How low can the i1 display pro read - AVS Forum
Forum Jump: 
 4Likes
Reply
 
Thread Tools
post #1 of 51 Old 06-13-2014, 07:51 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 847
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 70
How low can the i1 display pro read

I'm considering getting an i1display pro so that I can properly adjust the gamma of my FW900 using Argyll. Right now, the black level of my display is below 0.01 nits, which means that my DTP-94 can't pick it up. As a result, I'm unable to generate an appropriate LUT without raising the black level.

According to Tom Huffman, it is capable of reading down to 0.003 cd/m2, which might be good enough for my needs (I'm gonna try and borrow a Minolta LS-100 from my lab to check, since that unit reads down to 0.001 cd/m2).

Here is the page where Tom describes his data:

http://www.chromapure.com/newgear_display3.asp

I'm wondering if anyone else can confirm this low light capability of the i1d3. Note: I'm not interested in accurate chromaticity readings, I'm just concerned with its ability to accurately measure low luminances.
spacediver is online now  
Sponsored Links
Advertisement
 
post #2 of 51 Old 06-14-2014, 09:22 AM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 4,619
Mentioned: 29 Post(s)
Tagged: 0 Thread(s)
Quoted: 194 Post(s)
Liked: 341
That is what I've found as well looking at this data set, at low light levels the 1-sigma uncertainty in Y is 0.002 nits when using the Klein as a reference.

-----

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.

Last edited by zoyd; 06-14-2014 at 09:26 AM.
zoyd is online now  
post #3 of 51 Old 06-14-2014, 10:03 AM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 847
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 70
Quote:
Originally Posted by zoyd View Post
That is what I've found as well looking at this data set, at low light levels the 1-sigma uncertainty in Y is 0.002 nits when using the Klein as a reference.
Yea, I looked at your thread recently and was amazed at that level of precision at low light levels. I don't see any measurements at levels lower than 0.01 nits in that data set (though I haven't looked through the .ti3 files). But I'm assuming in different cases you've found the i1d3 can measure below 0.01 nits right? Off hand, do you remember the lowest reading it reported?
spacediver is online now  
post #4 of 51 Old 06-14-2014, 10:08 AM
AVS Special Member
 
Chad B's Avatar
 
Join Date: Oct 2002
Location: Piqua, OH
Posts: 2,077
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 68 Post(s)
Liked: 412
In CalMAN I've had my i1D3 report readings as low as .0006 fL with good repeatability with the LLH on 5 seconds.

ISF/THX calibrator, HDTVbyChadB
Jeti 1211

Latest reviews:

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.
,
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.
Chad B is online now  
post #5 of 51 Old 06-14-2014, 10:28 AM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 847
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 70
wow, Chad, that is remarkable.
spacediver is online now  
post #6 of 51 Old 06-14-2014, 10:33 AM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 4,619
Mentioned: 29 Post(s)
Tagged: 0 Thread(s)
Quoted: 194 Post(s)
Liked: 341
Quote:
Originally Posted by spacediver View Post
Quote:
Originally Posted by zoyd View Post
That is what I've found as well looking at this data set, at low light levels the 1-sigma uncertainty in Y is 0.002 nits when using the Klein as a reference.
Yea, I looked at your thread recently and was amazed at that level of precision at low light levels. I don't see any measurements at levels lower than 0.01 nits in that data set (though I haven't looked through the .ti3 files). But I'm assuming in different cases you've found the i1d3 can measure below 0.01 nits right? Off hand, do you remember the lowest reading it reported?
Based on the precision seen 0.002 nit differences are measurable given enough averaging. Both ChadB and LarryinRI have reliably measured VT60s at these levels with the D3.

-----

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 #7 of 51 Old 06-14-2014, 10:40 AM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 847
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 70
ah, right - I didn't make the connection, understood.
spacediver is online now  
post #8 of 51 Old 06-15-2014, 05:40 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 632
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 76 Post(s)
Liked: 80
It will depend on the driver a little (i.e. don't expect the X-Rite and ArgyllCMS drivers to have precisicely the same behavior), but ultimately the resolution is limited by the maximum integration time (20 sec) and the L2F sensors. So one transition in 20 seconds is the lowest possible limit. In practice there may be a threshold below which the sensor returns 0.

