MadVR - ArgyllCMS - Page 34 - AVS Forum
Forum Jump: 
 36Likes
Reply
 
Thread Tools
post #991 of 2640 Old 11-17-2013, 02:30 PM
Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 192
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 37 Post(s)
Liked: 34
Quote:
Originally Posted by mightyhuhn View Post

did some one know how to fix a broken 3d lut creater tool?

http://abload.de/img/3dlutwutimfp7.png

reinstalling / reseting to default did nothing. i even searched in the registry.
It's not broken, you just have set a device link profile as source, which can't be used as input for colllink, and so the number of options is limited and a madVR 3D LUT cannot be generated in this case (thus the create 3D LUT button is grayed). Just click browse and select a non-device link input profile, the default is Rec709.icm (when restoring defaults the 3D LUT settings are currently deliberately not reset).
Quote:
at the moment the white is way to red when loaded in windows.
Did you use a colorimeter correction (probably white LED)?
fhoech is offline  
Sponsored Links
Advertisement
 
post #992 of 2640 Old 11-17-2013, 02:32 PM
 
MonarchX's Avatar
 
Join Date: Jul 2012
Posts: 669
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
OK, continuous measure does work, but the 0 tab is not changing no matter what.

I don't think that madVR brightness settings works with patterns... It does work in movies, even with 3D LUT file loaded, but patterns are not changing for some reason... or am I measuring the wrong thing?
MonarchX is offline  
post #993 of 2640 Old 11-17-2013, 02:46 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 4,514
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 118 Post(s)
Liked: 324
Of course they work, I wouldn't have said what I said if they didn't. They will update slowly however due to the low luminance. The column values don't update, you have to watch the free measures values in the view window and the digital values in the column next to it. Adjust your brightness at the display, not in madVR.

-----

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
zoyd is online now  
post #994 of 2640 Old 11-17-2013, 02:56 PM
 
MonarchX's Avatar
 
Join Date: Jul 2012
Posts: 669
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
Quote:
Originally Posted by zoyd View Post

Of course they work, I wouldn't have said what I said if they didn't. They will update slowly however due to the low luminance. The column values don't update, you have to watch the free measures values in the view window and the digital values in the column next to it. Adjust your brightness at the display, not in madVR.


Errr... why not adjust in madVR? It has adjustments which do work in movies, but they do not work in madBR pattern generators. It would be a lot easier to adjust madVR settings, so that all non-madVR things like desktop and games would not require changing TV settings...

There is also the fact that with TV settings I was unable to achieve a flawless gamma... It was close, but nowhere near the original, the perfect one... madVR adjustments may resolve that...
MonarchX is offline  
post #995 of 2640 Old 11-17-2013, 03:03 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 616
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 69 Post(s)
Liked: 75
Quote:
Originally Posted by zoyd View Post

personally I would disable dithering in madVR during calibration especially if using a d3. They have a fairly large aperture and require very uniform illumination to read properly.
Hmm. Dithering should be fine with an i1d3 - it has a large aperture and a diffusion screen in the optical path, so it should average any reasonably fine spatial dithering very well. The dithering will help (ArgyllCMS) calibration a lot - you will find that it converges much more quickly and accurately with the MadVR dithering, rather than having to fight the 8 bit quantization. [The native ArgyllCMS patch display will use more than 8 bits if the hardware rendering path supports it. This is often the case with modern graphics cards and VGA, but not with DVI. It may work with display port, and could work with HDMI, but the only way to get some indication is using dispcal -R] The profiling measurements will be slightly more accurate by avoiding quantizing the device values, and the resulting 3dLut output will be rendered more accurately, and be capable of finer control. Overall it should be a big plus if a >8 bit hardware path is not available.
gwgill is offline  
post #996 of 2640 Old 11-17-2013, 03:11 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 4,514
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 118 Post(s)
Liked: 324
Ok, I mistakenly thought the "large blocking" noise that the user was complaining about was due to dithering, and this might be an issue with the d3 and uniformity. Dithering in general should be beneficial to the calibration path.

