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 82

post #2431 of 3439
Thread Starter 
Here is a new version for testing prior to release on sourceforge. In addition to the new Argyll meter code the following version also adds:

  • Color difference formulas, CIE76, CIE2000, and CMC. Select on the advanced properties page.
  • Color difference handling of gray scale improved with added options.
  • Color coding of dE values in the measurement boxes and for the window border based on average.
  • Display of color checker and dE formula selection in the measurements window title
  • Switch for turning off the madVR 3dlut when sending test patterns through madVR

The criteria for green/yellow/red highlighting is slightly different based on which color difference formula you are using but generally < 2 = green, >2 and < 4 = yellow, > 4 = red

The gray scale dE handling has been updated to use all of the above formulas and has 3 options for handling the white reference.

1. Relative -> Y White = Y measured.
2. Absolute w/gamma -> Color reference Y set to desired Y gamma target.
3. Absolute w/o gamma -> Color reference Y set to measured Y. (perfect gamma)

The recommended default for grayscale tuning is CIE2000 with option 3. For more sensitivity to chromaticity errors use any of the color formulas with option 1 although this option over-emphasizes errors at low luminance.

For final proofing use option 2 which folds in the error caused by missing your gamma targets.

Comments are welcome!

Test it
Edited by zoyd - 10/31/13 at 2:11pm
post #2432 of 3439
Zoyd you're da man, your tremendous work is really appreciated, thanks for upgrading colorHCFR and keep this project alive smile.gif
post #2433 of 3439
Quote:
Originally Posted by zoyd View Post


Test it

Link doesn't seem to work?
post #2434 of 3439
Thread Starter 
yeah sorry, continued with updates, link is now available.
post #2435 of 3439
Thanks! The link works now. Great work with the HCFR updates. Software has come a long way since I last used it maybe a year ago.

I've just recently bought the P50GT60 and I'm in the middle of calibrating and trying to figure out where my i1d3 stands regarding its accuracy and what pattern method to use with calibration. Read a lot of threads here about different patterns and I won't get into that discussion here too much. smile.gif

I didn't get the pattern generator in HCFR to work properly in previous versions but I installed madVR and I'll be using it's pattern generator. I did a quick test with it and I'll probably do more testing with it tonight with different settings regarding window size and background intensity and see how they measure up. I used blu ray disc but it's awfully slow compared to PC pattern generator so I'd rather use the PC to calibrate once I'm sure that it sends correct signals and the results are transferable to the Blu Ray player that I have.
post #2436 of 3439
Thread Starter 
Great, let us know how it goes. You don't really need madVR if you're calibrating for other sources (BD, DVD etc.), just use the internal pattern generator. If you will be using madVR for video rendering then sure, use it as your pattern generator.
post #2437 of 3439
Zoyd ,

I just launched your new beta,it showed an error after I ran it but it still worked after I cleared it....Also it crashed on me a few times.

Another note...The Spectral Samples are "NOT SHOWING" I need my Plasma EDR Spectral as it works a treat.

Btw so what do I use with CIE2000 ....

Relative Y
Absolute w/ gamma
Absolute w/o gamma

I'll tell you that CIE2000 formula made my last calibration look like a Chad B calibration with many de in the 0.? ....
Edited by PeterLewis - 11/1/13 at 5:34am
post #2438 of 3439
Thread Starter 
Quote:
Originally Posted by PeterLewis View Post

Zoyd ,

I just launched your new beta,it showed an error after I ran it but it still worked after I cleared it....Also it crashed on me a few times.

Another note...The Spectral Samples are "NOT SHOWING" I need my Plasma EDR Spectral as it works a treat.

Btw so what do I use with CIE2000 ....

Relative Y
Absolute w/ gamma
Absolute w/o gamma

I'll tell you that CIE2000 formula made my last calibration look like a Chad B calibration with many de in the 0.? ....

To everyone: In order to help me debug any bugs please include in your bug report the following information -
  • HCFR version# (should be 3.0.5.3 for this test version)
  • Type of meter
  • Screen shots or exact text of any error messages
  • HCFR.log (you need to enable logging on the meter properties page *before* you encounter an error)
  • If the program crashes give the circumstances (what sequences was being run, did it just stop and quit or hang, etc.)

