MadVR - ArgyllCMS - Page 70 - AVS Forum
Forum Jump: 
 15Likes
Reply
 
Thread Tools
post #2071 of 2437 Old 06-09-2014, 05:58 AM
Member
 
TimHamburg's Avatar
 
Join Date: Oct 2013
Posts: 46
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 11
Quote:
Originally Posted by Stereodude View Post

Any ideas why the 3D LUT I attempted on my RP-CRT is raising the black levels? I did not enable black point correction.

Here are some pictures with the camera's exposure locked.

With no 3D LUT levels below 16 are not visible. (I realize you can't really see the bars over 16 in this picture, but they are there)



3D LUT with no gamma/tone correction:



3D LUT with 1886 absolute 2.4:



3D LUT with 1886 relative 2.4:



HCFR measures the contrast of the display around 15,000:1. 100% white is right about 120cd/m^2. I followed the procedure in the first post except I did not precondition the tone curve to 1886 absolute 2.4. I left it at Rec.709 because I was not intending to apply any gamma/tone correction in the 3D LUT, but rather rely on the displays natural gamma. Also, the white level did not change pre/post 3D LUT. It displays up to 235 and stops. FWIW, this is not my first 3D LUT. I've made several before for a LCD with a ~3,000:1 contrast ratio and it did not raise the black level or start displaying bars that weren't visible before.

Hi Stereodude,

I have similar problems right now.

In earlier 3dluts (created with this procedure)

I didn't have any problems (16 is black, 235 is white).

My 0-255 chain is OK (if no 3dlut or older 3dlut applied --> no probs).

 

Additionally if I project 0%black patch through MadTPG (clicking on 0% in HCFR and doing a continuous reading),

the patch is visible brighter than the black background.

Could you check if that's the same with your 3dlut?

 

Thanks.

TimHamburg is offline  
Sponsored Links
Advertisement
 
post #2072 of 2437 Old 06-09-2014, 07:01 AM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 4,427
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 75 Post(s)
Liked: 299
The video card calibration step has led to some black point shifts in the past, try just the 3dLUT procedure (profile only button),
zoyd is online now  
post #2073 of 2437 Old 06-09-2014, 07:03 AM
Member
 
TimHamburg's Avatar
 
Join Date: Oct 2013
Posts: 46
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 11
Quote:
Originally Posted by zoyd View Post

The video card calibration step has led to some black point shifts in the past, try just the 3dLUT procedure (profile only button),


Thanks. I'll try that.

TimHamburg is offline  
post #2074 of 2437 Old 06-09-2014, 07:12 AM
Member
 
Francois76l's Avatar
 
Join Date: May 2014
Posts: 57
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10

For both (TimHamburg & Stereodude), have you added your colormeter correction in DispCalGUI?

Francois76l is offline  
post #2075 of 2437 Old 06-09-2014, 07:38 AM
Member
 
TimHamburg's Avatar
 
Join Date: Oct 2013
Posts: 46
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 11
Quote:
Originally Posted by Francois76l View Post
 

For both (TimHamburg & Stereodude), have you added your colormeter correction in DispCalGUI?


tonight...

TimHamburg is offline  
post #2076 of 2437 Old 06-09-2014, 08:31 AM
Member
 
Icaro's Avatar
 
Join Date: Jul 2007
Location: Italy
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 15
Quote:
Originally Posted by Francois76l View Post

For both (TimHamburg & Stereodude), have you added your colormeter correction in DispCalGUI?
Quote:
Originally Posted by TimHamburg View Post


tonight...