-----

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
zoyd is online now  
post #997 of 2640 Old 11-17-2013, 03:13 PM
Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 192
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 37 Post(s)
Liked: 34
Quote:
Originally Posted by MonarchX View Post

Errr... why not adjust in madVR? It has adjustments which do work in movies, but they do not work in madBR pattern generators. It would be a lot easier to adjust madVR settings, so that all non-madVR things like desktop and games would not require changing TV settings...

There is also the fact that with TV settings I was unable to achieve a flawless gamma... It was close, but nowhere near the original, the perfect one... madVR adjustments may resolve that...
Previously you indicated that contrast ratio was important to you. In which case, do not use software controls to alter brightness, ever.
fhoech is offline  
post #998 of 2640 Old 11-17-2013, 03:15 PM
Advanced Member
 
mightyhuhn's Avatar
 
Join Date: Nov 2013
Posts: 609
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 253 Post(s)
Liked: 88
Quote:
Originally Posted by fhoech View Post


It's not broken, you just have set a device link profile as source, which can't be used as input for colllink, and so the number of options is limited and a madVR 3D LUT cannot be generated in this case (thus the create 3D LUT button is grayed). Just click browse and select a non-device link input profile, the default is Rec709.icm (when restoring defaults the 3D LUT settings are currently deliberately not reset).

doesn't change anything i just tryed to change it. it was on Rec709.icm befor. 

it doesn't matter what i choice as source profile it will stay at "C:\Users\Administrator\Desktop\PL2210HD 2013-11-16 Rec. 1886 VF-S XYZLUT\PL2210HD 2013-11-16 Rec. 1886 VF-S XYZLUT.icm" (btw i created a 3d lut from this icm)

the second line changes that's all. here a new screen http://abload.de/img/3dlut2bvc2q.png manuel imput doesn't work too

 

Quote:
Did you use a colorimeter correction (probably white LED)?

no, i don't have a clue about corrections so i leaved it at auto (none)

 

i got 2 pantone color manager cxf are they related to this or of any use?

 

Quote:
Ok, I mistakenly thought the "large blocking" noise that the user was complaining about was due to dithering, and this might be an issue with the d3 and uniformity. Dithering in general should be beneficial to the calibration path.

i posted a pic in the madvr forum http://abload.de/img/madtpg01.mp4_snapshot3ef7p.png

 

i can easy see huge dots with my vg248 very clear without lightboost 

 

but i don't think is an good idea to talk but this in 2 threads at the same time.

mightyhuhn is online now  
post #999 of 2640 Old 11-17-2013, 03:23 PM
 
MonarchX's Avatar
 
Join Date: Jul 2012
Posts: 669
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
This is the best I can do. Any attempt go get closer to the reference line results in a complete mess with gamma - it starts to fall down to 2 past 70% IRE.



But my CR is back to its full pathetic glory!

I am happy once more!
MonarchX is offline  
post #1000 of 2640 Old 11-17-2013, 03:23 PM
Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 192
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 37 Post(s)
Liked: 34
Quote:
Originally Posted by mightyhuhn View Post

doesn't change anything i just tryed to change it. it was on Rec709.icm befor. 
it doesn't matter what i choice as source profile it will stay at "C:\Users\Administrator\Desktop\PL2210HD 2013-11-16 Rec. 1886 VF-S XYZLUT\PL2210HD 2013-11-16 Rec. 1886 VF-S XYZLUT.icm" (btw i created a 3d lut from this icm)
the second line changes that's all. here a new screen http://abload.de/img/3dlut2bvc2q.png manuel imput doesn't work too
Can you send/PM me some logs please? (C:\Users\<Your Username>\AppData\dispcalGUI\logs, probably best to zip up the whole folder)
Quote:
Originally Posted by mightyhuhn View Post

no, i don't have a clue about corrections so i leaved it at auto (none)
That's actually reasonable smile.gif I would recommend you to see how you go with one of the X-Rite provided corrections, though (you do have an i1 Display Pro, right?). Go to menu Tools, select "Import colorimeter corrections from other display profiling software...". If i1 Profiler is installed, clicking "Auto" should pick up the files automatically, otherwise you may have to point it at the i1 Profiler setup executable. After they are imported, they should be available under "Correction". The right choice for your screen is probably "WLED AC LG Samsung <WLEDFamily_07Feb11>".
Quote:
Originally Posted by mightyhuhn View Post

