MadVR - ArgyllCMS - Page 129 - AVS Forum | Home Theater Discussions And Reviews
Baselworld is only a few weeks away. Getting the latest news is easy, Click Here for info on how to join the Watchuseek.com newsletter list. Follow our team for updates featuring event coverage, new product unveilings, watch industry news & more!


Forum Jump: 
 285Likes
Reply
 
Thread Tools
post #3841 of 3853 Old 07-09-2015, 09:05 AM
Advanced Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 620
Mentioned: 22 Post(s)
Tagged: 0 Thread(s)
Quoted: 281 Post(s)
Liked: 180
Quote:
Originally Posted by mightyhuhn View Post
you can use madVR to disable the gamma ramp and use the 3d lut for full calibration. after madVR is closed the gamma ramp calibration is used again.
madVR will handle this automatically.

Quote:
Originally Posted by mightyhuhn View Post
is there really a way to force an icc profile for all games?
No, games don't use them. The only thing that affects games is the videoLUT (1D calibration affecting whitepoint/grayscale and single channels). There are ways to force the videoLUT, and some games may not reset them anyway.

DCG (dispcalGUI) - Graphical front-end for Argyll CMS display calibration and characterization
Current stable version 3.0.3 (released 2015-07-06) | DCG on Facebook

Current development snapshot: 3.0.3.3
Beta (released 2015-08-01) | Standalone | Changelog
fhoech is online now  
Sponsored Links
Advertisement
 
post #3842 of 3853 Old 07-11-2015, 04:39 PM
Member
 
Join Date: Jan 2015
Posts: 36
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 25 Post(s)
Liked: 1
I've been using dispcalGUI and ArgyllCMS for a while now to create 1DLUT's and 3DLUT's, but just now tested my 1DLUT properly and realized that 1DLUT's still raise my lowest black level point, even though the first set of entries for the .cal file are these:
BEGIN_DATA
0.00000000 0.00000000 0.00000000 0.00000000

Technically, doesn't that mean that the 1DLUT (.cal file) is using the lowest black level possible? When my TV is not using any LUT's at all, my brightness must be set to 46 for it to pass black level test (like this one - http://www.lagom.nl/lcd-test/black.php ). Anything below 46 and the top-left square disappears entirely. However, once I apply 1DLUT with "0, 0, 0, 0" data for the lowest black level, I can reduce brightness from 46 to 44 and still see the top-left square. I checked and re-checked and I am at a loss as to how that is even possible...
Masharak is online now  
post #3843 of 3853 Old 07-15-2015, 06:32 AM
Advanced Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 620
Mentioned: 22 Post(s)
Tagged: 0 Thread(s)
Quoted: 281 Post(s)
Liked: 180
I see no indication of raised black level. Near black values are not black. Black is only the 0,0,0 (or 16,1,6,16 in video levels) value.

Quote:
Originally Posted by Masharak View Post
When my TV is not using any LUT's at all, my brightness must be set to 46 for it to pass black level test (like this one - http://www.lagom.nl/lcd-test/black.php ). Anything below 46 and the top-left square disappears entirely. However, once I apply 1DLUT with "0, 0, 0, 0" data for the lowest black level, I can reduce brightness from 46 to 44 and still see the top-left square.
Not sure why you are puzzled by this. It just means your TV's native tone response is different from the target used for the 3D LUT, which is pretty normal. If your TV's black level is low enough, and the native response follows a pure power curve closely, then it may visually crush the first one or two levels above black (but they may still measure distinctly from black, so not necessarily clipping). The 3D LUT may raise an incoming signal of e.g. 1,1,1 to 2,2,2 or some other higher value, to meet the chosen target tone response. If you then lower the black level, you may actually end up clipping some near-black values.
Also note that your browser's color management settings may influence how the test page you linked is displayed. You'll only get visually correct results if an appropriate display profile is assigned, and your browser is set to color manage untagged content (assuming sRGB as source color space). To set the black level correctly, you actually want to disable any form of color management, or use a test disc with any 3D LUT or other processing disabled.

DCG (dispcalGUI) - Graphical front-end for Argyll CMS display calibration and characterization
Current stable version 3.0.3 (released 2015-07-06) | DCG on Facebook

Current development snapshot: 3.0.3.3
Beta (released 2015-08-01) | Standalone | Changelog

Last edited by fhoech; 07-15-2015 at 06:39 AM.
fhoech is online now  
post #3844 of 3853 Old 07-21-2015, 05:51 AM
Member
 
Sithuk's Avatar
 
Join Date: Mar 2011
Posts: 18
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 10 Post(s)
Liked: 10
Is there a guide online that would explain why video playback needs a 3D LUT for accurate calibration but gaming only needs a 1D LUT?

Quote:
Originally Posted by fhoech View Post
Quote:
Originally Posted by mightyhuhn View Post
you can use madVR to disable the gamma ramp and use the 3d lut for full calibration. after madVR is closed the gamma ramp calibration is used again.
madVR will handle this automatically.

