i1 profiler vs DisplayCAL. Colorimeter and display calibration software. - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 15Likes
Reply
 
Thread Tools
post #1 of 38 Old 01-05-2020, 04:02 PM - Thread Starter
Member
 
Join Date: Dec 2019
Posts: 22
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 18 Post(s)
Liked: 0
i1 profiler vs DisplayCAL. Colorimeter and display calibration software.

Hi,
I see that most people uses DisplayCAL over the XRite i1 Profiler even using the xrite i1 display pro colorimeter.

What is the sense of using a third party software when xrite give its own software?

Measurements are so different between those two software.
Even the white point measurement is way different with i1 profiler showing 6500K and DisplayCAL 6200K.

I really don't think that with an xrite colorimeter displaycal can be more accurate than the xrite sofyware,
so why people continue to use displaycal?

Have I missed something?
Davide Perini is offline  
Sponsored Links
Advertisement
 
post #2 of 38 Old 01-06-2020, 06:51 AM
AVS Forum Special Member
 
fhoech's Avatar
 
Join Date: Apr 2013
Location: Stuttgart, Germany
Posts: 1,059
Mentioned: 74 Post(s)
Tagged: 0 Thread(s)
Quoted: 582 Post(s)
Liked: 385
When you choose the exact same measurement parameters, i.e. in i1 Profiler called "Display Type", in DisplayCAL called "Correction", you'll get exactly the same measurements on white (within instrument/display noise and repeatability).

One reason you might choose DisplayCAL over i1 Profiler is because of technical limitations in the latter - you are limited to matrix or L*a*b* cLUT the last time I checked (quite a while ago). Due to the ICC encoding limitation of a* and b* in the cLUT from -128 to +127 this is not suitable for gamuts wider than, say, DCI P3 (usually saturated greens will get clipped first, as most wider gamuts are achieved by extending out the green primary the most).
Rolls-Royce, gwgill and L30Z3N like this.

DisplayCAL - Display Calibration and Characterization powered by ArgyllCMS
Current stable version 3.8.9.3 released 2019-12-14 | Changelog
DisplayCAL on Facebook
fhoech is offline  
post #3 of 38 Old 01-06-2020, 01:45 PM - Thread Starter
Member
 
Join Date: Dec 2019
Posts: 22
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 18 Post(s)
Liked: 0
Quote:
Originally Posted by fhoech View Post
When you choose the exact same measurement parameters, i.e. in i1 Profiler called "Display Type", in DisplayCAL called "Correction", you'll get exactly the same measurements on white (within instrument/display noise and repeatability).

One reason you might choose DisplayCAL over i1 Profiler is because of technical limitations in the latter - you are limited to matrix or L*a*b* cLUT the last time I checked (quite a while ago). Due to the ICC encoding limitation of a* and b* in the cLUT from -128 to +127 this is not suitable for gamuts wider than, say, DCI P3 (usually saturated greens will get clipped first, as most wider gamuts are achieved by extending out the green primary the most).
thanks for the reply.

there is no "correction" for my i1 display pro colorimeter in that menu. am I missing something?

are there some other reasons why to choose displaycal?
i1 profiler has been updated, now it supports even ICC v4 and it supports Automatic Screen Regulation.
Davide Perini is offline  
Sponsored Links
Advertisement
 
post #4 of 38 Old 01-06-2020, 04:42 PM
AVS Forum Special Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 1,134
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 382 Post(s)
Liked: 272
Quote:
Originally Posted by Davide Perini View Post
What is the sense of using a third party software when xrite give its own software?
Not a whole lot - that's one of the main benefits of bundling software for "free" with hardware -
stifle competition and minimize the need to improve the software!

Author of ArgyllCMS and ArgyllPRO ColorMeter
gwgill is offline  
post #5 of 38 Old 01-06-2020, 04:45 PM
AVS Forum Special Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 1,134
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 382 Post(s)
Liked: 272
Quote:
Originally Posted by Davide Perini View Post
i1 profiler has been updated, now it supports even ICC v4
Which would be great, if there was any demonstrable advantage of V4 over V2, which there is not.
Quote:
and it supports Automatic Screen Regulation.
Whatever that is !

(Can you explain what that does ?)

Author of ArgyllCMS and ArgyllPRO ColorMeter
gwgill is offline  
post #6 of 38 Old 01-07-2020, 08:54 AM
Member
 
stama's Avatar
 
Join Date: Feb 2013
Posts: 114
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 104 Post(s)
Liked: 53
I did several tests comparing the measurements between argyll (what DisplayCal uses), i1Profiler, and LightSpace (which is also using the Xrite SDK).

In the case of i1DisplayPro (i1D3), there's no notable difference between any of the above.
In case you are getting different measurements, make sure there are no other things interfering with your measurements. For instance both DisplayCal and i1Profiler have software running in the background which ask Windows to apply a color profile (including loading the gamma calibration from the vctag). It might be you are thinking you are measuring the display as it is, but in fact every value sent to the screen is modified by the currently active loaded gamma in the GPU.

Make sure to disable both DisplayCal's (DisplayCAL-apply-profiles) and i1Profiler's (i1ProfilerTray) when you are taking measurements, and load a unity neutral gamma to your GPU (there is an option in the DisplayCal menu for this or in its profile-setting tray app, or you can restart your computer after disabling these apps) to avoid one of them compromising the measurements you take with the other software.

One of the things you will find nice about argyll/DisplayCal is that you can make a 3D LUT which can either be loaded to your LG OLED TV (if you have one), to an external LUT box, or used in madVR (if you use it as a media player renderer), and that's better than using color profiles and expecting applications to use them (most don't).

PS: i1pro2, on the other hand, give very different measurement results between argyll and X-Rite's SDK on WRGBA OLEDs, but I couldn't say which are better, as I don't have any reference equipment to verify any measurements.

