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 81

post #2401 of 3434
Thank you Zoyd.
post #2402 of 3434
Further information about the Pattern intensity measurments, the (Primary/Secondary) white Pattern seems correct but the (Primary/Secondary) white data value is the one of the grey 100 % (White) .
post #2403 of 3434
Thread Starter 
Quote:
Originally Posted by Kerlucun View Post

Further information about the Pattern intensity measurments, the (Primary/Secondary) white Pattern seems correct but the (Primary/Secondary) white data value is the one of the grey 100 % (White) .

I was unable to reproduce either error you mentioned. I ran 75% intensity primary/secondary measurements and the white value used was from that run with dE's as expected. The grayscale white value is only used for MCD color checker patterns. I also ran simultaneous saturation measurements using a D3 and i1pro and they were fine but you must use 100% intensity primary/saturation run prior to these to get correct dE.

Please send .chc file with the error.
post #2404 of 3434
Quote:
Originally Posted by zoyd View Post

I was unable to reproduce either error you mentioned. I ran 75% intensity primary/secondary measurements and the white value used was from that run with dE's as expected. The grayscale white value is only used for MCD color checker patterns. I also ran simultaneous saturation measurements using a D3 and i1pro and they were fine but you must use 100% intensity primary/saturation run prior to these to get correct dE.

Please send .chc file with the error.

I have made some more tests on my XP computer and I can reproduce the first but not the second , so that I wil try again on my W7 Laptop.
Please find the asked files plus two images.










HCFR.zip 151k .zip file
post #2405 of 3434
Thread Starter 
Quote:
Originally Posted by Kerlucun View Post

I have made some more tests on my XP computer and I can reproduce the first but not the second , so that I wil try again on my W7 Laptop.
Please find the asked files plus two images.

Did you run anything after doing the primaries/secondaries? The only way I can reproduce what you are seeing is if I do this:

1. Run grayscale
2. Run primaries/secondaries at 50% intensity
3. Run contrast

The third step will replace YWhite on the contrast page with 100% White and that then get's used for primaries/secondaries.

By the way, 50% intensity would be (0.5)^gamma * 54 = 11.75 cd/m^2 If you want something close to 50% luminance use 75% intensity.
post #2406 of 3434
Quote:
Originally Posted by zoyd View Post

Did you run anything after doing the primaries/secondaries?

No I run straight 'Mesure grey scale, primary and secondary colors'.

Quote:
Originally Posted by zoyd 
The third step will replace YWhite on the contrast page with 100% White and that then get's used for primaries/secondaries.
Perhaps is it the same behaviour for the one step measurement?

Quote:
Originally Posted by zoyd 
By the way, 50% intensity would be (0.5)^gamma * 54 = 11.75 cd/m^2 If you want something close to 50% luminance use 75% intensity.

Correct I forgot the gamma.mad.gif
post #2407 of 3434
Thread Starter 
Quote:
Originally Posted by Kerlucun View Post

No I run straight 'Mesure grey scale, primary and secondary colors'.
Perhaps is it the same behaviour for the one step measurement?

yep, it's the combination gray scale+primaries/secondaries. Fixed now. I'll update the current release with the fix and new patgen features. In the near future John will have the Argyll sensor 1.6 updates and I plan on folding those in and releasing a version for probe testing.
post #2408 of 3434
Thread Starter 
Minor update release:

Modifications as of 10/18/2013
-Version 3.0.5.2
-Added targeting mode for internal pattern generator
-Added continuous measures capability to pattern display for grayscale andprimaries/secondaries
-Fixed use of primaries/secondaries 100% white when running grayscale+primaries/secondaries in 1 shot
-corrected small error in color generator for 1 checker pattern, fixed simulated probe use of gamma
post #2409 of 3434
Thanks again, great work cool.gif

Is there something wrong with xy position and dE calculation with latest version, CIE Diagram shows different than chart

Just noticed that you cant use calibration settings measured by PS3 via HDMI for DLNA usage with PS3 Media Server, grayscale is ok but colors not. So you have to calibrate DLNA with Mascior´s mp4 version.
Edited by Make73 - 10/23/13 at 11:04pm
post #2410 of 3434
Thread Starter 
Quote:
Originally Posted by Make73 View Post

Thanks again, great work cool.gif

Is there something wrong with xy position and dE calculation with latest version, CIE Diagram shows different than chart

Thanks for picking that up. For the special color space mode HDTV(75%) the CIE chart is using your measured average gamma (2.46) to calculate dxy and dE but the primaries/secondaries page is still using 2.22. Will be fixed in next update.
post #2411 of 3434
Hi, zoyd. In the CIE Diagram the item 'Display free measures' option is missing in version 3.0.5.2. Without this item the display can get messy.

