MadVR - ArgyllCMS - Page 103 - AVS | 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: 
 180Likes
Reply
Thread Tools
post #3061 of 3391 Old 03-24-2015, 01:14 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,490
Mentioned: 61 Post(s)
Tagged: 0 Thread(s)
Quoted: 741 Post(s)
Liked: 640
Thanks - I'm not sure if there was another way to do it but because I had my own executable path set with the standalone version the 0install version picked that up and I needed to reset it.
zoyd is online now  
Sponsored Links
Advertisement
 
post #3062 of 3391 Old 03-24-2015, 01:20 PM
Senior Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 480
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 123
Quote:
Originally Posted by zoyd View Post
Thanks - I'm not sure if there was another way to do it but because I had my own executable path set with the standalone version the 0install version picked that up and I needed to reset it.
Ah, yes. The Argyll versions are the same for the two devel builds (both identify as 1.7b), so in this case it doesn't recognize there's a newer version available. There's a slightly easier way though: Rename the standalone Argyll directory temporarily, then use "Select Argyll CMS exes..." in DCG's file menu which should offer the 0install one automatically.
fhoech is online now  
post #3063 of 3391 Old 03-24-2015, 02:58 PM
Member
 
Join Date: Nov 2014
Posts: 105
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 78 Post(s)
Liked: 5
fhoech,

I did as you instructed and the process continued where it ended last night (after the calibration). It did more 154 patches and it began the "characterization" phase.

It is still on for the last 20 min:

"Acceleration grid res=23
There are 241 unique fixed point to add (247 total fixed points)
There are 8976 far spread points to add"

The Log file was not updated so far. Last entries were:

Code:
2015-03-24 21:37:56,710 Doing some initial measurements
2015-03-24 21:37:59,884 Measured display update delay of 0 msec, using delay of 100 msec & 0 msec inst reaction
2015-03-24 21:38:03,974 Red   = XYZ 87.967 45.267  3.243
2015-03-24 21:38:03,976 Green = XYZ 77.215 151.204 22.829
2015-03-24 21:38:03,976 Blue  = XYZ 32.788 14.570 172.767
2015-03-24 21:38:04,857 White = XYZ 207.896 218.328 238.231
2015-03-24 21:38:04,858 
2015-03-24 21:38:04,858 Adjust R,G & B gain to get target x,y. Press space when done.
2015-03-24 21:38:04,858    Target Br 218.33, x 0.3127 , y 0.3290
2015-03-24 21:39:05,956 dispcalGUI: Sending buffer: ' '
2015-03-24 21:39:06,618 / Current Br 220.11, x 0.3128-, y 0.3289+  DE  0.2  R-  G++ B-
2015-03-24 21:39:06,621 
2015-03-24 21:39:06,622 Press 1 .. 7
2015-03-24 21:39:06,630 1) Black level (CRT: Offset/Brightness)
2015-03-24 21:39:06,631 2) White point (Color temperature, R,G,B, Gain/Contrast)
2015-03-24 21:39:06,631 3) White level (CRT: Gain/Contrast, LCD: Brightness/Backlight)
2015-03-24 21:39:06,631 4) Black point (R,G,B, Offset/Brightness)
2015-03-24 21:39:06,632 5) Check all
2015-03-24 21:39:06,632 6) Measure and set ambient for viewing condition adjustment
2015-03-24 21:39:06,634 7) Continue on to calibration
2015-03-24 21:39:06,637 8) Exit
2015-03-24 21:39:06,640 dispcalGUI: Waiting for send buffer
2015-03-24 21:39:08,253 dispcalGUI: Sending buffer: '8'
2015-03-24 21:39:08,257 Exiting
2015-03-24 21:39:08,260 The instrument can be removed from the screen.
2015-03-24 21:39:08,512 dispcalGUI: Reached EOF (OK)
Inside folder there were no new files (besides the log and ccmx file).
CPU level is around 46% with "targen" process consuming around 37%.