Using the ArgyllCMS driver in it's default mode, it will integrate for 20 seconds, and the lowest non-zero value of Y seems to be about around 0.002 cd/m^2.
gwgill is online now  
post #9 of 51 Old 06-15-2014, 06:22 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 847
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 70
thanks for this info, quite a remarkable piece of hardware.
spacediver is online now  
post #10 of 51 Old 06-18-2014, 07:50 PM
AVS Special Member
 
TomHuffman's Avatar
 
Join Date: Jan 2003
Location: Springfield, MO
Posts: 6,420
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 52 Post(s)
Liked: 217
Since this old issue has come up again, I thought that I would re-do my initial test, this time using greater precision.

You can see the result by going to our i1Display Pro New Gear article, scrolling down to the update under Low-Light Sensitivity Test. Or if you can just go directly to the video.
spacediver likes this.

Tom Huffman
ChromaPure Software/AccuPel Video Signal Generators
ISF/THX Calibrations
Springfield, MO

TomHuffman is online now  
post #11 of 51 Old 06-19-2014, 04:41 AM
Member
 
22point8's Avatar
 
Join Date: Mar 2012
Posts: 114
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 30 Post(s)
Liked: 25


Sony PVM 2044QM (made in January 1990), thats as good as the greyscale gets because the front panel green gain pot doesn't work so theres only green bias. Anyway, 0.002cdm2 black (0.002278).

Last edited by 22point8; 06-19-2014 at 04:46 AM.
22point8 is offline  
post #12 of 51 Old 06-19-2014, 04:42 AM
Advanced Member
 
trans_lux's Avatar
 
Join Date: Oct 2001
Location: In my theater
Posts: 974
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 64 Post(s)
Liked: 136
In order to get accurate reading of the black floor you would need a Klein K10 that reads all the way down to 0.00002 fl.?
trans_lux is offline  
post #13 of 51 Old 06-19-2014, 07:47 AM
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,330
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 97 Post(s)
Liked: 477
Quote:
Originally Posted by trans_lux View Post
In order to get accurate reading of the black floor you would need a Klein K10 that reads all the way down to 0.00002 fl.?
Hi, Klein K-10A takes 256 measurements per second at any luminance level, low/ultra low/ high..whatever..... From these 256 measurements it's doing an average internally and returns the results to the software 8 times per second, this means that it's measurement time is fixed at 125 ms. (0.125 sec).

The advantage of that fixed 125ms measuring time is that the patch you are measuring it's measured instantly, when you hit measure.

There tons of displays that are autodimming when you display a patch for some seconds more than normally, a lot of displays like LED's are shutting down their backlight or some plasma (like KURO) are entering to a panel 'shutdown' mode.

When you will try to measure these displays black level with a meter other that Klein, you are not sure that the returned results are accurate because during the about 10-20 seconds that will require to return you the results, the panel's luminance level will not be the same during all this time.
buzzard767 likes this.


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

S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #14 of 51 Old 06-19-2014, 08:34 AM
AVS Special Member
 
Chad B's Avatar
 
Join Date: Oct 2002
Location: Piqua, OH
Posts: 2,077
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 68 Post(s)
Liked: 412
Ted, do you know if that means that a single read sample of the K-10a in CalMAN is actually an average of 32 readings (256/8)? And if that's the case, would setting Calman to 3 read average mean that readings would take less than half a second and be an average of 96 readings? That would make CalMAN's LLH settings for the Klein seem like overkill.
What about ChromaPure, do you know if it's the same? Thanks.

ISF/THX calibrator, HDTVbyChadB
Jeti 1211

Latest reviews:

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.
,
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.
Chad B is online now  
post #15 of 51 Old 06-19-2014, 08:45 AM
AVS Special Member
 
Chad B's Avatar
 
Join Date: Oct 2002
Location: Piqua, OH
Posts: 2,077
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 68 Post(s)
Liked: 412
When I take black level readings with my i1D3 on a display that dims with no signal, I put up a pattern with a bit of content somewhere else on the screen to keep the display "awake". With displays like the Samsung F8500 with it's floating blacks and Black Optimizer, that can return very different results than a totally black screen, but arguably more relevant.