Also in HCFR for linearize white 100%,
used the same matrix correction or spectral sample in HCFR and DispcalGui

Sorry for my bad English
Icaro is offline  
post #2077 of 2437 Old 06-09-2014, 09:16 AM
AVS Special Member
 
Stereodude's Avatar
 
Join Date: Jan 2002
Location: Detroit Metro Area
Posts: 9,615
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 297 Post(s)
Liked: 379
Quote:
Originally Posted by Francois76l View Post

For both (TimHamburg & Stereodude), have you added your colormeter correction in DispCalGUI?
What do you mean by correction? I did not pick anything since none of the canned spectral responses are suitable to a RP-CRT.
Stereodude is online now  
post #2078 of 2437 Old 06-09-2014, 09:21 AM
AVS Special Member
 
Stereodude's Avatar
 
Join Date: Jan 2002
Location: Detroit Metro Area
Posts: 9,615
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 297 Post(s)
Liked: 379
Quote:
Originally Posted by TimHamburg View Post

Hi Stereodude,
I have similar problems right now.
In earlier 3dluts (created with this procedure)
I didn't have any problems (16 is black, 235 is white).
My 0-255 chain is OK (if no 3dlut or older 3dlut applied --> no probs).
I've used v1.7.5.7 DispCalGUI before, this is my first attempt with 2.1.x.x. I used 2.1.0.0 to create the profile/calibration, I created the 3D LUTs with 2.1.0.6. Likewise my 0-255 chain is fine with no 3D LUT. I haven't tried the 3D LUT I created with 1.7.5.7 (for another monitor).
Quote:
Additionally if I project 0%black patch through MadTPG (clicking on 0% in HCFR and doing a continuous reading),
the patch is visible brighter than the black background.
Could you check if that's the same with your 3dlut?
Yes, "black" through madVR (0) is visibly brighter than black from the Windows GUI.
Stereodude is online now  
post #2079 of 2437 Old 06-09-2014, 12:38 PM
Advanced Member
 
Iron Mike's Avatar
 
Join Date: Oct 2006
Posts: 775
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 11 Post(s)
Liked: 349
Quote:
Originally Posted by TimHamburg View Post

Thanks again!
That's great info!
I hope some of the others might also profit from it...
One thing I learned in calibrating my projector is:
No matter what the calibration data says and how perfect
the graphs look,
if the movie picture looks strange, the calibration counts for nothing.
Doubly so, if your colorimeter is giving you inacurate readings.
Trust your eyes and not just the calibration data!

yes, that is why Pros always do a visual evaluation of the calibration, no matter how fantastic the dE results are

calibration & profiling solutions: Lightspace, Spaceman, Calman, Argyll, ColorNavigator, basICColor
profiling & calibration workflow tools: Display Calibration Tools
meter: Klein K-10 A, i1Pro, i1D3
AVS thread: Lightspace & Custom Color Patch Set & Gamma Calibration on Panasonic 65VT60
Iron Mike is offline  
post #2080 of 2437 Old 06-09-2014, 02:48 PM
Member
 
Icaro's Avatar
 
Join Date: Jul 2007
Location: Italy
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 15
How I promise,
I took some pictures live at the projected image, (a projection screen 250"x170")

unfortunately I have to apologize in advance for the low quality,
and especially with the greyscale do not reflect in any way what I looks on the screen in real life.
the picture were captured with a phone Samsung Galaxy S4 (I do not have a camera),
but unfortunately immediately puts the phone in front of the projection screen,
through the display of the phone the picture changes,
with the greyscale can not be displayed the steps to bottom scale close to level 16,
and also displays a colorization on some steps,
I made ​​a lot of shots to be able to capture the real picture, but it was impossible,
the phone changes the image probably because of the display or sensor.
and this is the maximum I was able to obtain

But what I looks on the screen in real life,
it is another thing, with all steps well visible from 16 at 235,
without colorization but only Black and White at the different nuances for each step,
none black crush or ruins to bottom scale, and none clipping on the highlights.

in sincerity I was very undecided if show or not, but for fairness this is what.
However, for what they are worth, here they are












Sorry for my bad English
Icaro is offline  
post #2081 of 2437 Old 06-09-2014, 03:30 PM
Senior Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 278
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 21

@Icaro I notice your 2.2 gamma settings use profile type = XYZ LUT + matrix as opposed to the default of XYZ LUT + swapped matrix, why is that?