@Peter - I'll look into the .ccss missing file issue, they should be available if they were previously available.

For final measurement of the display response performance I would use CIE2000 + option 2 for grayscale (Absolute w/gamma). Compared to previous versions the dE number you will get should be very close on the primaries/secondaries and saturation measurements since CIE94 and CIE2000 are very similar. The numbers you obtain for grayscale calibration will be lower because the previous formula used exclusively for grayscale (CIELUV) is a more sensitive metric. Of course none of this changes the actual calibration, these options are just different ways of quantifying the error for different purposes.
Edited by zoyd - 11/1/13 at 6:06am
post #2439 of 3439
Ok,I uninstalled it and reloaded it and the error it say's.....

RED X ....."Unable to execute file:
C/programfiles/hcfrcalibration/colorhcfr.exe

Createprocess failed;code 740.
The requested operation requires elevation."

No Spectral samples...shows none.

Also you have XYZ as default instead of xyY (newbie's will get confused)

EDIT: I like how you incorporated the word awesome calibration if many fall in the green.You should write "Try Again Dummy" if one does a bad cali.smile.gif
post #2440 of 3439
First of... thanks for the continuation of this project.

My meters have been dormant for some time now and is time to wake them up. I did some reading and installed the the latest version to give it a try.
I have two meters and Eye-1 (Great Mcbeth) and a Spyder2. I got the Eye-1 working, but read that the firmware for the Spyder2 needs to be converted. Can someone point me on where or how to do that?
post #2441 of 3439
Quote:
Originally Posted by JohnAd View Post


Yep, drives me mad too
it's on the list, I am focussing in driver issues for now though so won't get fixed in the next version but probably the one after that.


John
This seems to still be a problem as I have 3.0.5 and still cannot scroll down to see the luma errors...?

Edit: Lost the original quote here. The issue is about being able to scroll down far enough in some windows to see all data.
post #2442 of 3439
Thread Starter 
Quote:
Originally Posted by Barsk View Post

This seems to still be a problem as I have 3.0.5 and still cannot scroll down to see the luma errors...?

Edit: Lost the original quote here. The issue is about being able to scroll down far enough in some windows to see all data.

To get luminance errors to display click the little arrow shown below.

post #2443 of 3439
Thread Starter 
Quote:
Originally Posted by PeterLewis View Post

Ok,I uninstalled it and reloaded it and the error it say's.....

RED X ....."Unable to execute file:
C/programfiles/hcfrcalibration/colorhcfr.exe

Createprocess failed;code 740.
The requested operation requires elevation."

No Spectral samples...shows none.

Also you have XYZ as default instead of xyY (newbie's will get confused)

EDIT: I like how you incorporated the word awesome calibration if many fall in the green.You should write "Try Again Dummy" if one does a bad cali.smile.gif

wink.gif Don't know if that will stay in, I'll see what people think. (see if you can find "Super Awesome"). Are you running the program with administrator privileges? try that.


edit:

I've updated the linked package to include installation of .ccss files to the user's [APPDATA] location, this might fix the problem you are having with finding .ccss files.
Edited by zoyd - 11/1/13 at 8:10am
post #2444 of 3439
Thread Starter 
Quote:
Originally Posted by superleo View Post

First of... thanks for the continuation of this project.

My meters have been dormant for some time now and is time to wake them up. I did some reading and installed the the latest version to give it a try.
I have two meters and Eye-1 (Great Mcbeth) and a Spyder2. I got the Eye-1 working, but read that the firmware for the Spyder2 needs to be converted. Can someone point me on where or how to do that?

You'll need the installation disk setup.exe file or the CVSpyder.dll for the probe.

Run the program oeminst.exe in the tools directory to extract and install the .PLD file

oeminst -v infile

where infile = setup.exe or (path to installation disk setup.exe) or CVSpyder.dll
post #2445 of 3439
Quote:
Originally Posted by zoyd View Post


You'll need the installation disk setup.exe file or the CVSpyder.dll for the probe.

