AVS Forum banner

301 - 310 of 310 Posts

·
Registered
Joined
·
1,533 Posts
Cutting my teeth with this i1 pro 2 and took some measurements on my laptop and got some weird results when I plugged them into Leon's spreadsheet. When I compared my profiled i1d3 measurements with the i1pro2 .bpd from CS, things are nearly perfectly aligned.

Is this right or am I missing something in how Leon's spreadsheet works?

See below:

Note how the reference meter readings on the left "Reference Meter Readings" do not match the "Reference Meter" values under "Profiled Meter Comparison on the right: : Red x is .6062 on the left, but .6768 on the right.
View attachment 3108030

Meanwhile, when I input the data into my own spreadsheet, it's nearly perfect.

View attachment 3108029

Hi,

I'm so very sorry. When I created the sheet the other night for ZRO users (after a 14 hour work day, I was shattered 😝 ), I forgot to link the public sheet to the external calculation sheet, which means it did not import your BPD data correctly (effectively comparing you results against dummy data).

I have corrected this now, here are you results

3108200


Feel free to revisit the provided link to verify this for yourself, and sorry for the confusion and inconvenience.
 

·
Registered
Joined
·
192 Posts
Hi,

I'm so very sorry. When I created the sheet the other night for ZRO users (after a 14 hour work day, I was shattered ), I forgot to link the public sheet to the external calculation sheet, which means it did not import your BPD data correctly (effectively comparing you results against dummy data).

I have corrected this now, here are you results

View attachment 3108200

Feel free to revisit the provided link to verify this for yourself, and sorry for the confusion and inconvenience.
All good, and thanks for doing this! Glad there wasn't more math I was missing.

Also, don't feel bad -- I posted a .csv file the other day for other users that was missing a whole patch value. I had no idea until someone tried to use it and pointed it out. Part of being in a community of volunteer enthusiasts!


Circling back -- was pleased that I got an immediate probe match with contact mode. Hopefully I can reproduce on my 75" 950H, which may pose more of a challenge.

Will begin working on figuring out a setup for that today.

Sent from my Pixel 2 using Tapatalk
 

·
Registered
Joined
·
1,533 Posts
All good, and thanks for doing this! Glad there wasn't more math I was missing.

Also, don't feel bad -- I posted a .csv file the other day for other users that was missing a whole patch value. I had no idea until someone tried to use it and pointed it out. Part of being in a community of volunteer enthusiasts!


Circling back -- was pleased that I got an immediate probe match with contact mode. Hopefully I can reproduce on my 75" 950H, which may pose more of a challenge.

Will begin working on figuring out a setup for that today.

Sent from my Pixel 2 using Tapatalk
Doing it this way, there isn't really any math involved other than to compare one set of readings to another. The formula to create the actual corrections is a different thing though.

For full clarity, using the method you have above, all you have done is take two sets of readings with your i1 Pro2 and i1d3, with that correction applied in software, you have then taken a final set of readings to compare to those taken with the spectro. Which of course is totally fine, and shows the correction is good. However, if you calculate the actual correction between the two probes, it is (and always will be) perfect when using FCVM.

The correction is hidden in this image but the outcome is an accurate result of the correction applied to your i1d3 meter BPD readings

3108224


The differences in the two sets of results are a consequence of display and/or probe instability and is normal.

And for completion, here is FCMM (with luma correction). Again, the matrix is a direct calculation of your BPD readings.

3108225


On your display type, you should always get a good correction. It is with WRGB OLED where people have issues.
 

·
Registered
Joined
·
892 Posts
Hi,

I'm so very sorry. When I created the sheet the other night for ZRO users (after a 14 hour work day, I was shattered 😝 ), I forgot to link the public sheet to the external calculation sheet, which means it did not import your BPD data correctly (effectively comparing you results against dummy data).

I have corrected this now, here are you results

View attachment 3108200

Feel free to revisit the provided link to verify this for yourself, and sorry for the confusion and inconvenience.
Where can I download this spreadsheet?
Thanks,
CFC
 

·
Registered
Joined
·
1,533 Posts
Where can I download this spreadsheet?
Thanks,
CFC
It is not a downloadable spreadsheet, it is an online “ Google Sheets” spreadsheet that was created specifically for ColourSpace ZRO users to verify their FCVM probe match.

Although the spreadsheet has been set to full public access, I haven’t published the link yet. I will publish it in the ColourSpace ZRO thread at some point in the near future.
 
  • Like
Reactions: CFC

·
Registered
Joined
·
1,723 Posts
Can anyone point to a definitive source document for the NIST tolerance values? I'm curious how they were arrived at. For example, if I take the Blue BT.709 primary for 100 nits White, and compare it with x+.001, y+.001, Y-1.5%, I calculate a dEITP of 2.8, which seems pretty high.
 

·
Registered
Joined
·
1,723 Posts
OK, I've found and read NIST Colorimetric Calibration Facility for Displays - Part 2 from 2001 and I've also purchased and read the prior NIST Colorimetric Calibration Facility for Displays from 1999. From the 2001 abstract: "the facility is capable of measuring colors of sources with an expanded uncertainty 0.001 for chromaticity (x,y) and 1% for luminance (Y)". That's referring to the calculated uncertainty (expanded, k=2) in the measurements obtained from their reference spectroradiometer for a target display. (The facility does FCMM profiling for a target meter and display.) In the body of the paper, they actually show separately calculated uncertainties for White, Red, Green, and Blue, but oddly, for LCD displays I only see uncertainty values for x and y, not for Y; they do show uncertainty values for Y for CRTs. (The 1999 paper has uncertainty values for Y for LCDs, but the 1999 values are all standard uncertainty, not expanded uncertainty.) Although the abstract says .001 for x and y, the body of the paper actually shows .0011 for y for White for LCDs.

But, these are uncertainty values for their spectro measurements, they are NOT tolerance values for FCMM comparison measurements. (And perhaps the uncertainty has improved after 20 years?) So, is there some other reference that actually specifies tolerances, or have those uncertainty values been misapplied, or have I not understood why they validly translate to tolerances?
 

·
Registered
Joined
·
1,723 Posts
I contacted Dr. Yoshi Ohno, one of the authors of those papers, now a NIST Fellow, and he kindly informed me that the display colorimetry project was stopped around the time of those papers. He said he was not aware of any NIST publication about expected or acceptable differences for FCMM measurements.
 

·
Registered
Joined
·
227 Posts
Hello,
even though there are already some ways to calculate correction matrices, I would like to point to another post on the subject.
Preview Probe Profiler2
I am not sure if there is still interest in such solutions at all. :unsure:
Peter
 
  • Like
Reactions: ConnecTEDDD

·
Registered
Joined
·
1,723 Posts
I hope you are changing the label on the Pass/Fail boxes and changing the criteria to something more reasonable, or eliminating them altogether.
 
301 - 310 of 310 Posts
Top