ISF/THX calibrator, HDTVbyChadB
Jeti 1211

Latest reviews:

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.
,
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.
Chad B is online now  
post #16 of 51 Old 06-19-2014, 09:21 AM
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,330
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 97 Post(s)
Liked: 477
Quote:
Originally Posted by Chad B View Post
Ted, do you know if that means that a single read sample of the K-10a in CalMAN is actually an average of 32 readings (256/8)? And if that's the case, would setting Calman to 3 read average mean that readings would take less than half a second and be an average of 96 readings? That would make CalMAN's LLH settings for the Klein seem like overkill.
What about ChromaPure, do you know if it's the same? Thanks.
If I remember well, one time I was checking the measurements log files with CalMAN it takes to measure about 50ms with single measurement, but I will confirm it the next time

125 ms is for LightSpace.


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

S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #17 of 51 Old 06-19-2014, 10:15 AM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,403
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 134
Quote:
Originally Posted by 22point8 View Post

Sony PVM 2044QM (made in January 1990), thats as good as the greyscale gets because the front panel green gain pot doesn't work so theres only green bias. Anyway, 0.002cdm2 black (0.002278).
As I've said before ... sometimes our fond memories and high regards for CRT tech is very often ... unwarranted.

I wonder how many transfers and color grades, over the years, were created using CRT monitors with a blue bias in the lower range of the grayscale. ...
ConnecTEDDD likes this.
HDTVChallenged is offline  
post #18 of 51 Old 06-19-2014, 11:07 AM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 847
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 70
Quote:
Originally Posted by HDTVChallenged View Post
As I've said before ... sometimes our fond memories and high regards for CRT tech is very often ... unwarranted.

I wonder how many transfers and color grades, over the years, were created using CRT monitors with a blue bias in the lower range of the grayscale. ...
I'd like to think that studios wouldn't be that sloppy (plus they probably used BVMs, though not sure about this).

The FW900 was also used in practice in many studios.

Here are some recent measurements of my FW900 after only a hardware calibration using the service port. The gamma is way dark because I brought the G2 voltage to a ridiculously low level, but after Argyll, I'm able to get a virtually perfect 2.4 gamma response (and without any quantization artifacts). The values shown are before I used Argyll. I may be receiving a minolta LS-100 soon and will share the near black response when I do.

The grayscale tracks remarkably well despite the process only involving essentially a 2 point grayscale adjustment. (edit: note that these are CIE76 values, which are lower than CIE 2000)



Here are the same measurements but using the CIE2000 delta E formula:


Last edited by spacediver; 06-19-2014 at 11:49 AM.
spacediver is online now  
post #19 of 51 Old 06-19-2014, 04:19 PM
Member
 
Synnove's Avatar
 
Join Date: Feb 2014
Posts: 61
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 20
Nice results. Only limited to 80nits though? I'm sure if you get a LUT box hooked up to it you'll end up with fantastic results. What's the gamut like?
Synnove is offline  
post #20 of 51 Old 06-19-2014, 05:29 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 847
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 70
Quote:
Originally Posted by Synnove View Post
Nice results. Only limited to 80nits though? I'm sure if you get a LUT box hooked up to it you'll end up with fantastic results. What's the gamut like?
More like 89 nits, which is about factory spec for these units. I could drive it higher but that would probably reduce the lifespan of the tube. Plus, halation is an issue with these units (lighter areas spread their glow to surrounding areas), and that would likely be magnified if I increased peak luminance.