i got 2 pantone color manager cxf are they related to this or of any use?
Hmm, I don't think so. These are propably color definitions (afaik CxF is a format for color exchange http://www.colorexchangeformat.com/).
fhoech is offline  
post #1001 of 2640 Old 11-17-2013, 04:41 PM
Advanced Member
 
mightyhuhn's Avatar
 
Join Date: Nov 2013
Posts: 609
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 253 Post(s)
Liked: 88

Quote:

Originally Posted by fhoech View Post


Can you send/PM me some logs please? (C:\Users\<Your Username>\AppData\dispcalGUI\logs, probably best to zip up the whole folder)

done

 

Quote:
That's actually reasonable smile.gif I would recommend you to see how you go with one of the X-Rite provided corrections, though (you do have an i1 Display Pro, right?). Go to menu Tools, select "Import colorimeter corrections from other display profiling software...". If i1 Profiler is installed, clicking "Auto" should pick up the files automatically, otherwise you may have to point it at the i1 Profiler setup executable. After they are imported, they should be available under "Correction". The right choice for your screen is probably "WLED AC LG Samsung <WLEDFamily_07Feb11>".

worked thx i give it a try tomorrow. but this display is still very special. an older lcd is an ccfl right?

mightyhuhn is online now  
post #1002 of 2640 Old 11-17-2013, 05:19 PM
 
MonarchX's Avatar
 
Join Date: Jul 2012
Posts: 669
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
So why does 3D LUT calibration guide suggest to use Auto (None) correction for my i1Display Pro? I have only a CCFL LCD screen, but a correction file is present - I have the latest i1Profiler from X-Rite.
MonarchX is offline  
post #1003 of 2640 Old 11-17-2013, 05:42 PM
Member
 
paulkono's Avatar
 
Join Date: Aug 2013
Posts: 23
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Is there something wrong with the monitor if after repeated calibrations (after 1 hour of warm-up) the resultant LUT doesn't correct the visual color temperature? I have three of the exact same panel and the other two monitors are fully corrected using the exact same procedures outlined in the first post. There's no OSD on either of them so the settings aren't different either.

The panel in question does have a lower uncalibrated visual color temperature compared to my other two panels though. Are there certain limitations to how much the LUT can correct from uncalibrated?
paulkono is offline  
post #1004 of 2640 Old 11-17-2013, 08:14 PM
 
MonarchX's Avatar
 
Join Date: Jul 2012
Posts: 669
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
OK, so I selected madVR Gamma 2.2 on dispcalGUI main menu. Ran the calibration and profiling, for 3D LUT I selected Rec/709_Gamma22. I got exactly the same result as I got with selecting regular Rec.709. I measured with HCFR and it provided 2.2 gamma, but it reduced my CR from 1850:1 to 600:1 again...

My correct 1D LUT settings are:
Contrast - 77
Brightness - 49

For BT.1886 3D LUT, to achieve the same black levels as with regular 1D LUT, I had to set this:
Brightness - 47
Contrast - 82

For Rec.709 I had to set
Brightness - 37/38
Contrast - ???

I spent an hour trying to get a correct gamma with contrast and gamma settings. NOTHING came even close to a straight line. I tried every damn combination.... It is not working.
Also, this time around, dE for many saturation sweeps were higher and ColorChecker dE went as high as 4.0. For this run I actually improved my 1D LUT color space somewhat...

What is going on?

Why does it reduce black levels to begin with?
MonarchX is offline  
post #1005 of 2640 Old 11-17-2013, 09:44 PM
 
MonarchX's Avatar
 
Join Date: Jul 2012
Posts: 669
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
I got it to work with sRGB profile!
MonarchX is offline  
post #1006 of 2640 Old 11-17-2013, 10:12 PM
 
MonarchX's Avatar
 
Join Date: Jul 2012
Posts: 669
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
There is something sketchy going on.... Regardless of the profile (BT.1886 or 2.2) or settings, there this weird unnatural gray thing going on, I can't explain it, its this smeared/smudges gray kind of crush.... Am I the only one getting it? Is it possible to fix?