I think Automatic Screen Regulation refers to the feature that the software constantly monitors the ambient luminosity (if you leave your i1D3 plugged in in ambient measurement mode) and does ambient adaptation. That's a feature that no other color calibration software has, as far as I know.

Last edited by stama; 01-07-2020 at 09:18 AM.
stama is offline  
post #7 of 38 Old 01-07-2020, 03:11 PM
AVS Forum Special Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 1,134
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 382 Post(s)
Liked: 272
Quote:
Originally Posted by stama View Post
PS: i1pro2, on the other hand, give very different measurement results between argyll and X-Rite's SDK on WRGBA OLEDs, but I couldn't say which are better, as I don't have any reference equipment to verify any measurements.
What sort of differences ? i.e. do you have examples ?
(There should be no significant differences, and I've never seen any.)

Quote:
I think Automatic Screen Regulation refers to the feature that the software constantly monitors the ambient luminosity (if you leave your i1D3 plugged in in ambient measurement mode) and does ambient adaptation. That's a feature that no other color calibration software has, as far as I know.
That sort of "feature" has been around a long time, and is supported by other vendors as well (i.e. see the Datcolor Spyder software and their "Room light monitoring").

But no-one serious about color would use such features, since it messes up the color accuracy. For the most accurate color display you want to keep everything constant - the behavior of the display (so that the profile remains an accurate recording of its behavior), and the surroundings you view the display in (so that you perception of the display remains constant). Changing the display in response to ambient intensity and/or color may be cute, but it isn't what you do if color accuracy is your goal.

Author of ArgyllCMS and ArgyllPRO ColorMeter
gwgill is offline  
post #8 of 38 Old 01-08-2020, 04:16 AM
Member
 
stama's Avatar
 
Join Date: Feb 2013
Posts: 114
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 104 Post(s)
Liked: 53
Quote:
Originally Posted by gwgill View Post
What sort of differences ? i.e. do you have examples ?
(There should be no significant differences, and I've never seen any.)
Yes, I have. I intend to make a post with more details, but here is quick example of (Y, x, y) triplets:

115.30 0.3119 0.3288 Argyll adaptive mode, normal resolution
114.25 0.3130 0.3289 Argyll adaptive mode, High resolution
113.92 0.3132 0.3289 Argyll non-adaptive mode, High resolution (-Y A)
113.84 0.3085 0.3263 LightSpace (there's no way to configure the i1Pro2 in LS)

My i1Pro2 is an OEM probe, so it' not recognized by iProfiler and I don't have measurements done with it. I expect it to give the same results as LightSpace though, since LightSpace is said to use the X-Rite SDK with no extra processing done by itself.

Luminosity kept varying between high 113 nits to low 115 nits during the measurements, and the x and y did change in time as well. But while there are slight variations in time, they remain mostly in the same target area, the LS x coordinate measurements never go over 0.31, while Argyll measures always above 0.31. Same story for y.

Measurements were carried out in isfExpert Bright mode, OLED light 30, brightness 49 (for really no light emitted at RGB 0,0,0), Contrast 85, with no white point or LUT corrections, Warm 1 color temperature, 2.2 gamma, and Auto gamut (which means the TV attempts to reproduce Rec709 gamut).

Considering the TV EDID advertises standard Rec709 colorimetry with Rec709 white point, I was surprised to see the Argyll High resolution mode being extremely close to it, while nothing else was. But, like I said, I don't have reference equipment to actually validate whether this panel really measures like that or not.

Quote:
Originally Posted by gwgill View Post
For the most accurate color display you want to keep everything constant - the behavior of the display (so that the profile remains an accurate recording of its behavior), and the surroundings you view the display in (so that you perception of the display remains constant). Changing the display in response to ambient intensity and/or color may be cute, but it isn't what you do if color accuracy is your goal.
I am not going to dispute that having a controlled viewing environment is going to be best, but most people do not view their TVs or PC monitors in controlled ambient luminosity conditions, and adaptation to ambient luminosity is a part of the "proper" way to reproduce color in case of relative tone mapping curves. That being said, nobody actually uses this (but that's another story)!

Now, I feel like I have to give a proper reply to the original question: why use something like argyll and DisplayCal instead of i1Profiler?

For me it simply comes down to i1Profiler being extremely limited in what it can do:
- it relies on very constraining workflows
- no easy way to make it display a patch and see what the luminance, x, and y coordinates are for that patch (this should be hassle-free, and it's not); there's no workflow for that
- creating a custom list of patches is such a complicated affair, you have to be in one of the measurement workflows to do it, and the patch generator is nowhere near what argyll's targen can do
- the workflow for displays basically forces you into having a profile generated and applied using the Windows color management
- it can't do 3D LUTs
- it is not able to use other patch generators
- etc.

With argyll you can do (almost) anything. You can even modify its source code to try new things, or just record measurements and then process them in some other way. I've installed i1Profiler only twice in the past ten years, and only for short periods of time: the first time I got my first i1Display Pro because I was curious to see what it can do, and a second time this winter because I wanted to check if it gives different results for a i1Display Pro Plus probe than other software (no, it doesn't).

Last edited by stama; 01-08-2020 at 08:08 AM. Reason: added details on what i1Profiler and argyll can do
stama is offline  
post #9 of 38 Old 01-08-2020, 03:35 PM
AVS Forum Special Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 1,134
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 382 Post(s)
Liked: 272
Quote:
Originally Posted by stama View Post
Yes, I have. I intend to make a post with more details, but here is quick example of (Y, x, y) triplets:

115.30 0.3119 0.3288 Argyll adaptive mode, normal resolution
114.25 0.3130 0.3289 Argyll adaptive mode, High resolution
113.92 0.3132 0.3289 Argyll non-adaptive mode, High resolution (-Y A)
113.84 0.3085 0.3263 LightSpace (there's no way to configure the i1Pro2 in LS)

My i1Pro2 is an OEM probe, so it' not recognized by iProfiler and I don't have measurements done with it. I expect it to give the same results as LightSpace though, since LightSpace is said to use the X-Rite SDK with no extra processing done by itself.