I will let it run for some time and see what happens.
Pepelegal is offline  
post #3064 of 3391 Old 03-24-2015, 03:06 PM
Senior Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 480
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 123
Quote:
Originally Posted by Pepelegal View Post
Inside folder there were no new files (besides the log and ccmx file).
CPU level is around 46% with "targen" process consuming around 37%.
Note that you need to go into the temporary folder (C:\Users\Username\AppData\Local\Temp\dispcalGUI-xxxx), not the storage directory. An icm file should be there, along with the current targen log.
fhoech is online now  
post #3065 of 3391 Old 03-24-2015, 03:15 PM
AVS Special Member
 
sillysally's Avatar
 
Join Date: Jun 2006
Location: Chicago
Posts: 3,964
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 248 Post(s)
Liked: 332
Latest, argyllCMS 1.7b 03/23, dispcalGUI 2903, latest madVR and 0install.
Intel 2.6 quad, Win 7 Pro 64 bit, 20+ gigs of ram, SSD

I did a "create testchart" and used my last profile as pre condition (checked)
It took about 30 min to complete my 10508 point, normally about 4 min.

Ran a non-checked calibration and profile, just the RGBW and set meter then dispcalgui stated the profile.
The normal pop up "current calibration and curves, liner or enable calibration and curves". Didn't pop up.
Check on the progress about 6 hours later, I could see it was still processing the profile, readings were done.
The display screen had the black back ground with the dark gray window, madTPG bar on top.
I just unplugged my K10 and left the PC running. Checked about 3 hours later and the process had completed.

It would seem that processing the profile once done with the patch readings takes much longer than normal.
That probably is why others had reported the profile process hung up, and then stopped the process.

The pop up window after the profile process is done looks to be normal in what it is reporting.

"Create 3DLUT" takes about the same time to process as older versions.
Got a Warning, Black hack used more than once (125).
What does the above mean.????????????????????????????

Last night I was also looking for the argyllCMS installed folder, found it as fhoech said but its a string of numbers, no name.

ss

Last edited by sillysally; 03-24-2015 at 03:26 PM.
sillysally is offline  
post #3066 of 3391 Old 03-24-2015, 03:32 PM
Senior Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 480
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 123
Quote:
Originally Posted by sillysally View Post
Ran a non-checked calibration and profile, just the RGBW and set meter then dispcalgui stated the profile.
The normal pop up "current calibration and curves, liner or enable calibration and curves". Didn't pop up.
This only pops up when interactive adjustment is disabled and calibration tone curve is set to "As measured".

Quote:
Originally Posted by sillysally View Post
Check on the progress about 6 hours later, I could see it was still processing the profile, readings were done.
The display screen had the black back ground with the dark gray window, madTPG bar on top.
I just unplugged my K10 and left the PC running. Checked about 3 hours later and the process had completed.

It would seem that processing the profile once done with the patch readings takes much longer than normal.
That probably is why others had reported the profile process hung up, and then stopped the process.
The Argyll 1.7b devel builds I made available have been compiled with optimization disabled, and it seems this makes a big difference for processing times as testcharts get larger.

Quote:
Originally Posted by sillysally View Post
"Create 3DLUT" takes about the same time to process as older versions.
Got a Warning, Black hack used more than once (125).
What does the above mean.????????????????????????????
It's the new black level hack in Argyll CMS which forces input to output black for 3D LUTs. As far as I'm aware, the warning is harmless.
sillysally likes this.
fhoech is online now  
post #3067 of 3391 Old 03-24-2015, 03:58 PM
Member
 
Join Date: Nov 2014
Posts: 105
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 78 Post(s)
Liked: 5
Quote:
Originally Posted by fhoech View Post
Note that you need to go into the temporary folder (C:\Users\Username\AppData\Local\Temp\dispcalGUI-xxxx), not the storage directory. An icm file should be there, along with the current targen log.
Ok, "targen" process is still running after 1h16min (CPU 37%).

I did a copy of the folder, but I think will cancel the process as it taking a long time...

Let me know if you need something else
Attached Files
File Type: zip dispcalGUI-hd3y0b.zip (62.8 KB, 2 views)
Pepelegal is offline  
post #3068 of 3391 Old 03-24-2015, 05:19 PM
AVS Special Member
 
sillysally's Avatar
 
Join Date: Jun 2006
Location: Chicago
Posts: 3,964
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 248 Post(s)
Liked: 332
The 3DLUT that was made from the above post profile is only 13.7mb.
Normal has been 14.4mb.
Why the difference??????
Yes I did have eecolor checked.
sillysally is offline  
post #3069 of 3391 Old 03-24-2015, 05:39 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,490
Mentioned: 61 Post(s)
Tagged: 0 Thread(s)
Quoted: 741 Post(s)
Liked: 640
zoyd is online now  
post #3070 of 3391 Old 03-24-2015, 06:19 PM
AVS Special Member
 