3ll3d00d is online now  
post #2082 of 2437 Old 06-09-2014, 04:55 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 545
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 68
Quote:
Originally Posted by Francois76l View Post

Ok Icaro, thanks for the information!
I will try it tonight. I will start with the default preset 6000K (which is close to real 6500K) and I will tune it to have a linear 6500K on ColorHCFR.
Then I will launch the autocal.
Is it ok?
Again 3h30 of autocal wink.gif
It's hard to know what you are referring to - nothing in ArgyllCMS is called "autocal" - perhaps you are in the wrong forum ??
gwgill is offline  
post #2083 of 2437 Old 06-09-2014, 05:40 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 545
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 68
Quote:
Originally Posted by Francois76l View Post

And in DispalGui I let LCD (generic) for Display? And for a projector?
It depends on what type of projector it is. An LCD type projector will be best with an LCD type matrix, which is non-refresh. A DLP projector will probably have a different calibration matrix, and will be refresh type (I think the projector that X-Rite includes in their .edr files is a DLP).
gwgill is offline  
post #2084 of 2437 Old 06-09-2014, 05:54 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 545
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 68
Quote:
Originally Posted by Francois76l View Post

Finally what is the differrence between the ccmx and the ccss?

.ccmx is a matrix calibration for a particular instrument to a particular display. It probably won't work so well for a different (but same make and model) instrument.

.ccss is a spectral sample of a particular display. It will work well with different instruments that support it (i1d3 or Spyder 4).

In both cases the instrument calibration may, or may not work so well to a different (but same make and model) display, or a same technology display, depending on how similar the spectral response of the display is compared to what the calibration was setup for.

A .ccmx created for a particular instrument on a particular display will likely be more accurate than a .ccss, but a .ccss will likely be more accurate than a .ccmx that was created for a different instrument.
gwgill is offline  
post #2085 of 2437 Old 06-09-2014, 05:56 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 545
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 68
Quote:
Originally Posted by Francois76l View Post

So the CCMX file is correct for all types of displays
CCSS is correct for a particular display

Well, no. See my previous post on .ccmx vs. .ccss.
gwgill is offline  
post #2086 of 2437 Old 06-09-2014, 06:04 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 545
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 68
Quote:
Originally Posted by xvfx View Post

Is there any side effects having this installed or using HCFR on it's own calibrating the TV externally? As I recall reading one time long ago Argyll caused issues? You had to have it uninstalled to use one or the other?
There are some tradeoffs if you are mixing desktop color managed apps with MadVR, but we setup MadTPG and the 3DLut to automatically let MadVR "do the right thing". It will flip to (any) 3DLut specific Video card Lut calibration curves when video is being displayed viad MadVR/MadTPG, and then flip back again when you have finished it to any calibration curves in the currently installed ICC profile (vcgt tag).
gwgill is offline  
post #2087 of 2437 Old 06-09-2014, 06:08 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 545
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 68
Quote:
Originally Posted by Holiday121 View Post

That's odd because from what I was reading I had to do the calibration first with hcfr.

There are several optional variations. The core part is profiling your display and then creating a device link/3DLut between the video colorspace you want the setup to emulate, and your display.

Optional parts are tweaking the controls of the display to put it closer to your target, and/or creating per channel calibration curves that can be implemented by the graphics card hardware, or (alternately) be incorporated in the 3DLut. The latter can be used to shift the white point and/or improve grey balance. These steps would be done before profiling the display.
gwgill is offline  
post #2088 of 2437 Old 06-09-2014, 07:29 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 545
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 68
Quote:
Originally Posted by Icaro View Post

the real problem is the instability of the UHP lamp of the projetor,
at least on the old series "HD" as my HD990, there is a substantial and visible deviation of the brightness,
also during the normal vision of a movie.
So I think, that for this reason, every 40 measurements, the process of autocal must recalculate all around these variations
(and in fact the process of autocal runs all around the white level)
Nothing in ArgyllCMS is called "Autocal", so I can't figure out which step you are talking about.

Is this the display calibration step, or the display profiling step ?