Luminosity kept varying between high 113 nits to low 115 nits during the measurements, and the x and y did change in time as well. But while there are slight variations in time, they remain mostly in the same target area, the LS x coordinate measurements never go over 0.31, while Argyll measures always above 0.31. Same story for y.
Seems like the luminance variation is due to your display. It's hard to know the significance or otherwise of the x,y differences, since x,y is not a perceptual space - a*b* or u'v' would be better spaces to make such an accuracy comparison. It would also be better to be comparing values from the X-Rite driver rather than via LightSpace just in case, but it could simply be that the LightSpace values are more heavily quantized due to the choice of a lower integration time. As I said, I've never seen any significant differences, in my testing the Argyll driver result is consistent with both the X-Rite driver and a reference instrument, and I've had no other reports of any accuracy issues.

I'm surprised that iProfiler doesn't work with the i1Pro2 OEM you have - the i1Pro doesn't have any locking codes, although it does have different feature sets for the OEM. It should at least run in backwards compatible i1Pro1 mode.

Author of ArgyllCMS and ArgyllPRO ColorMeter
gwgill is offline  
post #10 of 38 Old 01-09-2020, 11:26 AM
Member
 
stama's Avatar
 
Join Date: Feb 2013
Posts: 114
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 104 Post(s)
Liked: 53
Quote:
Originally Posted by gwgill View Post
Seems like the luminance variation is due to your display. It's hard to know the significance or otherwise of the x,y differences, since x,y is not a perceptual space - a*b* or u'v' would be better spaces to make such an accuracy comparison. It would also be better to be comparing values from the X-Rite driver rather than via LightSpace just in case, but it could simply be that the LightSpace values are more heavily quantized due to the choice of a lower integration time. As I said, I've never seen any significant differences, in my testing the Argyll driver result is consistent with both the X-Rite driver and a reference instrument, and I've had no other reports of any accuracy issues.

I'm surprised that iProfiler doesn't work with the i1Pro2 OEM you have - the i1Pro doesn't have any locking codes, although it does have different feature sets for the OEM. It should at least run in backwards compatible i1Pro1 mode.
The luminance variations are indeed due to the display, the OLED luminosity is not stable at all.

Regarding the choice of coordinates used for comparison, I don't understand the reason why perceptual space coordinates are better than absolute coordinates though. I would have thought that absolute coordinates are best, since you don't have to make any assumptions, like what illuminant to use (for conversion to D65 Lab values for instance) and so on. Conversion to some space with a certain illuminant requires chromatic adaptation, which introduces errors by itself.

I was surprised to see the i1Pro2 does not get recognized by the retail iProfiler software too. But it seems that enabling the use of a OEM i1Pro2 (or other X-Rite spectro) in the X-Rite software can only be done by paying for having the probe serial to be enabled in the software, or you have to use the software that comes with some OEM packages (like those for Canon press printers) which have a serial printed on the CD cover. This OEM probe does not seem to be pre-enabled in the retail software. Despite the fact that it is visible in LightSpace or Calman (which both use the X-Rite driver to access it), the i1Profiler software does not list it as available. The X-Rite i1Diagnostic software also sees it and can run the tests on it, so I really think it has to do with the retail i1Profiler software locking out OEM device serials.

Btw, I've used LightSpace ZRO for readings, which does not require a license. It can be used to measure patches (a click on the "Patch color" rectangle in its "Calibration interface" window will show the patch in a separate window that can be resized and moved anywhere on the screen, while the color sliders can be used to set the color to be displayed) but it only displays Yxy for readings.

Regarding other reports on the accuracy of the i1Pro2 on OLED displays, I have not seen other people posting results from both argyll and some other software which uses the X-Rite driver. So, these differences could be, just that nobody did experiments or reported them. Please believe me, I was very surprised by the large difference in measurements.

It does not help that it's quite complicated for most people to switch between the two drivers: that you need to have the argyll driver installed first (with Windows 10 restarted with driver signing disabled), then restart and install the Xrite Device Services which installs the X-rite driver, and after that you can use the properties in the Device Manager to rollback to the argyll driver when needed, or to reinstall the X-rite one.

I asked Anger.miki in the LightSpace thread if he tried using the argyll high resolution mode with his i1Pro2 and how it compares to a JETI on a OLED panel, but got no reply unfortunately. Steve gave a reply instead, where as expected, he said to use what LightSpace is using (the X-Rite driver) because it gives more stable measurements, not because it would be more accurate or not.

PS: I had a look in the argyll code, and it seems that it uses a linear interpolation for spectros with less than 5nm half-width for converting the measured spectral response to a 1nm spaced one, and a 3rd order polynomial (Lagrange?) interpolation for those with a larger half-width like the i1pro2. I 've read that the recommendation for such interpolation is to use either the method developed by Sprague for functions having a uniformly spaced independent variable, or a Cubic Spline method for non-uniformly spaced independent variable. And to use the last measured value for any extrapolation (but I've noticed argyll does that already). At least that's what it says here (needs a search in the page for [CIET13805a]). I'm not saying that following these rules gives a more accurate result than what argyll achieves, only that maybe these are the rules that have to be followed to get the same result as the X-Rite software. I should try and save the spectra to a file and try this myself, just have to find the time to do it. I don't know if there is a way to save the spectra from LightSpace too.

Last edited by stama; 01-09-2020 at 12:59 PM.
stama is offline  
post #11 of 38 Old 01-09-2020, 12:47 PM
Senior Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 215
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 163 Post(s)
Liked: 232
Quote:
Originally Posted by stama View Post

I was surprised to see the i1Pro2 does not get recognized by the retail iProfiler software too. But it seems that enabling the use of a OEM i1Pro2 (or other X-Rite spectro) in the X-Rite software can only be done by paying for having the probe serial to be enabled in the software, or you have to use the software that comes with some OEM packages (like those for Canon press printers) which have a serial printed on the CD cover.
Not wanting to jump in on the above discussion, but regarding use of the i1Pro2 OEM in the retail version of i1Profiler.