And again, why does this software set make it so that I have to increases contrast and reduce brightness to keep my contrast ratio and the deepest blacks? Maybe this is why I get this gray kind of crush BUT I tripple-checked that I did NOT crush my blacks at all!
MonarchX is offline  
post #1007 of 2640 Old 11-18-2013, 05:27 AM
Member
 
Kukulcan's Avatar
 
Join Date: Sep 2012
Location: Italy
Posts: 117
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 20
Quote:
Originally Posted by gwgill View Post

By normal display standards we are talking about a very, very small error - the above it is 0.004 cd/m^2, way below the expected accuracy of the instrument. Compared to a full scale of (say) 100 cd/m^2, that's an inaccuracy of 0.004%! It's miraculous that it is so small. I'm not clear under what circumstances it will be visible, other than observing the display in pitch dark with a black raster. Or am I underestimating how dark some scenes can get ?

Hi Graeme,

I can say that with a projector, in a darkened environment, you can perceive (and appreciate!!) a 0,004 cd/m2 lower black level in some very dark scenes, but that's not the crucial point! BT1886 processing, making calculations based on that higher black level, gives pretty visible increases in the first 10 IRE levels! The undue brightning of these first IRE levels is very noticeable, and of bigger amplitude than 0.004 cd/m2
Quote:
Originally Posted by gwgill View Post


Forcing the black to map device 0,0,0 -> 0,0,0 makes the assumption that device 0,0,0 really is the device black point. I've seen a number of examples where this isn't true

Please could you be so kind to explain me this? If 0,0,0 is NOT the device black point, which one could it be? Thanks!
Kukulcan is online now  
post #1008 of 2640 Old 11-18-2013, 07:27 AM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 4,514
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 118 Post(s)
Liked: 324
I think one could make the assumption that the user can more accurately map a black input to the displays minimum Luminance than the model can. Under that assumption you could have the code smoothly return the calibration or 3dlut mapping to input values, say below an input level that measures at or just above minimum display luminance. I post process my 3dluts in this fashion and have not observed or measured any bad side effects.

-----

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
zoyd is online now  
post #1009 of 2640 Old 11-18-2013, 07:41 AM
 
MonarchX's Avatar
 
Join Date: Jul 2012
Posts: 669
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
I just re-ran BT.1886 on top of a more accurate original hardware calibration and it is even better than before! There are no black/gray issues or artifacts. The black is the lowest possible. All is in order.

So, why am I getting such bad results with sRGB and Gamma 2.2 files? They require me to set my brightness WAY down (much more than BT.1886 calibration) to reach the lowest black and at that level blacks are crushed!

I also get these strange square artifacts in some of my videos that do not occur with my hardware-set calibration. I do not install the created .icc/.icm profiles, just make 3D LUT. Overall, even if I adjust my brightness and contrast to reach the same lowest black as my own calibration, the blacks seem off, the grays also seem off... even though HCFR readings are near-perfect.

I saw zoyd had similar issues with several builds, so I guess this is more or less common and will be fixed in the future? I am going through the whole thread and see a lot of people say the same thing. These issues with blacks and gray only occur with sRGB gamma 2.2, Rec.709 Gamma 2.2. Maybe I should try making a profile with black compensation enabled?

I just wonder why can't they simply FORCE the system to use an already-set black and white points. Maybe it will result in a less accurate profile, but getting blacks right is very important!
MonarchX is offline  
post #1010 of 2640 Old 11-18-2013, 02:16 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 616
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 69 Post(s)
Liked: 75
Quote:
Originally Posted by Kukulcan View Post

Quote:
Originally Posted by gwgill View Post


Forcing the black to map device 0,0,0 -> 0,0,0 makes the assumption that device 0,0,0 really is the device black point. I've seen a number of examples where this isn't true

Please could you be so kind to explain me this? If 0,0,0 is NOT the device black point, which one could it be? Thanks!