Because these CRTs are so linear, not sure how much difference a LUT box or a 3D LUT would make, but it is something I'm looking forward to experimenting with.

phosphors are based on the SMPTE-C primaries (which is a gamut associated with Rec 601). Not quite Rec 709 but close enough. Interestingly, much HD material was until quite recently mastered with these primaries.
spacediver is online now  
post #21 of 51 Old 06-20-2014, 09:56 AM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,403
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 134
Quote:
Originally Posted by spacediver View Post

The grayscale tracks remarkably well despite the process only involving essentially a 2 point grayscale adjustment. (edit: note that these are CIE76 values, which are lower than CIE 2000)



Here are the same measurements but using the CIE2000 delta E formula:

Here is the pitfall in relying on dE numbers alone ... you're clearly off the rails at 20% and below. y=0.3176 and y=0.3095 is *quite* blue relative to D65. Whether or not this is a "major" problem visually is another topic.

PS: Studio's not sloppy? How about "Bram Stoker's Dracula" (BD) ... "Blade Runner" (DVD) ... just off the top of my noggin.
HDTVChallenged is offline  
post #22 of 51 Old 06-20-2014, 10:23 AM
AVS Special Member
 
TomHuffman's Avatar
 
Join Date: Jan 2003
Location: Springfield, MO
Posts: 6,420
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 52 Post(s)
Liked: 217
Quote:
Originally Posted by HDTVChallenged View Post
Here is the pitfall in relying on dE numbers alone ... you're clearly off the rails at 20% and below. y=0.3176 and y=0.3095 is *quite* blue relative to D65. Whether or not this is a "major" problem visually is another topic.

PS: Studio's not sloppy? How about "Bram Stoker's Dracula" (BD) ... "Blade Runner" (DVD) ... just off the top of my noggin.
dE isn't the problem, it is the way it is being calculated in this case. The CIELUV dE of x0.3108, y0.3176 is 9.3. CIELAB is 6.9. CIEDE2000 is 11.4. However, this screenshot reports only 1.1. This is not correct.

Tom Huffman
ChromaPure Software/AccuPel Video Signal Generators
ISF/THX Calibrations
Springfield, MO

TomHuffman is online now  
post #23 of 51 Old 06-20-2014, 10:44 AM
Member
 
22point8's Avatar
 
Join Date: Mar 2012
Posts: 114
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 30 Post(s)
Liked: 25
The absolute Y W/O Gamma might account for it.
22point8 is offline  
post #24 of 51 Old 06-20-2014, 02:19 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 847
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 70
Quote:
Originally Posted by TomHuffman View Post
dE isn't the problem, it is the way it is being calculated in this case. The CIELUV dE of x0.3108, y0.3176 is 9.3. CIELAB is 6.9. CIEDE2000 is 11.4. However, this screenshot reports only 1.1. This is not correct.
The formula takes the luminance into account, since this is perceptually relevant when perceiving color difference.
spacediver is online now  
post #25 of 51 Old 06-20-2014, 02:25 PM
AVS Special Member
 
TomHuffman's Avatar
 
Join Date: Jan 2003
Location: Springfield, MO
Posts: 6,420
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 52 Post(s)
Liked: 217
Quote:
Originally Posted by 22point8 View Post
The absolute Y W/O Gamma might account for it.
Even taking that into account (which you shouldn't), the CIEDE2000 dE is 3.5, assuming a 2.22 gamma.

Tom Huffman
ChromaPure Software/AccuPel Video Signal Generators
ISF/THX Calibrations
Springfield, MO

TomHuffman is online now  
post #26 of 51 Old 06-20-2014, 02:44 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 847
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 72 Post(s)
Liked: 70
Quote:
Originally Posted by TomHuffman View Post
Even taking that into account (which you shouldn't), the CIEDE2000 dE is 3.5, assuming a 2.22 gamma.
Why shouldn't you?

There are contexts where one is only interested in chromaticity errors, discounting what the luminance should be at any given input level, and this is most certainly one of them, as I am only interested in how the grayscale tracked.
spacediver is online now  
post #27 of 51 Old 06-20-2014, 03:00 PM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 4,619
Mentioned: 29 Post(s)
Tagged: 0 Thread(s)
Quoted: 194 Post(s)
Liked: 341
Quote:
Originally Posted by TomHuffman View Post
dE isn't the problem, it is the way it is being calculated in this case. The CIELUV dE of x0.3108, y0.3176 is 9.3. CIELAB is 6.9. CIEDE2000 is 11.4. However, this screenshot reports only 1.1. This is not correct.
The screenshot is correct if you use 88.97 nits for Ywhite when calculating Lab (or Luv) coordinates for reference and target. This suppresses color differences when comparing dark patches which one could argue is appropriate if you are luminance adapted to scenes with a peak white present. If you want a more sensitive measure use the "relative Y" option, this will scale the white reference for conversion to Lab (Luv) such that L* = 100 at Y=0.47 nits for the 20% patch for example.