I have both a retail and OEM i1 Display Pro and an i1Pro2. I found (totally by accident) that if I had my retail i1 Display Pro and i1Pro2 connected when opening i1Profiler, the software moves out of trial mode (due to retail meter) and you can switch to using the OEM i1Pro2 that way. I only found this as I had both probes still connected from calibrating my OLED when I decided to do my laptop display too. If you have (or have access to) a retail i1d3 it may be worth checking to see if it is repeatable for you.
Rolls-Royce likes this.

LG OLED65C9MLB**Denon AVR-X3500H**Focal Sib Evo Atmos 5.1.2**Sky Q UHD**Xbox One X**Panasonic DP-UB820**ATV 4K**Harmony Elite
i1D3 OEM**i1D3 Retail**i1 Pro2 OEM**LightSpace HTP**CalMAN Home for LG**RPi Pgenerator**Ted's Disk & Media Files**S & M HD & UHD HDR Benchmark Disks**HD Fury Integral 2
liberator72 is online now  
post #12 of 38 Old 01-09-2020, 12:57 PM
Member
 
stama's Avatar
 
Join Date: Feb 2013
Posts: 114
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 104 Post(s)
Liked: 53
Quote:
Originally Posted by liberator72 View Post
Not wanting to jump in on the above discussion, but regarding use of the i1Pro2 OEM in the retail version of i1Profiler.

I have both a retail and OEM i1 Display Pro and an i1Pro2. I found (totally by accident) that if I had my retail i1 Display Pro and i1Pro2 connected when opening i1Profiler, the software moves out of trial mode (due to retail meter) and you can switch to using the OEM i1Pro2 that way. I only found this as I had both probes still connected from calibrating my OLED when I decided to do my laptop display too. If you have (or have access to) a retail i1d3 it may be worth checking to see if it is repeatable for you.
Hmm.. I do have retail i1 Display Pro probes, but it did not work for me. The i1Pro2 did not show up in the list of available probes, from what I can remember. Another thing I have to try again, I suppose.
stama is offline  
post #13 of 38 Old 01-09-2020, 01:24 PM
AVS Forum Special Member
 
Rolls-Royce's Avatar
 
Join Date: Apr 2003
Location: Victorville, CA
Posts: 4,407
Mentioned: 42 Post(s)
Tagged: 0 Thread(s)
Quoted: 1273 Post(s)
Liked: 1345
Quote:
Originally Posted by stama View Post
Hmm.. I do have retail i1 Display Pro probes, but it did not work for me. The i1Pro2 did not show up in the list of available probes, from what I can remember. Another thing I have to try again, I suppose.
Pro and Pro2 licensing is very different from that in the Display Pro. In the case of the retail Pro/Pro2, a license or licenses from X-Rite will be installed in the meter itself and can be added to by purchasing a license for the desired function (printer and camera profiling, etc) and installed via I1 Diagnostics. These licenses allow the meter to work with XRite's profiling software. The OEM Pro/Pro2 come with no licenses installed and although they will work with the diagnostics as well as third-party software, they do not with I1 Profiler. At one time, you could register and purchase licenses for the OEM I1Pro from XRite (I did it years ago for mine to use it for monitor profiling), but I don't know if the OEM Pro2 can have licenses added. I believe Ted has written that they are "seen" by the I1 drivers as essentially an I1 Pro, with some of the Pro 2 enhancements missing.

...Royce...

"I never drink...wine."
Bela Lugosi, DRACULA, 1931

Last edited by Rolls-Royce; 01-09-2020 at 01:37 PM.
Rolls-Royce is offline  
post #14 of 38 Old 01-09-2020, 01:56 PM
Advanced Member
 
Vishwa Somayaji's Avatar
 
Join Date: Jun 2001
Location: Edmonton,Alberta Canada
Posts: 572
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 110 Post(s)
Liked: 104
I had tried using OEM iPro 2 with i1 Profiler about a year ago as soon as I bought it. My experience was that it went through the whole process of monitor calibration as usual and at the end refused to make any changes to the 1D LUT of the video card. When I talked to X-Rite support people they said there was nothing they could do as it was OEM. I left it at that as I already had i1 Display Pro at the moment.
Rolls-Royce likes this.
Vishwa Somayaji is offline  
post #15 of 38 Old 01-09-2020, 01:58 PM
Senior Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 215
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 163 Post(s)
Liked: 232
Quote:
Originally Posted by Rolls-Royce View Post
Pro and Pro2 licensing is very different from that in the Display Pro. In the case of the retail Pro/Pro2, a license or licenses from X-Rite will be installed in the meter itself and can be added to by purchasing a license for the desired function (printer and camera profiling, etc) and installed via I1 Diagnostics. These licenses allow the meter to work with XRite's profiling software. The OEM Pro/Pro2 come with no licenses installed and although they will work with the diagnostics as well as third-party software, they do not with I1 Profiler. At one time, you could register and purchase licenses for the OEM I1Pro from XRite (I did it years ago for mine to use it for monitor profiling), but I don't know if the OEM Pro2 can have licenses added. I believe Ted has written that they are "seen" by the I1 drivers as essentially an I1 Pro, with some of the Pro 2 enhancements missing.
While this is true, I definitely 100% used my OEM i1Pro2 to profile my laptop display with i1Profiler, and as I say, it was completely by accident that I found I could use it. It was before I purchased my OEM i1d3, so I was using my retail i1d3 in combination with my OEM i1Pro2 to profile my LG OLED, and as both probes were still connected I decided to try i1Profiler to do my laptop (just for fun). I found I could select my OEM i1Pro2 and use it within i1Profiler.