Quote:
Originally Posted by mightyhuhn View Post
is there really a way to force an icc profile for all games?
No, games don't use them. The only thing that affects games is the videoLUT (1D calibration affecting whitepoint/grayscale and single channels). There are ways to force the videoLUT, and some games may not reset them anyway.
Sithuk is offline  
post #3845 of 3853 Old 07-21-2015, 06:01 AM
Advanced Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 620
Mentioned: 22 Post(s)
Tagged: 0 Thread(s)
Quoted: 281 Post(s)
Liked: 180
Quote:
Originally Posted by Sithuk View Post
Is there a guide online that would explain why video playback needs a 3D LUT for accurate calibration but gaming only needs a 1D LUT?
It's not about need, it's about support. Games usually don't support any form of color management, so in essence you have three options (which are not mutually exclusive):

  • Adjust the display (if possible) to match e.g. Rec. 709 / Rec. 1886 (i.e. use something like HCFR)
  • Use 1D LUT calibration to correct the gray balance and target a specific tone response (e.g. Rec. 1886)
  • Use an external 3D LUT box (e.g. eeColor)

DCG (dispcalGUI) - Graphical front-end for Argyll CMS display calibration and characterization
Current stable version 3.0.3 (released 2015-07-06) | DCG on Facebook

Current development snapshot: 3.0.3.3
Beta (released 2015-08-01) | Standalone | Changelog
fhoech is online now  
post #3846 of 3853 Old 07-21-2015, 07:15 AM
Member
 
VerGreeneyes's Avatar
 
Join Date: May 2013
Posts: 116
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 13
Or hook into the game or the compositor (in windowed mode) and apply the 3DLUT on top of the game's graphics (though it will be an 8-bit -> 8-bit correction). I don't think a ready-made solution like that currently exists though (the extent of my ability to write hooks has been to keep a game from resetting the 1DLUT, but I think it'd be great if someone with more experience made a solution to apply 3DLUTs to everything). It'd be great if SweetFX could do it for instance.

Edit: Actually, I did find MasterEffect (works on top of ReShade) which claims to support 3DLUTs. I haven't tried it though, so I don't know what formats it supports.

Edit2: It seems to use a 256x16 pixel png file for the 3DLUT, which I assume corresponds with a 16x16x16 3DLUT. Looking at the code (line 2909 and on of ReShade.fx from MasterEffect 1.1.287) it might not be hard to change the size though. The calculation would be more straightforward if it were stored in a 3D texture, but I don't know if ReShade supports that.

Edit3: The ReShade forums seem to indicate that ReShade does not support 3D textures, hence the workaround. In theory you could probably translate a 4096x4096 2D texture (something GPUs should be able to fit) to 256x256x256 coordinates though.

Last edited by VerGreeneyes; 07-21-2015 at 10:10 AM.
VerGreeneyes is offline  
post #3847 of 3853 Old 07-21-2015, 05:35 PM
Advanced Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 620
Mentioned: 22 Post(s)
Tagged: 0 Thread(s)
Quoted: 281 Post(s)
Liked: 180
Quote:
Originally Posted by VerGreeneyes View Post
Edit: Actually, I did find MasterEffect (works on top of ReShade) which claims to support 3DLUTs. I haven't tried it though, so I don't know what formats it supports.

Edit2: It seems to use a 256x16 pixel png file for the 3DLUT, which I assume corresponds with a 16x16x16 3DLUT. Looking at the code (line 2909 and on of ReShade.fx from MasterEffect 1.1.287) it might not be hard to change the size though. The calculation would be more straightforward if it were stored in a 3D texture, but I don't know if ReShade supports that.
Interesting. The 3D LUT functionality in MasterEffect seems to be a fairly recent addition. 16x16x16 is somewhat limited compared to madVR, but it's a starting point. I'm inclined to put the generation of compatible PNGs on the feature list for the next DCG devel snapshot, but it would be great if people could test if the current 3D LUT implementation in MasterEffect is actually working and usable first.

DCG (dispcalGUI) - Graphical front-end for Argyll CMS display calibration and characterization
Current stable version 3.0.3 (released 2015-07-06) | DCG on Facebook

Current development snapshot: 3.0.3.3
Beta (released 2015-08-01) | Standalone | Changelog
fhoech is online now  
post #3848 of 3853 Old 07-24-2015, 05:49 AM
AVS Special Member
 
mightyhuhn's Avatar
 
Join Date: Nov 2013
Posts: 1,072
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 522 Post(s)
Liked: 169
got it working and it's pretty easy too. after i learned that # is not to commentated a line.

here a screen with a custom ms paint splitscreen 3dlut: http://abload.de/img/working3dlutpys78.png

how about a new thread. i mean this has nothing to do with madVR.
|Tch0rT| likes this.
mightyhuhn is offline  
post #3849 of 3853 Old Yesterday, 03:38 AM
Senior Member
 
ttnuagmada's Avatar
 
Join Date: Jun 2011
Posts: 266
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 14
Forgive me if this has been covered somewhere, but I'm having an issue getting gamma/blacks to behave how I want them on my ST30 (plasma for those who dont know). I haven't seem to have been able to find a good setting for gamma calibration that doesn't end up doing something like raising my MLL, or blowing out the near-blacks. I'm actually happy with the sets native gamma response, but even if I set it all to native/measured/whatever with Dispcal, whenever I go to make the 3Dlut, it forces me to select a gamma setting, which ends up messing it up anyway.