Larry
post #2412 of 3434
Thread Starter 
Quote:
Originally Posted by LarryInRI View Post

Hi, zoyd. In the CIE Diagram the item 'Display free measures' option is missing in version 3.0.5.2. Without this item the display can get messy.

Larry

Thanks Larry, it got lost with addition of color checker display options. Current build has been updated with fix.

Modifications as of 10/24/2013
-Added display free measures option back to CIE plot
-Fixed 75%/75% primaries/saturations dE and dxy use of measured gamma
-Fixed 75%/75% blue pattern for internal generator
post #2413 of 3434
Yeah, I figured that it got commented out.

Great job that you're doing. The program certainly has almost all that the commercial versions have.

Larry
post #2414 of 3434
Zoyd,

Does " Include L in greyscale delta E" get checked if i'am using itu-1886 ?

Whats it for specifically?

Thanks
post #2415 of 3434
Thread Starter 
Quote:
Originally Posted by PeterLewis View Post

Zoyd,

Does " Include L in greyscale delta E" get checked if i'am using itu-1886 ?

no
Quote:
Whats it for specifically?

Thanks

When checked, delta L (lightness difference) is included in the calculation of the deltaE CIELUV formula, this avoids overemphasizing chromaticity errors at the low end of the gray scale and gives a more perceptually uniform measure for the gray scale errors.
post #2416 of 3434
I´m sorry to bother You zoyd but is there still something strange with dE calculations between chart and CIE or is it because of used 75% patterns?

btw, you can´t get USB/DLNA properly calibrated...however colors are pretty good at 75%
Edited by Make73 - 10/28/13 at 9:34pm
post #2417 of 3434
Thread Starter 
Quote:
Originally Posted by Make73 View Post

I´m sorry to bother You zoyd but is there still something strange with dE calculations between chart and CIE or is it because of used 75% patterns?

No problem, I'm glad you are finding the hard to find bugs. The last fix was applied to primaries but not secondaries, new executable attached you can just drop in your program directory.
Quote:
btw, you can´t get USB/DLNA properly calibrated...however colors are pretty good at 75%

What is the signal path/renderer and why does it measure differently than the DVD patterns? The gamut is undersaturated compared to the other two.

ColorHCFR_3052_fixed75secondary.zip 2582k .zip file
Edited by zoyd - 10/25/13 at 5:17am
post #2418 of 3434
Zoyd, when calibrating with 4%apl and you want to calibrate to 40FTL, do you use the 100% white 4%apl pattern knowing that the brightness is purposely being lowered by the tv's abl or do you go by the checkerboard pattern or some other pattern?
post #2419 of 3434
Quote:
Originally Posted by zoyd View Post

No problem, I'm glad you are finding the hard to find bugs. The last fix was applied to primaries but not secondaries, new executable attached you can just drop in your program directory.
What is the signal path/renderer and why does it measure differently than the DVD patterns? The gamut is undersaturated compared to the other two.

ColorHCFR_3052_fixed75secondary.zip 2582k .zip file

Big thanks for the fix smile.gif

Have you tried calibrate your D-serie with Mascior´s mp4 version via USB or PS3 Media Server?
At least with E (also LED ES) it´s totally different than via HDMI. Grayscale is ok but colors not.
I don´t know why gamut is so undersaturated, had to set color from default 50 to 82 so you can get green closer to where it belongs.
I have send feedback to Samsung but seems they are not interested.
post #2420 of 3434
Thread Starter 
@PL
The whole purpose of using constant energy patterns like those is that they better represent an operating point that actual content will generate, so if you can get your desired peak white using them you should. And, ABL should not be a problem with those patterns anyway.
post #2421 of 3434
Thread Starter 
Quote:
Originally Posted by Make73 View Post

Big thanks for the fix smile.gif

Have you tried calibrate your D-serie with Mascior´s mp4 version via USB or PS3 Media Server?
At least with E (also LED ES) it´s totally different than via HDMI. Grayscale is ok but colors not.
I don´t know why gamut is so undersaturated, had to set color from default 50 to 82 so you can get green closer to where it belongs.
I have send feedback to Samsung but seems they are not interested.

No, I've never tried to calibrate that path but if you have to up the color control that much something is broken in rendering the MP4 to RGB. When you say USB, is that attached to the PS3 or directly to display?
post #2422 of 3434
Quote:
Originally Posted by zoyd View Post


I've added this capability to the internal pattern generator for HCFR. Please test and comment using linked file below (drop files in current HCFR directory). Select Target from the drop-down box in the information windows and then select a gray scale level or primary color from the measurements grid. This color will be sent to the pattern generator for display in either GDI or madVR view images mode during single and continuous measures. During continuous measures you can select other colors and it will switch on the fly.