It was about 6 months ago, and my retail is currently boxed up, but I can pull it out tomorrow and check if it works still, as i1Profiler has been updated since then

LG OLED65C9MLB**Denon AVR-X3500H**Focal Sib Evo Atmos 5.1.2**Sky Q UHD**Xbox One X**Panasonic DP-UB820**ATV 4K**Harmony Elite
i1D3 OEM**i1D3 Retail**i1 Pro2 OEM**LightSpace HTP**CalMAN Home for LG**RPi Pgenerator**Ted's Disk & Media Files**S & M HD & UHD HDR Benchmark Disks**HD Fury Integral 2
liberator72 is online now  
post #16 of 38 Old 01-09-2020, 02:15 PM
AVS Forum Special Member
 
Rolls-Royce's Avatar
 
Join Date: Apr 2003
Location: Victorville, CA
Posts: 4,407
Mentioned: 42 Post(s)
Tagged: 0 Thread(s)
Quoted: 1273 Post(s)
Liked: 1345
Quote:
Originally Posted by liberator72 View Post
While this is true, I definitely 100% used my OEM i1Pro2 to profile my laptop display with i1Profiler, and as I say, it was completely by accident that I found I could use it. It was before I purchased my OEM i1d3, so I was using my retail i1d3 in combination with my OEM i1Pro2 to profile my LG OLED, and as both probes were still connected I decided to try i1Profiler to do my laptop (just for fun). I found I could select my OEM i1Pro2 and use it within i1Profiler.

It was about 6 months ago, and my retail is currently boxed up, but I can pull it out tomorrow and check if it works still, as i1Profiler has been updated since then
I don't doubt you. Perhaps having the Retail I1D3 connected "tricks" the software. On his website, Ted has described the OEM Pro 2 as "Display only", so it's possible that the meter comes with a display profiling license installed. Did you ever try to run Profiler with just the OEM Pro 2 connected? My OEM Pro was purchased from SpectraCal many moons ago when they first began selling them (before the Display Pro), and it definitely didn't work with XRite's then-factory-software until a license for display profiling was purchased from XRite.

...Royce...

"I never drink...wine."
Bela Lugosi, DRACULA, 1931
Rolls-Royce is offline  
post #17 of 38 Old 01-09-2020, 03:07 PM
Senior Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 215
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 163 Post(s)
Liked: 232
Quote:
Originally Posted by Rolls-Royce View Post
I don't doubt you. Perhaps having the Retail I1D3 connected "tricks" the software. On his website, Ted has described the OEM Pro 2 as "Display only", so it's possible that the meter comes with a display profiling license installed. Did you ever try to run Profiler with just the OEM Pro 2 connected? My OEM Pro was purchased from SpectraCal many moons ago when they first began selling them (before the Display Pro), and it definitely didn't work with XRite's then-factory-software until a license for display profiling was purchased from XRite.
No that’s right and why I agreed with you above

I fully didn’t expect it to work as I knew it wasn’t licensed to be used in any of X-Rites own software. I purchased via Ted so was well aware of the OEM limitation with i1Profiler. I planned to do my laptop display with my retail i1d3 but as the i1Pro2 showed up in list of connected meters, I set it as default meter and it worked.

If I connect it without the retail unit connected, the OEM is not selectable and the software stays in demo mode only. This was all I was trying to point out to stama with my original post, and that if he had a retail meter it might be something he could try.

Hope that makes sense
liberator72 is online now  
post #18 of 38 Old 01-09-2020, 03:13 PM
AVS Forum Special Member
 
Rolls-Royce's Avatar
 
Join Date: Apr 2003
Location: Victorville, CA
Posts: 4,407
Mentioned: 42 Post(s)
Tagged: 0 Thread(s)
Quoted: 1273 Post(s)
Liked: 1345
Quote:
Originally Posted by liberator72 View Post
No that’s right and why I agreed with you above

I fully didn’t expect it to work as I knew it wasn’t licensed to be used in any of X-Rites own software. I purchased via Ted so was well aware of the OEM limitation with i1Profiler. I planned to do my laptop display with my retail i1d3 but as the i1Pro2 showed up in list of connected meters, I set it as default meter and it worked.

If I connect it without the retail unit connected, the OEM is not selectable and the software stays in demo mode only. This was all I was trying to point out to stama with my original post, and that if he had a retail meter it might be something he could try.

Hope that makes sense
Perfect sense. I just didn't remember whether or not you had tried the Pro 2 with Profiler on its own...
liberator72 likes this.

...Royce...

"I never drink...wine."
Bela Lugosi, DRACULA, 1931
Rolls-Royce is offline  
post #19 of 38 Old 01-09-2020, 03:22 PM
Senior Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 215
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 163 Post(s)
Liked: 232
Quote:
Originally Posted by Rolls-Royce View Post
Perfect sense. I just didn't remember whether or not you had tried the Pro 2 with Profiler on its own...
Actually tried again 2 nights ago forgetting I had my OEM i1d3 connected and not my Retail. But I will definitely try it again tomorrow (it's late here in the UK) and report back. It is purely to see if stama has a way to use his OEM i1Pro2 in i1Profiler

LG OLED65C9MLB**Denon AVR-X3500H**Focal Sib Evo Atmos 5.1.2**Sky Q UHD**Xbox One X**Panasonic DP-UB820**ATV 4K**Harmony Elite
i1D3 OEM**i1D3 Retail**i1 Pro2 OEM**LightSpace HTP**CalMAN Home for LG**RPi Pgenerator**Ted's Disk & Media Files**S & M HD & UHD HDR Benchmark Disks**HD Fury Integral 2
liberator72 is online now  
post #20 of 38 Old 01-09-2020, 03:31 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 7,577
Mentioned: 150 Post(s)
Tagged: 0 Thread(s)
Quoted: 5384 Post(s)
Liked: 2129
Quote:
Originally Posted by liberator72 View Post
While this is true, I definitely 100% used my OEM i1Pro2 to profile my laptop display with i1Profiler, and as I say, it was completely by accident that I found I could use it. It was before I purchased my OEM i1d3, so I was using my retail i1d3 in combination with my OEM i1Pro2 to profile my LG OLED, and as both probes were still connected I decided to try i1Profiler to do my laptop (just for fun). I found I could select my OEM i1Pro2 and use it within i1Profiler.