sillysally's Avatar
 
Join Date: Jun 2006
Location: Chicago
Posts: 3,964
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 248 Post(s)
Liked: 332
Quote:
Originally Posted by zoyd View Post
File properties should report 13.7 MB (14,417,920 bytes) for the eeColor.txt file.
The bigger MB is from dispcalGUI 2.3.0.0
The Smaller is from dispcalGUI 2.9.0.3

ss
Attached Thumbnails
Click image for larger version

Name:	Properities Made with 2.3.0 dispcal.jpg
Views:	27
Size:	67.5 KB
ID:	622977   Click image for larger version

Name:	Properties Made with dispcal 2.9.0.3.jpg
Views:	21
Size:	72.2 KB
ID:	622985  
sillysally is offline  
post #3071 of 3391 Old 03-24-2015, 06:37 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,490
Mentioned: 61 Post(s)
Tagged: 0 Thread(s)
Quoted: 741 Post(s)
Liked: 640
The older Argyll code would write the 65th node (1.0) which is not adjustable so the newer code drops those points and produces smaller files.
sillysally likes this.
zoyd is online now  
post #3072 of 3391 Old 03-24-2015, 08:39 PM
AVS Special Member
 
sillysally's Avatar
 
Join Date: Jun 2006
Location: Chicago
Posts: 3,964
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 248 Post(s)
Liked: 332
Using dispcalGUI 2.9.0.3, argyllCMS Beta b (03/23/2015) and latest madVR. Profile only

Did a 493 point measurement report. See picture.

ss
Attached Thumbnails
Click image for larger version

Name:	Measurement Report 493 patches.jpg
Views:	45
Size:	92.3 KB
ID:	623497  
sillysally is offline  
post #3073 of 3391 Old 03-25-2015, 02:59 AM
Senior Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 480
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 123
Quote:
Originally Posted by Pepelegal View Post
Ok, "targen" process is still running after 1h16min (CPU 37%).

I did a copy of the folder, but I think will cancel the process as it taking a long time...

Let me know if you need something else
Thanks. The reason targen hangs in your case is because the profile black point is zero, and it actually hangs on all profiles with zero blackpoint when they are used for preconditioning. I switched from MinGW GCC to MS Visual C compiler, and the problem went away. I updated the Argyll 1.7b beta devel build and also re-enabled optimization, so the newest executables (2015-03-25) should be back to roughly the same ballpark computation speed as 1.6.3.
Pepelegal likes this.
fhoech is online now  
post #3074 of 3391 Old 03-25-2015, 03:01 AM
AVS Special Member
 
Brian Hampton's Avatar
 
Join Date: Apr 2000
Posts: 7,271
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 318 Post(s)
Liked: 132
Hope to get to try this process for the first time in a few hours.

I will start by using HCFR to set my projector's greyscale and also ensure my color mode is wider than standard. (I've heard these 2 steps are required to make a good 3D Out,.. though they aren't mentioned.. I guess everyone figures it's known.)

Is there a way to use the correction matrix from HCFR with DisplayCALgui?... I will profile my i1d3 to my i1pro and I see from the first post there is a field to input correction file but maybe that file has to be made with DisplayCal.

(I'm so excited to try this out.. and so happy to think I finally have the required hardware.)

-Brian
Brian Hampton is offline  
post #3075 of 3391 Old 03-25-2015, 03:07 AM
Senior Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 480
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 123
Quote:
Originally Posted by Brian Hampton View Post
I will start by using HCFR to set my projector's greyscale and also ensure my color mode is wider than standard. (I've heard these 2 steps are required to make a good 3D Out,.. though they aren't mentioned.. I guess everyone figures it's known.)
This can help the 3D LUT process. On my own TV, there's not much to adjust, so I usually just make sure that none of the channels clip, and adjust the whitepoint with DCG. It's up to you though, when you're using HCFR anyway it makes sense to adjust whitepoint there.