Make sure that all three boxes under preferences->general->user measures are checked and note that this only works for grayscale and primaries/secondaries for full colorspace modes (rec709, rec601, etc.), not special modes like CC6

ColorHCFR_test_internalpat.zip 3775k .zip file

Thank you, Zoyd!

I just did another calibration round:

The function to select individual targets and have the madTPG display the right patch

seems to work very well!

Thanks again! Especially when calibrating RGB levels I find it very helpful getting continous feedback for one level and search

for the right RGB balance.

 

One question:

With my JVC RS40 projector I'm calibrating for BT.1886.

Two weeks ago HCFR painted a Gamma reference line that fell at 30% to 5% from 2.3 Gamma to 2.1.

 

Now I always get a completely horizontal reference line (painted in white) at 2.4.

 

Is that right?

Any ideas?

 

FYI: I'm using the madvr preset in the new DispcalGUI versions after a manual calibration of the RS40.


Edited by TimHamburg - 10/25/13 at 7:18am
post #2423 of 3434
Thread Starter 
Quote:
Originally Posted by TimHamburg View Post


Now I always get a completely horizontal reference line (painted in white) at 2.4.

Check your measurement of black, this should only happen when black is measured as 0 (or missing).
post #2424 of 3434
Quote:
Originally Posted by zoyd View Post

No, I've never tried to calibrate that path but if you have to up the color control that much something is broken in rendering the MP4 to RGB. When you say USB, is that attached to the PS3 or directly to display?

Directly to PS64E8000 so there isn´t PS3 or PS3 Media Server causing this problem. (actually I tried both ways, directly to TV with USB and via DLNA with PS3 Media Server)
btw, this is not because of Evolution Kit, done measurements also before installing that and results was same kind of.

Maby if you have some extra time you could try copy patterns to USB and make measurement, it would be interesting to see if same is with D-series.
post #2425 of 3434
Quote:
Originally Posted by zoyd View Post


Check your measurement of black, this should only happen when black is measured as 0 (or missing).


Yes - there is only a very high value fr delta xy - the other fileds are empty at 0%.

What am I to do?

My living room isn't fully light-controlled - so I think a falling gamma-curve would make sense.

 

(Sensor is the I1D3 - projector a JVC RS40)

post #2426 of 3434
Thread Starter 
You need to get a good measurement of the black level off the screen and then just enter that value in the 0% Y field.

1. if checked->uncheck XYZ adjustment
2. check editable data
3. Enter black Y value in 0% position
4. Recheck XYZ adjustment if you unchecked it 1.
post #2427 of 3434
Thread Starter 
I have a new version for testing that includes John's recent update of the Argyll interface to handle ArgyllCMS 1.6 meter code. I've tested with the spyder2, D3, D2, and i1pro2. Please test and report any problems as I suspect some meters may not work right off the bat. Supported meters and some Driver installation notes can be found here. You will not have to update the driver for the D3.

Usage notes:

  • New driver installation files can be found in the program directory under drivers/usb - point to this directory when updating for those meters than need it.
  • Logging can be turn on by checking the appropriate box on the meter properties page, messages are sent to HCFR.log
  • The D3 integration parameters are at their default values so this meter will run a bit faster compared to previous versions, please report any unstable readings.

Updated test version can be found here.
Edited by zoyd - 10/31/13 at 9:27am
post #2428 of 3434
Does this add calibration for colormunki display?

I use the factory XRITE DLL for the munki,does the new argyll code benefit me?
post #2429 of 3434
Thread Starter 
Quote:
Originally Posted by PeterLewis View Post

Does this add calibration for colormunki display?

I use the factory XRITE DLL for the munki,does the new argyll code benefit me?

The refresh rate calibration can't be performed by the colormunki in an acceptable amount of time so that is not available, whether or not it will perform better than using the XRITE dll I don't know as I don't have a colormunki to test with, can you download it and try?
post #2430 of 3434
Quote:
Originally Posted by zoyd View Post

The refresh rate calibration can't be performed by the colormunki in an acceptable amount of time so that is not available, whether or not it will perform better than using the XRITE dll I don't know as I don't have a colormunki to test with, can you download it and try?

Thanks,

I only brought it up because greamegill said that there is a possibility...

I can dl new code and try early in the week as there is too much sports on at the moment (hardcore sports junkie)....
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