heavy bleed in darker scenes and because of that the gains came up a fair bit. Plus the richer colours no longer look neon. Some football parks at night the grass was nearly glowing like kryptonite.

It's strange how the colours seem to be matt like but looking so natural since I was used to the saturation bleed. Only some scenes like Catwomans suit the blacks look inky, black leather shiny seats and so on. Then when you see the Sky Sports HD channels how the colours pop as well as how lovely the greys and white score board dialogs pop. I can't get over how white the dialogs have become. Doing it by eye in the past with on screen disc patterns always looked a little too blue, then it became a little too red or a little too green. It was impossible to balance it all out to find this perfectly bright white that I'm currently viewing.

Another thing I've also noticed about the set, a long time ago in some fast action fighting scenes normally the dark scenes I could see a lot of ghosting. Either had to turn on motion plus and such to try and rectify it. It wasn't fully cured but then I had to put up with the soap opera effect to some degree via the custom settings. Now I'm seeing fast fluid action without any of the motion features.

Feels like for so long I've been viewing the screen behind nasty filters. Like someone mentioned in another thread here walking about with blue filters on for a while then taking them off.
Edited by xvfx - 9/23/13 at 8:19pm
post #2320 of 3446
Quote:
Originally Posted by zoyd View Post

@Kerlucun

Did you run a primaries+white @100% sequence first?

Yes I do.
post #2321 of 3446
Thread Starter 
Ok, I'll have to see if I can reproduce the problem later, with the simulated probe everything looks fine. In the meantime, can you test this using French language. I want to see if the new generator functions work as well a new feature. Unzip this to the same directory as your current ColorHCFR.exe file. The new feature calculates dE and CIE chart target location of saturation runs based on your measured gamma (average from the grayscale run) or user specified gamma entered into the average gamma box on the references dialog.

And just a wild guess but try using Rec 709 primaries instead of PAL/SECAM to see if there is any difference in the saturation run pattern generator.
Edited by zoyd - 9/24/13 at 3:45am
post #2322 of 3446
Quote:
You will only get the same (DVD vs. computer generated) results if your video card does not touch the RGB levels, meaning it has to have linear gamma tables loaded, no ICC color profile etc. It looks like yours is set to produce a color temperature shift to blue.

As far as I can tell the computer does not have a profile loaded. How do I determine what the gamma table is?
post #2323 of 3446
Please find the test .chc file.
The French Sub-menu are correct but sat are still bad.

HCFR_Test.zip 1k .zip file
post #2324 of 3446
Thread Starter 
Quote:
Originally Posted by Kerlucun View Post

Please find the test .chc file.
The French Sub-menu are correct but sat are still bad.

Thanks, I think I found the problem. try this one.
post #2325 of 3446
@zoyd, did you remember the bug I reported a while ago, namely that I have to re-confirm madVR (open dialog and confirm without changing anything) as test pattern generator every time I start HCFR? Don't have time atm to test whether it's fixed in the newer builds.
post #2326 of 3446
Thread Starter 
yes, fixed that one. double clicking .chc file still doesn't work though.
post #2327 of 3446
Quote:
Originally Posted by zoyd View Post

Thanks, I think I found the problem. try this one.

Well Done, Sat works perfectly.

HCFR_Test2.zip 2k .zip file


Is it a way to find the ColorChecker values?
A view like 'Primaries and secondaries' will be helpfull.

Thanks very much.
post #2328 of 3446
Thread Starter 
Yes, that view exists in the drop down menu after magenta.

edit: oops, not in french version, only english. I'll fix it shortly.
post #2329 of 3446
Thread Starter 
Here is a 3.0.4.5 release candidate for folks to test (drop files in current HCFR directory). Alt languages should have all menu selections available and this version adds a "use measured gamma" option to the references prop page. Selecting this (default behavior) and HCFR will use your actual measured average gamma to calculate saturation targets for reporting dE and saturation shifts. If you deselect it will use the gamma entered into the average gamma text box above the button. That way you can see the effects of different gamma setting on the location of your target points in the CIE chart. Also added a saturation sweep called AXIS which expects an RGB sequence of 8 patterns per color at 12,24,36,48,60,72,84, and 96% channel stimulus. There are no DVD patterns for this so it only makes sense to use it with the internal generator. Please test and comment. thanks.
post #2330 of 3446
Thread Starter 
Quote:
Originally Posted by dschlic1 View Post

As far as I can tell the computer does not have a profile loaded. How do I determine what the gamma table is?

There is usually a control panel for the video card that might have gamma controls. The easiest way I know of to make sure linear tables are loaded is to use the command line utility "dispwin -c" from the argyllCMS package.

I did a lot of testing of this issue a few months ago before I moved over to the HCFR generator as my calibration source. I use it now for everything as it's much faster and more versatile than DVD based patterns.