Run the program oeminst.exe in the tools directory to extract and install the .PLD file

oeminst -v infile

where infile = setup.exe or (path to installation disk setup.exe) or CVSpyder.dll

I'll give it a try, sounds easy enough. thanks for the response and thank you for the continued development.
post #2446 of 3439
Zoyd,

I just tried doing a calibration with your new beta using de2000 w/gamma and I had problems with free measures in the low ire (20,30)

For instance the rgb levels would never remain consistent,one moment ,one would say 100% then bounce down to 92% the next and also the delta errors in free measures are in the 100's.So something in not jiving.

I had to switch back to 3.0.5.2.
post #2447 of 3439
Thread Starter 
Quote:
Originally Posted by PeterLewis View Post

Zoyd,

I just tried doing a calibration with your new beta using de2000 w/gamma and I had problems with free measures in the low ire (20,30)

For instance the rgb levels would never remain consistent,one moment ,one would say 100% then bounce down to 92% the next and also the delta errors in free measures are in the 100's.So something in not jiving.

I had to switch back to 3.0.5.2.

That's why it's a testing release. smile.gif First off, did you get the edr file listing correctly?
-The unstable readings at 20% are because the default integration time in the ArgyllCMS meter code is a little low when in non-refresh mode (0.2 sec). I've changed it to be close to the 3.0.5.2 value, it's now 400 ms vs. 500 ms previously. This should stabilize those readings.

-The dE values reported in the history plots during free measures are identical to previous versions so you shouldn't see any problems there, they don't depend on what dE formula you have chosen. Only the values on the measurement pages use the new dE formulas.

New version updates
  • Stabilized 20% red reading for D3/munki
  • Added ambient mode
  • Fixed issue with projector mode

Test it
post #2448 of 3439
Quote:
Originally Posted by zoyd View Post

That's why it's a testing release. smile.gif First off, did you get the edr file listing correctly?
-The unstable readings at 20% are because the default integration time in the ArgyllCMS meter code is a little low when in non-refresh mode (0.2 sec). I've changed it to be close to the 3.0.5.2 value, it's now 400 ms vs. 500 ms previously. This should stabilize those readings.

-The dE values reported in the history plots during free measures are identical to previous versions so you shouldn't see any problems there, they don't depend on what dE formula you have chosen. Only the values on the measurement pages use the new dE formulas.

New version updates
  • Stabilized 20% red reading for D3/munki
  • Added ambient mode
  • Fixed issue with projector mode

Test it

Yes,the EDR file loaded up.

Well when I do free measures with 3.0.5.2 it would show de errors <6 but with the new beta it would show in the 100's.

Also about the meter code being 400ms as to 500ms, I was using the factory Colormunki DLL from Xrite so why would that need manipulating.Shoot,you should make it 300ms,I would sacrifice speed for proper accuracy.

I will try this new patch tomorrow....I will be your test dummy tongue.gif
Edited by PeterLewis - 11/2/13 at 3:43pm
post #2449 of 3439
Quote:
Originally Posted by zoyd View Post

To everyone: In order to help me debug any bugs please include in your bug report the following information -
  • HCFR version# (should be 3.0.5.3 for this test version)
  • Type of meter
  • Screen shots or exact text of any error messages
  • HCFR.log (you need to enable logging on the meter properties page *before* you encounter an error)
  • If the program crashes give the circumstances (what sequences was being run, did it just stop and quit or hang, etc.)

HCFR 3.0.5.3

If I input values into sensor matrix with decimal comma (i.e. 1,1) instead of decimal dot (1.1), the software converts the visible value to 1.10000, but something funky is going on in the background. If I go to sensor configuration again and just click OK, the it adds the matrix values to the current readings instead of recalculating them and thus multiplying the correction.

To reproduce:
  1. Create new measurement file and input at least one value to the sensor matrix with comma.
  2. Take a grey scale reading
  3. Click sensor configuration and click ok and the value should shift.

post #2450 of 3439
Testing new patch...