Quote:
Originally Posted by Brian Hampton View Post
Is there a way to use the correction matrix from HCFR with DisplayCALgui?... I will profile my i1d3 to my i1pro and I see from the first post there is a field to input correction file but maybe that file has to be made with DisplayCal.
HCFR uses the Argyll format for corrections, so it'll work in DCG.
fhoech is online now  
post #3076 of 3391 Old 03-25-2015, 03:18 AM
Senior Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 480
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 123
Quote:
Originally Posted by sillysally View Post
Using dispcalGUI 2.9.0.3, argyllCMS Beta b (03/23/2015) and latest madVR. Profile only

Did a 493 point measurement report. See picture.
Looks good Did you use a custom testchart for the profile or the "auto" mode?
fhoech is online now  
post #3077 of 3391 Old 03-25-2015, 03:41 AM
Member
 
Join Date: Nov 2014
Posts: 105
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 78 Post(s)
Liked: 5
Last night I tried to create a 3D LUT using the previous .cal file I had, but this time the 154 patches didn't run and it went directly to the profile phase. I did set the "Tone Curve" to "as measured" and disabled the "interactive display adjustment".

For the test chart, I edited one of the already available for Rec.709 and added 9000 to the "iterative patches" to create a new test chart with 9268 patterns (if there is a better way to do this please let me know).

it took around 4h00 to measure all patches and to create the 3DLUT.

This morning it seemed it was everything alright : peak err = 1.995696, avg err = 0.362326, RMS = 0.429741

But when I saw the curves, I noted "Tone response" weren't very linear.





And when I played some content I could see some slightly artefacts in some areas. It's nothing like I had my first 3DLUT but it's there.

It may be some setting that I may be using wrong. Last night I used these settings (except I used the I1D3 instead of the I1Pro):






I do have some doubts on how to set "3D LUT" -> "Rendering Intent" as in some tutorials I see the recommendation to use "Relative Colorimetric" and in others the recommendation is to use "Absolute colorimetric with white point scaling".

Can anyone tell me how can I improve this results ?

Thank you,
Attached Thumbnails
Click image for larger version

Name:	Gamut madVR 2015-03-25  (absolute) M-S XYZLUT.png
Views:	126
Size:	63.2 KB
ID:	623785   Click image for larger version

Name:	Tone response curves madVR 2015-03-25  (absolute) M-S XYZLUT.jpg
Views:	126
Size:	27.2 KB
ID:	623793   Click image for larger version

Name:	Calibration curves madVR 2015-03-25  (absolute) M-S XYZLUT.jpg
Views:	126
Size:	24.0 KB
ID:	623801   Click image for larger version

Name:	Screen Shot 2015-03-25 at 10.18.04.png
Views:	125
Size:	152.9 KB
ID:	623809   Click image for larger version

Name:	Screen Shot 2015-03-25 at 10.17.55.png
Views:	125
Size:	139.0 KB
ID:	623817  

Click image for larger version

Name:	Screen Shot 2015-03-25 at 10.17.49.png
Views:	10
Size:	152.1 KB
ID:	623825   Click image for larger version

Name:	Screen Shot 2015-03-25 at 10.17.12.png
Views:	124
Size:	149.9 KB
ID:	623833   Click image for larger version

Name:	Screen Shot 2015-03-25 at 10.36.40.png
Views:	125
Size:	144.2 KB
ID:	623841  
Pepelegal is offline  
post #3078 of 3391 Old 03-25-2015, 03:58 AM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,490
Mentioned: 61 Post(s)
Tagged: 0 Thread(s)
Quoted: 741 Post(s)
Liked: 640
Quote:
Originally Posted by fhoech View Post


HCFR uses the Argyll format for corrections, so it'll work in DCG.
HCFR can read ccmx style files but it does not write them, it uses it's own format from pre-Argyll days. You'll either have to create it in DCG or copy/paste from the HCFR sensor matrix display to an existing ccmx file.
Brian Hampton likes this.
zoyd is online now  
post #3079 of 3391 Old 03-25-2015, 04:03 AM
Senior Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 480
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 123
Quote:
Originally Posted by Pepelegal View Post
Last night I tried to create a 3D LUT using the previous .cal file I had, but this time the 154 patches didn't run and it went directly to the profile phase. I did set the "Tone Curve" to "as measured" and disabled the "interactive display adjustment".
The 154 patches are only used for automatic preconditioning, so unless you've set testchart to "Auto" they won't run.