It was about 6 months ago, and my retail is currently boxed up, but I can pull it out tomorrow and check if it works still, as i1Profiler has been updated since then
I just tried that and can confirm that i1Profiler recognized my OEM i1Pro2, but only after I plug in my i1DisplayPro.

EDIT: Note that the reverse does not work; i.e., you cannot use am OEM i1DisplayPro with i1Profiler, even if a licensed i1Pro/Pro2 is plugged in.
liberator72 and Rolls-Royce like this.

Last edited by Dominic Chan; 01-14-2020 at 09:33 PM.
Dominic Chan is online now  
post #21 of 38 Old 01-09-2020, 03:36 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 7,577
Mentioned: 150 Post(s)
Tagged: 0 Thread(s)
Quoted: 5384 Post(s)
Liked: 2129
Quote:
Originally Posted by stama View Post
It does not help that it's quite complicated for most people to switch between the two drivers: that you need to have the argyll driver installed first (with Windows 10 restarted with driver signing disabled), then restart and install the Xrite Device Services which installs the X-rite driver, and after that you can use the properties in the Device Manager to rollback to the argyll driver when needed, or to reinstall the X-rite one.
Once both drivers are installed you can switch between the two drivers quite easily. And if you set them up with different USB ports, Windows will remember the "correct" driver to use when you plug the i1Pro into the corresponding USB port. I have to do that for my projector as the JVC Autocal software requires the X-Rite driver, whereas HCFR requries the ArgyllCMS driver.
Dominic Chan is online now  
post #22 of 38 Old 01-09-2020, 03:37 PM
Senior Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 215
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 163 Post(s)
Liked: 232
Quote:
Originally Posted by Dominic Chan View Post
I just tried that and can confirm that i1Profiler recognized my OEM i1Pro2, but only after I plug in my i1DisplayPro.
Ah great. And once detected it usable for you too? If so it will save me digging my retail i1d3 out to test tomorrow

LG OLED65C9MLB**Denon AVR-X3500H**Focal Sib Evo Atmos 5.1.2**Sky Q UHD**Xbox One X**Panasonic DP-UB820**ATV 4K**Harmony Elite
i1D3 OEM**i1D3 Retail**i1 Pro2 OEM**LightSpace HTP**CalMAN Home for LG**RPi Pgenerator**Ted's Disk & Media Files**S & M HD & UHD HDR Benchmark Disks**HD Fury Integral 2
liberator72 is online now  
post #23 of 38 Old 01-09-2020, 03:39 PM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 7,577
Mentioned: 150 Post(s)
Tagged: 0 Thread(s)
Quoted: 5384 Post(s)
Liked: 2129
Quote:
Originally Posted by liberator72 View Post
Ah great. And once detected it usable for you too? If so it will save me digging my retail i1d3 out to test tomorrow
Yes, I thought I'd save you some time, after learning that trick from your post.
liberator72 likes this.
Dominic Chan is online now  
post #24 of 38 Old 01-09-2020, 03:42 PM
Senior Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 215
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 163 Post(s)
Liked: 232
Quote:
Originally Posted by Dominic Chan View Post
Yes, I thought I'd save you some time, after learning that trick from your post.
Perfect, thank you for confirming I’m not going bonkers and remembering things wrong

You just need a retail unit to “unlock” the software from Demo Mode.

My post was not in vain then, my work here is done
Rolls-Royce likes this.

LG OLED65C9MLB**Denon AVR-X3500H**Focal Sib Evo Atmos 5.1.2**Sky Q UHD**Xbox One X**Panasonic DP-UB820**ATV 4K**Harmony Elite
i1D3 OEM**i1D3 Retail**i1 Pro2 OEM**LightSpace HTP**CalMAN Home for LG**RPi Pgenerator**Ted's Disk & Media Files**S & M HD & UHD HDR Benchmark Disks**HD Fury Integral 2
liberator72 is online now  
post #25 of 38 Old 01-09-2020, 05:08 PM
AVS Forum Special Member
 
Rolls-Royce's Avatar
 
Join Date: Apr 2003
Location: Victorville, CA
Posts: 4,407
Mentioned: 42 Post(s)
Tagged: 0 Thread(s)
Quoted: 1273 Post(s)
Liked: 1345
Quote:
Originally Posted by liberator72 View Post
Perfect, thank you for confirming I’m not going bonkers and remembering things wrong

You just need a retail unit to “unlock” the software from Demo Mode.

My post was not in vain then, my work here is done
Looks like it's time to head to the local for a pint...
liberator72 likes this.

...Royce...

"I never drink...wine."
Bela Lugosi, DRACULA, 1931
Rolls-Royce is offline  
post #26 of 38 Old 01-10-2020, 05:01 AM
Member
 
stama's Avatar
 
Join Date: Feb 2013
Posts: 114
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 104 Post(s)
Liked: 53
Ok guys, thank you for the tips. I will try them out.
stama is offline  
post #27 of 38 Old 01-10-2020, 04:27 PM
AVS Forum Special Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 1,134
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 382 Post(s)
Liked: 272
Quote:
Originally Posted by stama View Post

Regarding other reports on the accuracy of the i1Pro2 on OLED displays, I have not seen other people posting results from both argyll and some other software which uses the X-Rite driver. So, these differences could be, just that nobody did experiments or reported them. Please believe me, I was very surprised by the large difference in measurements.
Let's see:

115.30 0.3119 0.3288 Argyll adaptive mode, normal resolution
114.25 0.3130 0.3289 Argyll adaptive mode, High resolution
113.92 0.3132 0.3289 Argyll non-adaptive mode, High resolution (-Y A)
113.84 0.3085 0.3263 LightSpace (there's no way to configure the i1Pro2 in LS)

Delta x is about 0.0047, delta y about 0.0026. In x,y space, the width of the spectral
locus is about 0.7, and the height is about 0.8. So the delta's are about 0.7% and 0.3%
respectively. Given the typical inaccuracy, of non-laboratory instruments, that doesn't
seem large to me. As I said, it could be quantization errors. Some careful measurements
might confirm or refute this idea.