Is there anyway to create a 3Dlut that just fixes saturation/greyscale etc, without messing with the gamma? Or can anyone tell me the best settings to use on a panasonic plasma as far as getting a gamma result that doesn't mess with black levels?

I'm using a i1D3 profiled off of a Colormunki photo.

Thanks
ttnuagmada is online now  
post #3850 of 3853 Old Yesterday, 07:02 AM
Member
 
Join Date: Jan 2015
Posts: 36
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 25 Post(s)
Liked: 1
Oh wow - 3DLUT support in video games and dispcalGUI supports it! Where can I find a direct download link to ArgyllCMS 1.8 beta?
Masharak is online now  
post #3851 of 3853 Old Yesterday, 09:35 AM
Advanced Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 620
Mentioned: 22 Post(s)
Tagged: 0 Thread(s)
Quoted: 281 Post(s)
Liked: 180
Quote:
Originally Posted by ttnuagmada View Post
[...]even if I set it all to native/measured/whatever with Dispcal, whenever I go to make the 3Dlut, it forces me to select a gamma setting,
If the display's native tone curve is different from the target, then the 3D LUT will adjust it to hit the target. You can adjust the tone curve parameters on the "3D LUT" tab. Increasing the output offset for example moves the target curve closer to a pure power function.

DCG (dispcalGUI) - Graphical front-end for Argyll CMS display calibration and characterization
Current stable version 3.0.3 (released 2015-07-06) | DCG on Facebook

Current development snapshot: 3.0.3.3
Beta (released 2015-08-01) | Standalone | Changelog
fhoech is online now  
post #3852 of 3853 Old Yesterday, 10:32 AM
zel
Member
 
zel's Avatar
 
Join Date: Mar 2004
Location: California
Posts: 90
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 29 Post(s)
Liked: 20
Quote:
Originally Posted by ttnuagmada View Post
Forgive me if this has been covered somewhere, but I'm having an issue getting gamma/blacks to behave how I want them on my ST30 (plasma for those who dont know). I haven't seem to have been able to find a good setting for gamma calibration that doesn't end up doing something like raising my MLL, or blowing out the near-blacks.
I had similar issues trying to profile my plasmas.

This worked for me:
  1. Use the spectral sample for plasma. This gave me better accuracy than profiling off an i1pro (assuming my display hasn't drifted much from the pro calibration)
  2. Keep the meter off the screen (thanks sillysally for the tip!). I found that having a 1-2cm gap gave me much better results with not too much contribution from the ambient light.
  3. Run a slow calibration. I used a 2.2 gamma target with all input offset (which I guess may be a difficult target) but got very good results.
  4. Even a ~400 patches profile gave pretty good results with the calibration - you can try first with a low patch count and if the results are good, reuse the calibration and measure a few thousand patches.

The results on a zt60 are very close to the professional calibration with bt.1886 gamma, but with better dEs and improved near black response. Calibration with TV's controls clips level 17, the DCG/ArgyllCMS created 3dlut is able to control that too.

Last edited by zel; Yesterday at 10:50 AM.
zel is online now  
post #3853 of 3853 Old Today, 07:24 AM
Senior Member
 
ttnuagmada's Avatar
 
Join Date: Jun 2011
Posts: 266
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 14
Quote:
Originally Posted by zel View Post
I had similar issues trying to profile my plasmas.

This worked for me:
  1. Use the spectral sample for plasma. This gave me better accuracy than profiling off an i1pro (assuming my display hasn't drifted much from the pro calibration)
  2. Keep the meter off the screen (thanks sillysally for the tip!). I found that having a 1-2cm gap gave me much better results with not too much contribution from the ambient light.
  3. Run a slow calibration. I used a 2.2 gamma target with all input offset (which I guess may be a difficult target) but got very good results.
  4. Even a ~400 patches profile gave pretty good results with the calibration - you can try first with a low patch count and if the results are good, reuse the calibration and measure a few thousand patches.

The results on a zt60 are very close to the professional calibration with bt.1886 gamma, but with better dEs and improved near black response. Calibration with TV's controls clips level 17, the DCG/ArgyllCMS created 3dlut is able to control that too.
I'll give these tips a shot when i try again this weekend. Personally I'd be happy if I could just get dispcal to do gamut and greyscale without even touching gamma, but im not sure how to do that.
ttnuagmada is online now  
Sponsored Links
Advertisement
 
Reply Display Calibration

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off