Quote:
Originally Posted by Pepelegal View Post
For the test chart, I edited one of the already available for Rec.709 and added 9000 to the "iterative patches" to create a new test chart with 9268 patterns (if there is a better way to do this please let me know).
That's fine.

Quote:
Originally Posted by Pepelegal View Post
But when I saw the curves, I noted "Tone response" weren't very linear.
The default plot shows the curves of the PCS->device tables, but those tables are not used for 3D LUTs normally and thus created in lower precision to save some computation time. You have to switch the plot to device <- A2B <- PCS.

Quote:
Originally Posted by Pepelegal View Post
And when I played some content I could see some slightly artefacts in some areas. It's nothing like I had my first 3DLUT but it's there.
Without the 3D LUT file it's hard to comment.

Quote:
Originally Posted by Pepelegal View Post
It may be some setting that I may be using wrong. Last night I used these settings (except I used the I1D3 instead of the I1Pro):
Settings look fine.

Quote:
Originally Posted by Pepelegal View Post
I do have some doubts on how to set "3D LUT" -> "Rendering Intent" as in some tutorials I see the recommendation to use "Relative Colorimetric" and in others the recommendation is to use "Absolute colorimetric with white point scaling".
Generally I'd recommend adjusting the whitepoint using the TV's controls and then use relative colorimetric.
Pepelegal likes this.
fhoech is online now  
post #3080 of 3391 Old 03-25-2015, 04:25 AM
Member
 
Join Date: Nov 2014
Posts: 105
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 78 Post(s)
Liked: 5
Thanks fhoech.

I changed to "PCS<-B2A<-device" and..heck...this seems one of my son drawings

Is this "right" ?


This is the link for my Google Drive Project folder with all the files: LINK

Thank you,
Attached Thumbnails
Click image for larger version

Name:	Gamut madVR 2015-03-25  (absolute) M-S XYZLUT.jpg
Views:	118
Size:	40.6 KB
ID:	623849  
Pepelegal is offline  
post #3081 of 3391 Old 03-25-2015, 04:31 AM
Senior Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 480
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 123
Quote:
Originally Posted by Pepelegal View Post
Thanks fhoech.

I changed to "PCS<-B2A<-device" and..heck...this seems one of my son drawings

Is this "right" ?
Device -> A2B -> PCS would be the right choice. You've selected B2A (PCS -> device), which is the lower precision table that's not actually used.
fhoech is online now  
post #3082 of 3391 Old 03-25-2015, 04:41 AM
Member
 
Join Date: Nov 2014
Posts: 105
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 78 Post(s)
Liked: 5
Quote:
Originally Posted by fhoech View Post
Device -> A2B -> PCS would be the right choice. You've selected B2A (PCS -> device), which is the lower precision table that's not actually used.
Ok, I think this is what you mentioned:

But shouldn't the curve be more linear on the lower left ?
Attached Thumbnails
Click image for larger version

Name:	Screen Shot 2015-03-25 at 11.39.11.png
Views:	128
Size:	270.9 KB
ID:	623857  
Pepelegal is offline  
post #3083 of 3391 Old 03-25-2015, 04:44 AM
Senior Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 480
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 123
Quote:
Originally Posted by Pepelegal View Post
Ok, I think this is what you mentioned:

But shouldn't the curve be more linear on the lower left ?
"Waves"/"bumps" can be a sign that something isn't right. Can you attach the profile?
fhoech is online now  
post #3084 of 3391 Old 03-25-2015, 04:48 AM
Member
 
Join Date: Nov 2014
Posts: 105
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 78 Post(s)
Liked: 5
Quote:
Originally Posted by Pepelegal View Post
This is the link for my Google Drive Project folder with all the files: LINK
I shared my Project Folder on Google Drive. Did you had any problems accessing it ?

I thought the Google Drive would be better but I can make a Zip to attach here if you want.
Pepelegal is offline  
post #3085 of 3391 Old 03-25-2015, 05:44 AM
Senior Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 480
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 123
Quote:
Originally Posted by Pepelegal View Post
I shared my Project Folder on Google Drive. Did you had any problems accessing it ?