With some more specific information, it may be possible to figure out what's going on.
gwgill is offline  
post #2089 of 2437 Old 06-09-2014, 08:46 PM
Senior Member
 
Audionut11's Avatar
 
Join Date: May 2008
Posts: 347
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 70 Post(s)
Liked: 52
Quote:
Originally Posted by gwgill View Post

Optional parts are tweaking the controls of the display to put it closer to your target

One thing I am not clear on, is how to perform this display adjustment with argyll/dispcalGUI? There are plenty of other applications I can use, but I cannot get a consistent reading between applications.

I've tried both ChromaPure and HCFR. And despite creating my own meter calibration files, I always have a minimum of 3dE between these applications and the interactive display adjustment with dispcalGUI.


edit: Ok, there was an error in my meter calibration with argll. I have fixed that, but would still appreciate instructions for tweaking the display controls with argll/dispcalGUI.
Audionut11 is offline  
post #2090 of 2437 Old 06-09-2014, 10:22 PM
Member
 
Icaro's Avatar
 
Join Date: Jul 2007
Location: Italy
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 15
Quote:
Originally Posted by gwgill View Post

Nothing in ArgyllCMS is called "Autocal",......

I apologize,
for "Autocal" I mean the autocalibration procedures,
through ArgyllCMS+Dispcalgui
Quote:
Originally Posted by gwgill View Post

....so I can't figure out which step you are talking about.
Is this the display calibration step, or the display profiling step ?
With some more specific information, it may be possible to figure out what's going on.

The step is the second,
when they start all the 3778 patches.

This is my settings for the calibration and profiling of a projector JVC HD990 means a i1D3

http://www.avsforum.com/t/1471169/madvr-argyllcms/2010#post_24805613

I need to change something ?

Sorry for my bad English
Icaro is offline  
post #2091 of 2437 Old 06-09-2014, 10:27 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 545
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 68
Quote:
Originally Posted by ST60 View Post

Hi! Has there been any progress on new ArgylCMS updates? The issue of WTW not being properly clipped in 3DLUTs is still there.

It was reported by madshi and Asmodean here - http://www.avsforum.com/t/1471169/madvr-argyllcms/1530#post_24432596   .

I've looked into this, but don't see that there is a problem. The 3DLut is behaving as has been suggested when handling WTW - it extrapolates as far as it can, and then (naturally) clips to the sync level signals, doing a pass-through on them.

The response is (as far as I can tell) monotonic and proportional. Madshi's example doesn't show a problem, since he's overlooked the fact that white maps to 235 228 212, and that the out of gamut input does map to an out of gamut output :-

xicclu -s 255 test_a.icm
235.000000 235.000000 235.000000 [RGB] -> Lut -> 235.317180 227.900700 212.305930 [RGB]
240.000000 240.000000 240.000000 [RGB] -> Lut -> 240.001105 233.313260 217.144033 [RGB]

Since the R channel output is > 235, the color is clearly out of gamut (remember, color is a 3 dimensional coordinate). The proportions of RGB remain similar to the white point, as was requested, to allow (so I'm told) WTW content to be color managed.

If you want this changed to something else then I need to hear a cohesive rational as to how to deal with out of gamut input through the 3DLut that takes account of whether video content actually contains WTW or not, and how various displays deal with being fed WTW.

If some displays clip on a per component basis rather than displaying WTW or clipping in a way that preserves hue, then yes, feeding color managed WTW values to the display is probably a bad thing. But that's not what I was told was the situation. Changing it so that the 3DLut clips the input value is an obvoius alternate to the current behavior, but this is certainly not passing WTW through.
gwgill is offline  
post #2092 of 2437 Old 06-09-2014, 10:38 PM
Member
 
Icaro's Avatar
 
Join Date: Jul 2007
Location: Italy
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 15
Quote:
Originally Posted by 3ll3d00d View Post

@Icaro
 I notice your 2.2 gamma settings use profile type = XYZ LUT + matrix as opposed to the default of XYZ LUT + swapped matrix, why is that?


indeed I do not remember perfectly the reason for this choice,
I did numerous tests, and at the end I chose these settings,

but I posted time ago, the same settings,
wondering to the more experienced, if they were fine for Gamma 2.2 pure law,
unfortunately there was no answer,
but faces the results, I continued to use them.

however I take this opportunity,
to renew the question:
these settings, are fine for a Gamma 2.2 pure law ?
for the calibration and profiling of a projector JVC HD990 means a i1D3

http://www.avsforum.com/t/1471169/madvr-argyllcms/2010#post_24805613

or do I need to change something ?