Here is just one test comparing the GCD gray scale run (0-100%) to HCFR patterns. dE of HCFR peaked at 1.1 in the 30-40% range where pattern geometry has some small effects but everywhere else it's below 1 and xy shifts are also very low indicating that the RGB triplets getting sent are accurate.



and HCFR with respect to Mascior's disk


Edited by zoyd - 9/25/13 at 8:18am
post #2331 of 3446
How about saturation targets with 3045 when preferences like at picture?

post #2332 of 3446
Thread Starter 
I'm not sure saturation runs make sense for color spaces not fully saturated (like 75% and CC6). They are meant to be used as a validation tool for the entire color space so I think I will disable the feature for 75%/CC6 calibration targets.
post #2333 of 3446
Quote:
Originally Posted by dschlic1 View Post

As far as I can tell the computer does not have a profile loaded. How do I determine what the gamma table is?

CalibrationTester from GretaMacbeth is a good tool fot that.
http://torentilo.com/download/6324709/GretagMacbeth-ProfileMaker-Professional-5.0.10.html
post #2334 of 3446
Quote:
Originally Posted by zoyd View Post

I'm not sure saturation runs make sense for color spaces not fully saturated (like 75% and CC6). They are meant to be used as a validation tool for the entire color space so I think I will disable the feature for 75%/CC6 calibration targets.

Ok, just uploaded that because with 3.0.4.4 it showed saturation targets correctly even if it was done with same 75% preference/patterns.



Samsung PS64E8005 1030h 120cdm2 Cinema calibrated Masciors Disc Pana matrix.zip 4k .zip file
Edited by Make73 - 9/26/13 at 6:28am
post #2335 of 3446
Thread Starter 
gotcha, I've updated to previous behavior.
post #2336 of 3446
Quote:
Originally Posted by zoyd View Post

Note to anyone using Mascior's color checker patterns: Because there isn't a 100% white in that set you have to do primaries run at 100% intensity prior to running the color checkers to get the right deltaE values.

So I have to use 100A 100S patterns for primaries?

Did you test color checker with MCD?
Why HCFR asks black for last pattern when at MCD it´s not black, should I still use black pattern?

(btw, downloaded your 3.0.4.5 files again but didn´t affect for saturation points when 75% at preferences or did you updated something already?)
post #2337 of 3446
Quote:
Originally Posted by Make73 View Post

So I have to use 100A 100S patterns for primaries?

Did you test color checker with MCD?
Why HCFR asks black for last pattern when at MCD it´s not black, should I still use black pattern?

(btw, downloaded your 3.0.4.5 files again but didn´t affect for saturation points when 75% at preferences or did you updated something already?)

Hi Make73,

Yes, you still use the pattern labeled "black" from the color checker sweep as it provides the correct triplets that the hcfr software is looking for. I know it looks closer to 20% stimulus and the triplets are actually 62,61,62 unlike the standard video black we are all use to 16,16,16.

Ryan
post #2338 of 3446
Quote:
Originally Posted by mascior View Post

Hi Make73,

Yes, you still use the pattern labeled "black" from the color checker sweep as it provides the correct triplets that the hcfr software is looking for. I know it looks closer to 20% stimulus and the triplets are actually 62,61,62 unlike the standard video black we are all use to 16,16,16.

Ryan

Thanks Ryan smile.gif

I still don´t get it.
Tested with version 3.0.4.4 and 3.0.4.5
When use continuous measure it shows everything ok at CIE diagram but when run ColorChecker measure it shows something wierd places at CIE like this:

Edited by Make73 - 9/28/13 at 11:46pm
post #2339 of 3446
Thread Starter 
select "MCD" in the dropdown box preferences->references. Looks like some of the labels on the CIE chart for the measurement points still need fixing but dEs and target labels should be correct. Regarding the saturation target fix, I've updated that in the code but not uploaded an exe yet.
Edited by zoyd - 9/27/13 at 6:00am
post #2340 of 3446
Quote:
Originally Posted by Make73 View Post

Thanks Ryan smile.gif

I still don´t get it.
Tested with version 3.0.4.4 and 3.0.4.5
When use continuous measure it shows everything ok at CIE diagram but when run ColorChecker measure it shows something wierd places at CIE like this:


Also dE´s are all wrong calculated, so what I´m doing wrong redface.gif

Samsung PS64E8005 1249h 120cdm2 Movie calibrated Masciors Disc Pana matrix.zip 4k .zip file

Unfortunately I've only tested my color checker pattern sweep with ChromaPure software and I didn't see any discrepancies. Maybe zoyd could shed some light and see why you are encountering these issues.

Ryan
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Display Calibration
AVS › AVS Forum › Display Devices › Display Calibration › HCFR - Open source projector and display calibration software