I thought the Google Drive would be better but I can make a Zip to attach here if you want.
Didn't see the link initially. Looking at the measurements, they don't seem very stable at the low end as evidenced by the attached grayscale plot (the vertical axis is L*), this could be due to the TV or the meter, or both. Also, I wouldn't recommend so many grayscale steps, a good value is around 50-120 max.
Attached Thumbnails
Click image for larger version

Name:	madVR 2015-03-25  (absolute) M-S XYZLUT onlygray.png
Views:	19
Size:	20.7 KB
ID:	623865  
fhoech is online now  
post #3086 of 3391 Old 03-25-2015, 05:55 AM
Member
 
Join Date: Nov 2014
Posts: 105
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 78 Post(s)
Liked: 5
Quote:
Originally Posted by fhoech View Post
Didn't see the link initially. Looking at the measurements, they don't seem very stable at the low end as evidenced by the attached grayscale plot (the vertical axis is L*), this could be due to the TV or the meter, or both. Also, I wouldn't recommend so many grayscale steps, a good value is around 50-120 max.
Ok fhoech, thanks

Does that mean I need to reduce the number of "neutral patches" on my custom test chart? I currently have it at 140. Should I set to 85 or 100 perhaps?

Also, would it be better if I do a new full calibration + profile before the creation of the 3DLUT (the all full process)? Do you think it may help somehow?....

Thank you once more.

EDIT: This is my custom Test Chart parameters:
Attached Thumbnails
Click image for larger version

Name:	Screen Shot 2015-03-25 at 12.57.05.png
Views:	123
Size:	181.9 KB
ID:	623873  

Last edited by Pepelegal; 03-25-2015 at 05:58 AM.
Pepelegal is offline  
post #3087 of 3391 Old 03-25-2015, 06:02 AM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,490
Mentioned: 61 Post(s)
Tagged: 0 Thread(s)
Quoted: 741 Post(s)
Liked: 640
Quote:
Originally Posted by fhoech View Post
Didn't see the link initially. Looking at the measurements, they don't seem very stable at the low end as evidenced by the attached grayscale plot (the vertical axis is L*), this could be due to the TV or the meter, or both. Also, I wouldn't recommend so many grayscale steps, a good value is around 50-120 max.
If you plot L* vs. stimulus level they look normal to me.

Pepelegal likes this.
zoyd is online now  
post #3088 of 3391 Old 03-25-2015, 06:05 AM
Senior Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 480
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 123
Quote:
Originally Posted by Pepelegal View Post
Ok fhoech, thanks

Does that mean I need to reduce the number of "neutral patches" on my custom test chart? I currently have it at 140. Should I set to 85 or 100 perhaps?
I'd do that. More pressing is the issue of the bumpy display response though.

Quote:
Originally Posted by Pepelegal View Post
Also, would it be better if I do a new full calibration + profile before the creation of the 3DLUT (the all full process)? Do you think it may help somehow?....
Possibly, although the 1D calibration curves look fine.
Pepelegal likes this.
fhoech is online now  
post #3089 of 3391 Old 03-25-2015, 06:14 AM
Senior Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 480
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 196 Post(s)
Liked: 123
Quote:
Originally Posted by zoyd View Post
If you plot L* vs. stimulus level they look normal to me.
I still think some of the measured values are problematic. E.g. increasing RGB, but decreasing lightness:

Levels 12.23 (12%) and 12.95 (13%) have an L* of 6.04 and 5.5, or levels 15.827 (16%) and 16.547 (17%) which have an L* of 10.74 and 9.28 respectively, and there a few other occasions like this. I guess the RGB steps are too small and challenging the display/instrument repeatability(?).
Pepelegal likes this.

Last edited by fhoech; 03-25-2015 at 06:18 AM.
fhoech is online now  
post #3090 of 3391 Old 03-25-2015, 06:28 AM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,490
Mentioned: 61 Post(s)
Tagged: 0 Thread(s)
Quoted: 741 Post(s)
Liked: 640
Quote:
Originally Posted by fhoech View Post
I still think some of the measured values are problematic. E.g. increasing RGB, but decreasing lightness:

Levels 12.23 (12%) and 12.95 (13%) have an L* of 6.04 and 5.5, or levels 15.827 (16%) and 16.547 (17%) which have an L* of 10.74 and 9.28 respectively, and there a few other occasions like this. I guess the RGB steps are too small and challenging the display/instrument repeatability(?).
I agree that you don't need that fine a grid (50 grayscale patches are sufficient) but a peak-to-peak L* error less than 1 is probably fine.
zoyd is online now  
Sponsored Links
Advertisement
 
Reply Display Calibration



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