Originally Posted by JustaPlacebo
In rec. 709 SDR, these sets are so accurate you will be hard pressed to notice even the most thorough calibration. For the HDR standards, especially rec.2020, calibration may help but not dramatically, IMO. (Based on all the reviews I have read and experience over the years with pro and consumer OLEDs.) If you have an X-rite i1 Display Pro or Spyder 5 Elite, either should allow you to calibrate for 120 nits in SDR and basic 2-point white balance. Though those pucks are too slow to create and apply an ICC profile or LUT.
Hi, it's not available in consumer market any display that it's accurate with it's factory default settings using it's most accurate mode (like Movie or ISF), that kind of displays are available in pro market only, a good example are the FSI displays which are coming pre-calibrated with certification of performance, so they pre-calibrate them using 3D LUT display characterization generated from LightSpace and they are using CalMAN to verify the results.
In consumer world, usually magazine/reviews are taking only a few measurements to validate the performance of the display in color reproduction, even a calibrated C7, while at high luminance range the errors are low, if you measure different luminance levels you will see that errors are will be increased. That kind of errors obviously can't be corrected using internal calibration controls; this is why a lot of LG OLED users are using 3D LUT Boxes like eeColor to improve their color reproduction in SDR with low cost.
C7 - 10-Point Saturation @ 100% Luminance Level:
C7 - 10-Point Saturation @ 75% Luminance Level:
C7 - 10-Point Saturation @ 50% Luminance Level:
C7 - 10-Point Saturation @ 25% Luminance Level:
Credit for measurements to John (jrref).
Someone will say that sites telling that dE below 3.0 are not visible, this is not true.
dE of 1.0 is the smallest colour difference the human eye can see.
So, theoretically any dE less than 1.0 is imperceptible, while any dE greater than 1.0 is noticeable. However, some colour differences greater than 1 can be imperceptible, while some colour differences below 1.0 can be very visible, depending on the colour being measured.
If you calibrate your display and use the most popular DIY meters like iD3 + i1PRO, and then take your Grayscale+Saturations readings that got you average dE 0.5 for example, this means only that you are good versus your current meters.
So you have to add sometimes +1 / +2 dE or more because this is the average difference or your i1PRO compared to a better spectro like JETI/PhotoResearch.
Also same dE color difference between two colors, for example comparing 2 yellows and two greens may not look like the same difference to our eye.
Can you easily spot these errors using the color comparator feature of CalMAN.
For examle this is a CAlMAN Color Comparitor Screen Capture that one users has posted, CalMAN Reported these errors:
White: 0.3 dE2000 / Red: 1.3 dE2000 / Green: 2.1 dE2000 / Blue: 3.1 dE2000 / Cyan: 0.9 dE2000 / Magenta: 1.1 dE2000 / Yellow: 0.7 dE2000
and the above Color Comparitor Screen shows these errors...
But PC Monitors are not the ideal screens to make color comparisons of the different spot of two color because of they angle color shifting problems, but i'm moving this Color Comparison to the Next Level that will make all see the difference clearer to all
I made some AnimaTED Gifs with Color RGB Triplets taken from these Data's that CalMAN has at these small Color Comparator Boxes. Actual & Target Boxes.
I made 2 different 300px x 300px Full Color Images for each Color, One Frame with RGB Value from Your Actual Measured Color and another Frame with RGB Triplet Information from the Target Color,
All these Data's are coming from CalMAN Image Comparator Boxes.
You can verify them if you like, I did that already with screen capturing the AnimaTED GIF Frame Files to see if the values are matching with these at your picture.
So we have:
White Color Difference between 2 Frames of 0.3 dE2000
Red Color Difference between 2 Frames of 1.3 dE2000
Green Color Difference between 2 Frames of 2.1 dE2000
Blue Color Difference between 2 Frames of 3.1 dE2000
Cyan Color Difference between 2 Frames of 0.9 dE2000
Magenta Color Difference between 2 Frames of 1.1 dE2000
Yellow Color Difference between 2 Frames of 0.7 dE2000
You can spot the Red 1.1dE or the Cyan 0.9dE..... Green 2.1 dE or Red 1.4dE also, so 3.0 dE is not the limit of Human Perception.
1.0 - 3.0 dE Difference is visible but it has to do with color you are comparing, some low dE differencies can be more or less noticed.
BTW Spyder 5 is the worst example of meter as a choice for any kind of serious color application.
When you buy a new Spyder 5 meter, it's color accuracy is like playing a lottery, so even new meters have large unit-to-unit differences to the measurements, this is why they are not recommended from calibration software companies. (You can fix the color accuracy issue if you create a 4-color matrix meter correction table using a software that supports that (not Spyder software).
Here you can see Tom's Huffman (ChromaPure) test: New Spyder 5 vs. i1Display Pro / ColorMunki Display
LightSpace and CalMAN also are not recommending to use Spyder meters for accurate calibration work.
As you can see from the following picture (the crosses away from the REC.709 triangle of the CIE Chart are bad/failed readings) from a test Steve Shaw performed (CEO of LightIllusion) show the poor performance of Spyder 4/5 meters where they fail to read properly dark colors.
About your comment about slow speed of i1Display PRO for 3D LUT with OLED, using LightSpace HTL and taking 17-Point Cube measurements (about 5000 measurements) it takes about ~3 hours with 1 sec of delay before each meter read and 0.75sec meter exposure time. It will require less time if you set the delay to 0.5 sec; you will save about 30-40 minutes.