The whole point of calibration/profiling is not to trust that the device is set to behave in any particular way. So it's quite possible (and it does happen) for a display to have a "dead band" from black. So 0 -> black, 10-> black, 20 -> black, 21->first non-black. Assuming black starts at 0 and feeding signal into it will crush your blacks, and distort the response up to the first interpolation grid point.

No I don't know how common this is with particular display technologies. CRT - very common. LCD ? Plasma ? I'm not so sure. But the point is to ensure a progressive response from black, so the black point has to be determined, not assumed.
gwgill is offline  
post #1011 of 2640 Old 11-18-2013, 02:33 PM
Newbie
 
bmw525i's Avatar
 
Join Date: Nov 2013
Location: Hermaringen, BW, Germany
Posts: 13
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10

Hi,

 

has anyone figured out how to create a madVR LUT with a straight gamma of 2.2 (or any other number) yet?

No matter what gamma values or standards I set for measurements in dispcalGUI, the 3DLUT-maker always outputs some kind of curved gamma, looking like that bt1886 thing.

 

The problem is, that movies shown with my projector look a bit flat in scenes, where there are no real dark or bright areas. I suppose that's due to the reduced gamma at low luminance - providing more shadow-detail of course - but i would rather like that scenes with more contrast than with more shadow-detail. I just don't know how to achieve that ^^

 

Thanks for any ideas.

bmw525i is offline  
post #1012 of 2640 Old 11-18-2013, 02:41 PM
 
MonarchX's Avatar
 
Join Date: Jul 2012
Posts: 669
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
Quote:
Originally Posted by bmw525i View Post

Hi,

has anyone figured out how to create a madVR LUT with a straight gamma of 2.2 (or any other number) yet?
No matter what gamma values or standards I set for measurements in dispcalGUI, the 3DLUT-maker always outputs some kind of curved gamma, looking like that bt1886 thing.

The problem is, that movies shown with my projector look a bit flat in scenes, where there are no real dark or bright areas. I suppose that's due to the reduced gamma at low luminance - providing more shadow-detail of course - but i would rather like that scenes with more contrast than with more shadow-detail. I just don't know how to achieve that ^^

Thanks for any ideas.

Just don't use Rec.709 as the source. Use madVR settings, select sRGB gamma or simple Gamma 2.2. Then, if you used sRGB, select sRGB for source target. If you used Gamma 2.2, then select Rec.709_Gamma22 for source target. You will get your straight gamma that way. However, it ended up killing my contrast both times and I was unable to reduce brightness and set contrast to something that would preserve both - contrast ratio and the calibration accuracy.
MonarchX is offline  
post #1013 of 2640 Old 11-18-2013, 02:52 PM
Newbie
 
bmw525i's Avatar
 
Join Date: Nov 2013
Location: Hermaringen, BW, Germany
Posts: 13
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by MonarchX View Post


Just don't use Rec.709 as the source. Use madVR settings, select sRGB gamma or simple Gamma 2.2. Then, if you used sRGB, select sRGB for source target. If you used Gamma 2.2, then select Rec.709_Gamma22 for source target. You will get your straight gamma that way. However, it ended up killing my contrast both times and I was unable to reduce brightness and set contrast to something that would preserve both - contrast ratio and the calibration accuracy.


Hmm, I think that's because the source profile should be the one in which the video is encoded, that you want to play back.

As far as I understood that whole LUT thing as implemented in Argyll, you always have to specify exactly the two colorspaces between you want to transform. So the output should always be your monitor-profile of course and the source is whatever you have as source, like sRGB for web-/camera-images, or AdobeRBG for images with extended colorspace and so on.

 

Thanks for the tip though, but with that reduced contrast it just doesn't seem like the "right" way to do it ^^

bmw525i is offline  
post #1014 of 2640 Old 11-18-2013, 04:31 PM
Member
 
VerGreeneyes's Avatar
 
Join Date: May 2013
Posts: 112
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
One thing that might be worth trying is to use a matrix profile instead of a LUT profile. This will likely be less accurate in some ways, but if the measurements are unstable enough a LUT can cause some visually obvious outliers that mess up the black point (and dark shades near black).

