Difference between revisions of "AXIOM Beta TestsTODO"
From apertus wiki
Line 1: | Line 1: | ||
#raw12 snapshot of colorchecker passport with | #raw12 snapshot of colorchecker passport with 2010 offset (regs 87/88) gain x1 and x2 - exposure so nothing clips | ||
#grey out of focus wall, | #repeat Hutch and IT8 charts - same framing for both Nikon and Apertus, please! (surroundings do matter, and so does the camera orientation) | ||
#one small video of the out of focus wall (5 frames or so), with exposure on the dark side (around 7 stops below white), gain x2 | #grey out of focus wall, gain x1,x2 with 100 frames of different exposure times (listed below), then run a script and upload the results (no need to upload the raw files) | ||
#raw12 snapshots of real world scenes (outdoors, people, etc.) with high & low gain, long & short exposure times | #one small video of the out of focus wall (5 frames or so), with exposure on the dark side (around 7 stops below white), gain x2 (to check FPN correlation between frames) | ||
#colorchecker passport zoomed in on the bottom half (just the boxes area) | #raw12 snapshots of real world scenes (outdoors, people, etc.) with high & low gain, long & short exposure times | ||
#colorchecker passport zoomed in on the bottom half and defocused a bit (just the boxes area) - for FPN | |||
For all samples: | |||
* Please set the black offset registers (87,88) to 2010. 1910 is too low (clipped shadows = data lost). To get a normal picture, use raw2dng --black=100. | |||
* Please save the register metadata block (on past samples, many files were mislabeled). | |||
For the color charts: | |||
* Please use a static scene (same background in both Nikon and Apertus), and try to take the picture from the same location (they should have the same framing *after* cropping the Nikon image to APS-C). | |||
Exposure times for the gray frames: | Exposure times for the gray frames: |
Revision as of 14:12, 18 December 2015
- raw12 snapshot of colorchecker passport with 2010 offset (regs 87/88) gain x1 and x2 - exposure so nothing clips
- repeat Hutch and IT8 charts - same framing for both Nikon and Apertus, please! (surroundings do matter, and so does the camera orientation)
- grey out of focus wall, gain x1,x2 with 100 frames of different exposure times (listed below), then run a script and upload the results (no need to upload the raw files)
- one small video of the out of focus wall (5 frames or so), with exposure on the dark side (around 7 stops below white), gain x2 (to check FPN correlation between frames)
- raw12 snapshots of real world scenes (outdoors, people, etc.) with high & low gain, long & short exposure times
- colorchecker passport zoomed in on the bottom half and defocused a bit (just the boxes area) - for FPN
For all samples:
- Please set the black offset registers (87,88) to 2010. 1910 is too low (clipped shadows = data lost). To get a normal picture, use raw2dng --black=100.
- Please save the register metadata block (on past samples, many files were mislabeled).
For the color charts:
- Please use a static scene (same background in both Nikon and Apertus), and try to take the picture from the same location (they should have the same framing *after* cropping the Nikon image to APS-C).
Exposure times for the gray frames:
milliseconds: 0.04, 0.05, 0.06, 0.07, 0.08, 0.09, 0.10, 0.11, 0.12, 0.13, 0.14, 0.15, 0.16, 0.17, 0.19, 0.21, 0.22, 0.23, 0.25, 0.26, 0.28, 0.30, 0.32, 0.34, 0.37, 0.39, 0.42, 0.45, 0.48, 0.53, 0.57, 0.61, 0.65, 0.70, 0.75, 0.80, 0.87, 0.94, 1.01, 1.08, 1.17, 1.26, 1.35, 1.45, 1.57, 1.69, 1.83, 1.97, 2.12, 2.28, 2.46, 2.65, 2.86, 3.09, 3.32, 3.58, 3.86, 4.16, 4.49, 4.84, 5.22, 5.62, 6.07, 6.54, 7.06, 7.62, 8.21, 8.85, 9.55, 10.30, 11.11, 11.98, 12.92, 13.94, 15.04, 16.22, 17.49, 18.87, 20.36, 21.96, 23.68, 25.55, 27.56, 29.73, 32.07, 34.59, 37.32, 40.26, 43.43, 46.85, 50.54, 54.52, 58.81, 63.45, 68.44, 73.84, 79.65, 85.92, 92.69, 100.00. exp_time reg: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 17, 18, 19, 21, 22, 24, 26, 28, 30, 33, 35, 38, 41, 44, 48, 52, 56, 60, 65, 70, 75, 81, 88, 95, 102, 110, 119, 128, 138, 149, 161, 174, 188, 202, 218, 235, 254, 274, 296, 319, 344, 371, 400, 432, 466, 503, 542, 585, 631, 681, 735, 793, 855, 922, 995, 1074, 1158, 1249, 1348, 1454, 1569, 1692, 1826, 1970, 2125, 2292, 2473, 2668, 2878, 3105, 3349, 3613, 3898, 4205, 4537, 4894, 5280, 5696, 6145, 6629, 7152, 7715, 8323, 8979, 9687.
Script to be run on those 100 frames:
raw2dng *.raw12 python2 raw12_stats.py