-----

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 #28 of 51 Old 06-20-2014, 05:33 PM
AVS Special Member
 
TomHuffman's Avatar
 
Join Date: Jan 2003
Location: Springfield, MO
Posts: 6,420
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 52 Post(s)
Liked: 217
Quote:
Originally Posted by zoyd View Post
The screenshot is correct if you use 88.97 nits for Ywhite when calculating Lab (or Luv) coordinates for reference and target. This suppresses color differences when comparing dark patches which one could argue is appropriate if you are luminance adapted to scenes with a peak white present. If you want a more sensitive measure use the "relative Y" option, this will scale the white reference for conversion to Lab (Luv) such that L* = 100 at Y=0.47 nits for the 20% patch for example.
Lab and Luv ALWAYS use relative luminance to assign a value for lightness in the 0-100 range. Peak white in almost never present in film content, which is a relatively dim medium. In fact, it is rarely over 50%.

There are only two ways to do this. Either you look at the chromaticity difference only and ignore the luminance, in which case,

Reference
L: 100
a: 0.0
b: 0.0

Measured
L: 100
a: 4.88
b: -4.85

The CIEDE2000 for this pair of colors is 7.3.

or,

you include luminance with an assumed gamma (2.22 for the example), in which case

Reference
L: 19.26
a: 0
b: 0

Measured
L: 4.697
a: 0.599
b: -0.6039

The CIEDE2000 for this pair of colors is 9.4. Ordinarily, using this method would result in a much lower dE score, but in this particular instance both the chromaticity and the luminance have large errors.

What you doing is really weird. You assign the SAME L value of 19.26 to both reference and test color. In other words, for the measured color you use the ideal luminance for 20% regardless of how it actually measures. This is the worst of both worlds. You consider luminance in that you use the expected luminance of that video level, but then you also ignore luminance by using the ideal luminance for the test value. This results in an absurdly low color difference of 1.1 for a color that is visibly blue (relative to D65) and with an awful gamma of 3.26. I would suggest that this is not very helpful.

Tom Huffman
ChromaPure Software/AccuPel Video Signal Generators
ISF/THX Calibrations
Springfield, MO


Last edited by TomHuffman; 06-20-2014 at 05:37 PM.
TomHuffman is online now  
post #29 of 51 Old 06-20-2014, 05:36 PM
AVS Special Member
 
TomHuffman's Avatar
 
Join Date: Jan 2003
Location: Springfield, MO
Posts: 6,420
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 52 Post(s)
Liked: 217
Quote:
Originally Posted by spacediver View Post
Why shouldn't you?

There are contexts where one is only interested in chromaticity errors, discounting what the luminance should be at any given input level, and this is most certainly one of them, as I am only interested in how the grayscale tracked.
My point exactly. Grayscale should be about chromaticity errors only. If you want to consider luminance errors, then look to gamma. Perhaps I misunderstood what the previous poster was referring to by "absolute Y W/O Gamma".

Tom Huffman
ChromaPure Software/AccuPel Video Signal Generators
ISF/THX Calibrations
Springfield, MO

TomHuffman is online now  
post #30 of 51 Old 06-20-2014, 05:40 PM
AVS Special Member
 
TomHuffman's Avatar
 
Join Date: Jan 2003
Location: Springfield, MO
Posts: 6,420
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 52 Post(s)
Liked: 217
Quote:
Originally Posted by spacediver View Post
The formula takes the luminance into account, since this is perceptually relevant when perceiving color difference.
No, it really doesn't. See my post below. In one way it does, while in a more important way it does not.

Tom Huffman
ChromaPure Software/AccuPel Video Signal Generators
ISF/THX Calibrations
Springfield, MO

TomHuffman is online now  
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