Thanks

Sorry for my bad English
Icaro is offline  
post #2093 of 2437 Old 06-09-2014, 10:39 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 545
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 68
Quote:
Originally Posted by Stereodude View Post

I did not pick anything since none of the canned spectral responses are suitable to a RP-CRT.
So the CRT selection isn't suitable ?
gwgill is offline  
post #2094 of 2437 Old 06-09-2014, 10:46 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 545
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 68
Quote:
Originally Posted by Audionut11 View Post


One thing I am not clear on, is how to perform this display adjustment with argyll/dispcalGUI? There are plenty of other applications I can use, but I cannot get a consistent reading between applications.

I've tried both ChromaPure and HCFR. And despite creating my own meter calibration files, I always have a minimum of 3dE between these applications and the interactive display adjustment with dispcalGUI.
Whether they are exactly consistent is somewhat irrelevant - profiling trumps all that. I would imagine though, that they are different because they are aiming for slightly different things.
Quote:
I have fixed that, but would still appreciate instructions for tweaking the display controls with argll/dispcalGUI.
The display calibration step is somewhat limited, because it is working with the limitations of typical computer displays. The biggest gain would probably be from making sure the display white point is near the target. Per channel calibration curves may help the accuracy of the final 3DLut, but if the display response curve and grey balance is good to start with, is probably unnecessary.
gwgill is offline  
post #2095 of 2437 Old 06-09-2014, 10:50 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 545
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 68
Quote:
Originally Posted by Icaro View Post

The step is the second,
when they start all the 3778 patches.
Hmm. - well that's a puzzle then. The white patches may be being read a bit too often during calibration with the current code, but not when reading a large number of patches in one hit.

I'm unable to reproduce the problem myself.

If someone who has this problem is comfortable running the ArgyllCMS tools on the command line, then I can make a debug version of dispread available to help track this down.
gwgill is offline  
post #2096 of 2437 Old 06-09-2014, 11:16 PM
AVS Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 5,414
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 16 Post(s)
Liked: 108
Quote:
Originally Posted by gwgill View Post

I've looked into this, but don't see that there is a problem. The 3DLut is behaving as has been suggested when handling WTW - it extrapolates as far as it can, and then (naturally) clips to the sync level signals, doing a pass-through on them.

The response is (as far as I can tell) monotonic and proportional. Madshi's example doesn't show a problem, since he's overlooked the fact that white maps to 235 228 212, and that the out of gamut input does map to an out of gamut output :-

xicclu -s 255 test_a.icm
235.000000 235.000000 235.000000 [RGB] -> Lut -> 235.317180 227.900700 212.305930 [RGB]
240.000000 240.000000 240.000000 [RGB] -> Lut -> 240.001105 233.313260 217.144033 [RGB]

Since the R channel output is > 235, the color is clearly out of gamut (remember, color is a 3 dimensional coordinate). The proportions of RGB remain similar to the white point, as was requested, to allow (so I'm told) WTW content to be color managed.

If you want this changed to something else then I need to hear a cohesive rational as to how to deal with out of gamut input through the 3DLut that takes account of whether video content actually contains WTW or not, and how various displays deal with being fed WTW.

If some displays clip on a per component basis rather than displaying WTW or clipping in a way that preserves hue, then yes, feeding color managed WTW values to the display is probably a bad thing. But that's not what I was told was the situation. Changing it so that the 3DLut clips the input value is an obvoius alternate to the current behavior, but this is certainly not passing WTW through.