By the way if you feel like experimenting, I have a few more source profiles you can try here. The number in each file name indicates which power function was used for the gamma curves. The ones with no number in the name use their default (to specification) transfer functions, except Rec709-sRGB.icm which uses the sRGB transfer function. inverted.icm simply inverts all colours so you can easily test if an application is using the installed profile (I don't know what it would do with collink).
VerGreeneyes is offline  
post #1015 of 2640 Old 11-18-2013, 04:57 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 616
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 69 Post(s)
Liked: 75
I've released ArgyllCMS V1.6.2, a bug fix and feature release.
See http://www.freelists.org/post/argyllcms/ArgyllCMS-V162-has-been-released
gwgill is offline  
post #1016 of 2640 Old 11-18-2013, 05:00 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 616
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 69 Post(s)
Liked: 75
Quote:
Originally Posted by zoyd View Post

I think one could make the assumption that the user can more accurately map a black input to the displays minimum Luminance than the model can. Under that assumption you could have the code smoothly return the calibration or 3dlut mapping to input values, say below an input level that measures at or just above minimum display luminance. I post process my 3dluts in this fashion and have not observed or measured any bad side effects.
That would be a last resort. Instead, why don't you try out the new "targen -V" parameter in ArgyllCMS V1.6.2 to see if this reduces the black point error.
See http://www.argyllcms.com/doc/targen.html#V
gwgill is offline  
post #1017 of 2640 Old 11-18-2013, 06:03 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 4,514
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 118 Post(s)
Liked: 324
Quote:
Originally Posted by gwgill View Post

That would be a last resort. Instead, why don't you try out the new "targen -V" parameter in ArgyllCMS V1.6.2 to see if this reduces the black point error.
See http://www.argyllcms.com/doc/targen.html#V

Definitely will try it out, I know you favor the "weighting the important bits" approach so let's see how it goes. I noticed also -N for neutral axis weighting. The recommend parameter value for -V is 1.5 - 2.0, what is the effect on the distribution of patches and how colprof treats them?

-----

To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.


To view links or images in signatures your post count must be 0 or greater. You currently have 0 posts.
zoyd is online now  
post #1018 of 2640 Old 11-18-2013, 06:15 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 616
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 69 Post(s)
Liked: 75
Quote:
Originally Posted by zoyd View Post

The recommend parameter value for -V is 1.5 - 2.0, what is the effect on the distribution of patches and how colprof treats them?
I haven't done a great deal of testing yet, so feedback is welcome.
You can observe the effect by looking at the device point distribution using "targen -W".
The complementary flag in colprof distributes the profile cLUT grid points in a similar fashion.
gwgill is offline  
post #1019 of 2640 Old 11-18-2013, 06:32 PM
Member
 
VerGreeneyes's Avatar
 
Join Date: May 2013
Posts: 112
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by gwgill View Post

That would be a last resort. Instead, why don't you try out the new "targen -V" parameter in ArgyllCMS V1.6.2 to see if this reduces the black point error.
Quote:
This emphasis is only effective for perceptual patch distributions where a pre-conditioning profile is provided.
Does the parameter do anything if you don't provide such a profile? Regardless, it's interesting that this seems to turn profiling into more of a two-pass process - one pass to generate the pre-conditioning profile, then another pass with a patch distribution that takes this into account.
VerGreeneyes is offline  
post #1020 of 2640 Old 11-18-2013, 07:48 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 616
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 69 Post(s)
Liked: 75
Quote:
Originally Posted by VerGreeneyes View Post

Does the parameter do anything if you don't provide such a profile? Regardless, it's interesting that this seems to turn profiling into more of a two-pass process - one pass to generate the pre-conditioning profile, then another pass with a patch distribution that takes this into account.
No it does nothing without a pre-conditioning profile. As noted in the tutorial, for Video use it's quite reasonable to use the Rec709.icm (etc.) for a first/only pre-conditioning profile, since TV's don't have the range of behaviors that other devices like printers have.
gwgill is offline  
Reply Display Calibration

User Tag List

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