Low red reading still erratic.Bounces from 92%-100%.

Also still says 100 de in free measures.

I'am using original Xrite colormunki DLL driver.
post #2451 of 3439
Thread Starter 
Create a .chc file with the problem and send it to me along with the log. To do this follow these steps:

1. Delete HCFR.log in executable directory if there is one.
2. Start program and check the log box on the probe properties page.
3. Display a 20% pattern and do a grayscale run with just that pattern so you get 10 measurements of the same thing.
4. Save file and zip it up with log and attach to a post.

thanks
post #2452 of 3439
Thread Starter 
Quote:
Originally Posted by Aranis View Post

If I go to sensor configuration again and just click OK, the it adds the matrix values to the current readings instead of recalculating them and thus multiplying the correction.

Good catch, thanks! This will happen with either decimal decimal or comma decimal entry.

Work around for now:

If you want to manually edit your sensor matrix make sure the XYZ adustment button is unchecked before opening the sensor editor.
post #2453 of 3439
Thanks for the new version Zoyd. Using your latest posted one, I verified that my previous calibration done over 600 hours ago was still stable even with the new driver using my colormunki photo. I then did a new calibration and everything worked without any issues. Link is in my sig where I posted some screenshots of the results. I also left everything at defaults so using CIE2000 and absolute Y w/o gamma, only thing I changed was to change the reference gamma to 2.4 to go along with my higher gamma calibration.

Switching to Relative Y does give me big errors at 10 and 20, but I figure it's just my spectro not reading properly that low anyway.
post #2454 of 3439
Zoyd,

In the new version could you mute the color coding on the delta E lines? As it is now, the colors tend to pull one's eyes away from the other data. Maybe an option switch to turn the colors on or off?

Also, is the comment "Awesome Calibration" necessary? (I know that without the comment. smile.gif)


Keep up the great work,
Larry
post #2455 of 3439
Thread Starter 
I knew the messages would get old quick, I was gonna wait until someone unlocked "super awesome" before removing.

I will soften up the highlighting and see how that goes.
post #2456 of 3439
I actually was liking the highlighting on the dE's and the comments weren't bothering me, so keep that in mind when removing/adding stuff, some people like some may not. I guess the best solution is an option to toggle that stuff, though I understand that would mean more work for you.
post #2457 of 3439
Thread Starter 
How about this for a compromise?




I've also added custom white point.

post #2458 of 3439
Quote:
Originally Posted by zoyd View Post

I knew the messages would get old quick, I was gonna wait until someone unlocked "super awesome" before removing.

I will soften up the highlighting and see how that goes.

I vote to keep the messages in.

I also vote for Zoyd to fix the low level 20% ire reading.biggrin.gif

Zoyd, cant you just use your original Xrite dll that comes with your D3 and try that with mascior 4% apl 20% ire pattern and use de2000 w/gamma and put it in free measure for 5 minutes on that pattern and you will see the red% do the lambada dance.
post #2459 of 3439
Muy bueno.

My objection to the comments was only because we tend to clip and paste the data. To those not familiar with HCFR, it might appear to be bragging.

Larry
post #2460 of 3439
Thread Starter 
Quote:
Originally Posted by PeterLewis View Post

I vote to keep the messages in.

I also vote for Zoyd to fix the low level 20% ire reading.biggrin.gif

Zoyd, cant you just use your original Xrite dll that comes with your D3 and try that with mascior 4% apl 20% ire pattern and use de2000 w/gamma and put it in free measure for 5 minutes on that pattern and you will see the red% do the lambada dance.

I have already tested 20% patterns with my D3 and they are fine, how do you think I came up with the integration time changes? The xrite dll is irrelevant because the D3 and colormunki use the windows HID driver. So both the xrite software and HCFR can access the probe without installing special drivers. When you use HCFR you are using meter code from ArgyllCMS that is identical to that used in version 3.0.5.2 (after I updated the integration time), so you will have to produce some data showing a reproducible problem before I go and change anything else. I can not reproduce what you are seeing with my D3, the red readings are stable to +/- 2% on a 20% pattern.
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