Being in x,y space, the visual significance of these differences is very hard to say, since
it is not a perceptual linear or proportional space. If the delta's were measured in L*a*b*
space, then at least the significance would be obvious (because 1 L*a*b* unit is intended to
be a Just Noticeable Difference).

[ The habit of the TV calibration business to use x,y is a bad one. People should use u'v'
at least, since it is closer to perceptually uniform. Using L*a*b* and a modern Delta E
measure would be so much better. ]

Quote:
PS: I had a look in the argyll code, and it seems that it uses a linear interpolation for spectros with less than 5nm half-width for converting the measured spectral response to a 1nm spaced one, and a 3rd order polynomial (Lagrange?) interpolation for those with a larger half-width like the i1pro2. I 've read that the recommendation for such interpolation is to use either the method developed by Sprague for functions having a uniformly spaced independent variable, or a Cubic Spline method for non-uniformly spaced independent variable. And to use the last measured value for any extrapolation (but I've noticed argyll does that already). At least that's what it says here (needs a search in the page for [CIET13805a]). I'm not saying that following these rules gives a more accurate result than what argyll achieves, only that maybe these are the rules that have to be followed to get the same result as the X-Rite software. I should try and save the spectra to a file and try this myself, just have to find the time to do it. I don't know if there is a way to save the spectra from LightSpace too.
AFAIK I'm using the recommendations of the appropriate CIE standards. In any case, I'm quite satisfied that the ArgyllCMS driver returns CIE values that are very very close to those returned by the X-Rite driver - a maximum difference of 0.05 Delta E - a value 20 times smaller than the smallest perceivable difference.

Author of ArgyllCMS and ArgyllPRO ColorMeter
gwgill is offline  
post #28 of 38 Old 01-11-2020, 02:14 AM
Member
 
stama's Avatar
 
Join Date: Feb 2013
Posts: 114
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 104 Post(s)
Liked: 53
Quote:
Originally Posted by gwgill View Post
So the delta's are about 0.7% and 0.3%
respectively. Given the typical inaccuracy, of non-laboratory instruments, that doesn't
seem large to me. As I said, it could be quantization errors. Some careful measurements
might confirm or refute this idea.
The deltas are important because they are not between the i1Pro2 and a reference instrument, they are between two measurements made with the same i1Pro2, in the same conditions.

The "typical inaccuracy, of non-laboratory instruments" is relevant when comparing the i1Pro2 to a reference instrument, not for the case that is presented here.

Quote:
Originally Posted by gwgill View Post
AFAIK I'm using the recommendations of the appropriate CIE standards.
I assume you refer to the CIE 15:2004 "Colorimetry 3rd Edition" which says this:
Quote:
Use one of the four following methods to calculate needed but unmeasured values of <spectral, reflectance, transmittance curves> within the range of measurements:
1) the third-order polynomial interpolation (Lagrange) from the four neighbouring data points around the point to be interpolated, or
2) cubic spline interpolation formula, or
3) a fifth order polynomial interpolation formula from the six neighboring data points around the point to be interpolated, or
4) a Sprague interpolation (see Seve, 2003)
The CIE167:2005 I mentioned is more recent, and there is even a new edition of Colorimetry - CIE15:2018 "Colorimetry 4th edition" (but I don't have this one).

I found however a recent paper that compares the accuracy of the various proposed methods though - "Methods for improving the accuracy of Commission Internationale de L’Eclairage tristimulus values of object color by calculation" - which is open access: link.

Surprisingly, a table of weighting factors gave better results than all the other methods, and has the advantage of being easily computed for any illuminant or observer, since the equation to compute it is provided.

Quote:
Originally Posted by gwgill View Post
In any case, I'm quite satisfied that the ArgyllCMS driver returns CIE values that are very very close to those returned by the X-Rite driver - a maximum difference of 0.05 Delta E - a value 20 times smaller than the smallest perceivable difference.
Erm... am I supposed to understand that the readings I took don't exist?
They certainly don't show argyll and X-Rite's readings to be that close.

Regarding the link you posted, that example is for a nicely smooth wide band spectrum.
LED backlights and OLEDs are not that uniform. Here's a look at the non-uniform green, and the narrowish-band blue in the spectrum of a 2016 OLED taken by the HCFR developer with a JETI a couple of years ago (first screenshot). The second screenshot is the one apparently taken by FSI and LS for their i1D3 WRGB EDR file (with a CR-250 and a 2018 OLED panel?).

It would be great to have this kind of measurement validation between X-Rite's driver and argyll for a i1Pro2 performed on LED sources, too.



Later edit: I've realized this is turning into a "argyll vs X-rite driver" discussion. Between "how close are argyll's measurements to X-Rite's measurements" and "how accurate is argyll vs. a reference probe", the latter is what I am interested in, I don't care at all about the former (and I'd wager others feel the same way). After all, if the i1Pro2 can be made to read more accurate than X-Rite's driver does, I'd say that is much more interesting than reproducing the same level of (in)accuracy as X-Rite.

My observation is that argyll and the X-Rite driver (at least in the way LightSpace is using it) are reading differently. I don't know which is more accurate, as I don't have reference equipment to test this. But if anyone has a reference spectro, it would be great if they can do this test with OLED or LED light sources and share their results.