I suppose not every display will do the same thing. How about adding an option (if it's not already there) which simply clips every channel to black/white? You could clip the input to 16-235. Or the output to the calibrated black/white point. The end result would be somewhat identical to using a 0-255 levels 3dlut, where BTB/WTW are simply clipped. This clipping should be optional, not enabled by default.
madshi is online now  
post #2097 of 2437 Old 06-10-2014, 12:08 AM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 545
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 13 Post(s)
Liked: 68
Quote:
Originally Posted by madshi View Post

I suppose not every display will do the same thing. How about adding an option (if it's not already there) which simply clips every channel to black/white? You could clip the input to 16-235. Or the output to the calibrated black/white point. The end result would be somewhat identical to using a 0-255 levels 3dlut, where BTB/WTW are simply clipped. This clipping should be optional, not enabled by default.
From a hardware implementation point of view, I'm thinking that per-component clipping is highly likely in many TV's, so I'm inclined to change the code to just clip everything at the input.

I really don't want more options - every option is another possible wrong combination. Those with displays that display WTW will have to speak up, and explain how that can possibly work without per-component clipping causing colored highlights.
gwgill is offline  
post #2098 of 2437 Old 06-10-2014, 12:34 AM
Senior Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 278
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 21
Quote:
Originally Posted by Icaro View Post


indeed I do not remember perfectly the reason for this choice,
I did numerous tests, and at the end I chose these settings,

but I posted time ago, the same settings,
wondering to the more experienced, if they were fine for Gamma 2.2 pure law,
unfortunately there was no answer,
but faces the results, I continued to use them.

however I take this opportunity,
to renew the question:
these settings, are fine for a Gamma 2.2 pure law ?
for the calibration and profiling of a projector JVC HD990 means a i1D3

http://www.avsforum.com/t/1471169/madvr-argyllcms/2010#post_24805613

or do I need to change something ?

Thanks

OK well I tried with exactly your settings, here is a measurement report of the results. I'm not sure how to read the gamma one in particular, it seems quite different to that reported by chromapure.

 

 

I might try without black point compensation tonight.

3ll3d00d is online now  
post #2099 of 2437 Old 06-10-2014, 12:43 AM
AVS Special Member
 
madshi's Avatar
 
Join Date: May 2005
Posts: 5,414
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 16 Post(s)
Liked: 108
Quote:
Originally Posted by gwgill View Post

From a hardware implementation point of view, I'm thinking that per-component clipping is highly likely in many TV's, so I'm inclined to change the code to just clip everything at the input.

I really don't want more options - every option is another possible wrong combination. Those with displays that display WTW will have to speak up, and explain how that can possibly work without per-component clipping causing colored highlights.

I suppose some displays might not clip until the very last processing stage. Which means that feeding them WTW/BTB might help get some more headroom in earlier processing stages (e.g. converting to YCbCr, deinterlacing, scaling, motion interpolation etc etc).

If you really very much want to avoid adding another option, I have 2 ideas/suggestions:

(1) You already have a switch to decide whether the input is 0-255 or 16-235. Maybe you can extend this option to allow (a) 0-255 (b) 16-235 with clipping (c) 16-235 without clipping? This way there would not be a new option, just one more value for one of the already existing options.
(2) I could offer an option in madVR that clips the 3dlut input to 16-235. That way you'd not need to change anything in ArgyllCMS. However, this would only work for madVR, not for eeColor box users...
madshi is online now  
post #2100 of 2437 Old 06-10-2014, 02:42 AM
Member
 
Icaro's Avatar
 
Join Date: Jul 2007
Location: Italy
Posts: 164
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 15
Quote:
Originally Posted by Icaro View Post

The step is the second,
when they start all the 3778 patches.

This is my settings for the calibration and profiling of a projector JVC HD990 means a i1D3

http://www.avsforum.com/t/1471169/madvr-argyllcms/2010#post_24805613

I need to change something ?
Quote:
Originally Posted by gwgill View Post

Hmm. - well that's a puzzle then. The white patches may be being read a bit too often during calibration with the current code, but not when reading a large number of patches in one hit.

I'm unable to reproduce the problem myself.

If someone who has this problem is comfortable running the ArgyllCMS tools on the command line, then I can make a debug version of dispread available to help track this down.

it may be useful these are the files Log:

filesLOG.zip 10k .zip file
Attached Files
File Type: zip filesLOG.zip (9.7 KB, 1 views)

Sorry for my bad English
Icaro 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