HDR PQ Test Pattern - 0 to 10 000 nits in steps of 1% to calculate EOTF accuracy

Channel:
Subscribers:
472
Published on ● Video Link: https://www.youtube.com/watch?v=_9oKaf_mfn8



Duration: 0:09
8,465 views
20


I created this test pattern in order to be able to profile the accuracy of HDR TVs and monitors without paying $3000 for Calman licence, and to compare difference in EOTF tracking between various sources connected to the same display. For example, it's currently impossible to use Calman to measure HDR output of Windows, so this manual approach can be used instead. Please let me know if you're aware of an easier way to do this.

To use this test pattern, place colorimeter at the center square and run a measuring program on a secondary screen to continuously read values (I recommend hcfr: https://sourceforge.net/projects/hcfr/ ).

Every other frame is a black frame to prevent OLED auto dimming on static content, and because I noticed my QD-OLED (at least) displays somewhat different brightness if value moves from eg 74% to 75% brightness than it does when value moves from 0% to 75%. Black frames ensure that all values are compared on equal footing, and not being biased by the previously shown value.

This pattern is supposed to be used and navigated one frame at a time by pressing , and . keys. I redownloaded this pattern to check how much youtube compression affected it, and it's mostly unaffected but some values do deviate very slightly, for example: 32.8 nits is sent to the screen instead of 32.4479, or 0.134 instead of 0.132. The difference this makes in plotted EOTF graphs is not visible to the eye, but if you desire perfect accuracy you can download the near uncompressed (DNxHR 12 bit .mov file) video file here: https://drive.google.com/file/d/1I7JyU2H2fea6fA5V85lxtwCgHoLbZ9Kt/view?usp=sharing

I sized this pattern to be used with 55" screen. It's a 0.38% window. If you find the square too small to use on a smaller screen please leave a comment and I'll make a version of this video with the increased area.

This is the google sheet I created to be used with this test pattern with the purpose of creating EOTF graphs: https://docs.google.com/spreadsheets/d/18IGWDipM6Im5l22ihRl8FMRbFqykcBuCpOccSwV7yD0/edit?usp=sharing

PQ nits (cd/m²)
0 0.0000
0.01 0.0023
0.02 0.0087
0.03 0.0198
0.04 0.0365
0.05 0.0600
0.06 0.0914
0.07 0.1320
0.08 0.1833
0.09 0.2469
0.10 0.3246
0.11 0.4183
0.12 0.5303
0.13 0.6630
0.14 0.8189
0.15 1.0011
0.16 1.2127
0.17 1.4572
0.18 1.7386
0.19 2.0610
0.20 2.4293
0.21 2.8484
0.22 3.3239
0.23 3.8622
0.24 4.4698
0.25 5.1542
0.26 5.9234
0.27 6.7863
0.28 7.7525
0.29 8.8326
0.30 10.0382
0.31 11.3819
0.32 12.8775
0.33 14.5401
0.34 16.3861
0.35 18.4336
0.36 20.7022
0.37 23.2134
0.38 25.9906
0.39 29.0594
0.40 32.4479
0.41 36.1865
0.42 40.3086
0.43 44.8505
0.44 49.8521
0.45 55.3567
0.46 61.4117
0.47 68.0689
0.48 75.3849
0.49 83.4213
0.50 92.2457
0.51 101.9318
0.52 112.5601
0.53 124.2188
0.54 137.0039
0.55 151.0206
0.56 166.3842
0.57 183.2202
0.58 201.6663
0.59 221.8730
0.60 244.0052
0.61 268.2431
0.62 294.7842
0.63 323.8448
0.64 355.6618
0.65 390.4947
0.66 428.6282
0.67 470.3742
0.68 516.0750
0.69 566.1060
0.70 620.8794
0.71 680.8477
0.72 746.5079
0.73 818.4061
0.74 897.1427
0.75 983.3779
0.76 1077.8378
0.77 1181.3220
0.78 1294.7106
0.79 1418.9731
0.80 1555.1784
0.81 1704.5047
0.82 1868.2520
0.83 2047.8554
0.84 2244.8994
0.85 2461.1351
0.86 2698.4983
0.87 2959.1303
0.88 3245.4014
0.89 3559.9366
0.90 3905.6447
0.91 4285.7513
0.92 4703.8357
0.93 5163.8717
0.94 5670.2746
0.95 6227.9535
0.96 6842.3706
0.97 7519.6074
0.98 8266.4406
0.99 9090.4270
1 10000.0000