If the readings are different, I sincerely don't believe something like the interpolation method is responsible, but the way results are retrieved from the probe. I don't think argyll's i1Pro2 readings were tested with a LED light source, which is why it would be very useful to have such a test done against a reference probe (and not against X-rite's driver). Maybe the reasons for the difference in measurements is some side-effect of the way argyll reads data which was not evident for spectra of other light sources (I imagine the tests were carried out on CRTs).
Attached Thumbnails
Click image for larger version

Name:	WRGB OLED - JETI.png
Views:	8
Size:	113.8 KB
ID:	2667934   Click image for larger version

Name:	LightSpace FSI XM55U 23Jan19 (OLED WRGB) spectral graph.png
Views:	6
Size:	73.6 KB
ID:	2667962  
Attached Files
File Type: txt WRGB_OLED_LG.ccss.txt (21.9 KB, 1 views)

Last edited by stama; 01-11-2020 at 11:35 AM. Reason: added the "Later edit" section
stama is offline  
post #29 of 38 Old 01-11-2020, 11:54 AM - Thread Starter
Member
 
Join Date: Dec 2019
Posts: 22
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 18 Post(s)
Liked: 0
Quote:
Originally Posted by stama View Post
I did several tests comparing the measurements between argyll (what DisplayCal uses), i1Profiler, and LightSpace (which is also using the Xrite SDK).

In the case of i1DisplayPro (i1D3), there's no notable difference between any of the above.
In case you are getting different measurements, make sure there are no other things interfering with your measurements. For instance both DisplayCal and i1Profiler have software running in the background which ask Windows to apply a color profile (including loading the gamma calibration from the vctag). It might be you are thinking you are measuring the display as it is, but in fact every value sent to the screen is modified by the currently active loaded gamma in the GPU.

Make sure to disable both DisplayCal's (DisplayCAL-apply-profiles) and i1Profiler's (i1ProfilerTray) when you are taking measurements, and load a unity neutral gamma to your GPU (there is an option in the DisplayCal menu for this or in its profile-setting tray app, or you can restart your computer after disabling these apps) to avoid one of them compromising the measurements you take with the other software.

One of the things you will find nice about argyll/DisplayCal is that you can make a 3D LUT which can either be loaded to your LG OLED TV (if you have one), to an external LUT box, or used in madVR (if you use it as a media player renderer), and that's better than using color profiles and expecting applications to use them (most don't).

PS: i1pro2, on the other hand, give very different measurement results between argyll and X-Rite's SDK on WRGBA OLEDs, but I couldn't say which are better, as I don't have any reference equipment to verify any measurements.

I think Automatic Screen Regulation refers to the feature that the software constantly monitors the ambient luminosity (if you leave your i1D3 plugged in in ambient measurement mode) and does ambient adaptation. That's a feature that no other color calibration software has, as far as I know.
DisplayCAL continue to show different results in terms of white temperature than i1 Profiler.
since I'm using an xrite i1 display pro, I trust i1 profiler much more than display cal.


Quote:
Originally Posted by gwgill View Post
Which would be great, if there was any demonstrable advantage of V4 over V2, which there is not.

Whatever that is !

(Can you explain what that does ?)
automatic display calibration uses the monitor's DDC feature, on most monitors can automatically calibrate the screen brightness to the desired one,
on latest monitors it can automatically adjust both brightness and RGB gains.
I think that this feature is a must have, it does in few seconds what I can do manually with 10 mintues and most often it does it a lot better than me.
Davide Perini is offline  
post #30 of 38 Old 01-11-2020, 07:49 PM
AVS Forum Special Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 1,134
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 382 Post(s)
Liked: 272
Quote:
Originally Posted by stama View Post
The deltas are important because they are not between the i1Pro2 and a reference instrument, they are between two measurements made with the same i1Pro2, in the same conditions.
The point is that you are claiming that the differences are big, when I don't thing they are. There are lots of possible reasons for differences of less than 1%, including display variability, black drift with temperature, quantizion due to limited resolution at short integration times, etc.
Quote:
Surprisingly, a table of weighting factors gave better results than all the other methods, and has the advantage of being easily computed for any illuminant or observer, since the equation to compute it is provided.
You've made a few measurements of a not terribly stable display using the same instrument and two different drivers and noted a small difference, and then ignoring other possible sources of such differences, have made the assumption the ArgyllCMS measurements must be "wrong", and then jump to the conclusion that it is something to do with spectral interpolation algorithms. I don't think any of this can be justified on the evidence so far.

I'll see if I can find some time to do some measurements on the wide gamut display that I now have at my disposal, to see if I can see any systematic differences in driver behavior in emissive measurement mode.
Quote:
They certainly don't show argyll and X-Rite's readings to be that close.
They show them to be within 1%. That's not a big difference.
Quote:
Later edit: I've realized this is turning into a "argyll vs X-rite driver" discussion. Between "how close are argyll's measurements to X-Rite's measurements" and "how accurate is argyll vs. a reference probe", the latter is what I am interested in, I don't care at all about the former (and I'd wager others feel the same way). After all, if the i1Pro2 can be made to read more accurate than X-Rite's driver does, I'd say that is much more interesting than reproducing the same level of (in)accuracy as X-Rite.
The ArgyllCMS driver has little scope for being more accurate than the X-Rite one simply for the reason that it relies on the X-Rite calibration data for the particular instrument. Fundamentally the best it can be is not different to the X-Rite result, with the possible exception of the high resolution spectral mode being slightly more accurate due to throwing away less of the available information, but with the caveat that stretching the X-Rite calibration data to high resolution probably has limitations. [ I personally find the high res. mode valuable for being able to see more detail of spectral structure. I'm unconvinced without more evidence that it is reliably more accurate. ]

Author of ArgyllCMS and ArgyllPRO ColorMeter
gwgill is offline  
Sponsored Links
Advertisement
 
Reply Display Calibration

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


Forum Jump: 

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

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