techless wrote: ↑22 Jul 2020, 06:36
I was curious if pursuit camera photos with a 105mm macro lens would be possible. It took some tries
That is some fantastic pursuit camera work!
Normally I recommend that cameras be put at viewing distance of a human head, so that color distortions from LCD viewing angles more approximately simulate a human.
However, there is definite merit to macro work for close inspection.
From inspecting a zoomed version of your sync track (ladder)
-- Visually Estimated Vertical Tracking Error: ~0.2 pixel.
You have relatively accurate vertical tracking error margin (probably ~0.2 pixel over the course of 4 refresh cycles when I zoom into your sync track), given how the screendoor effect is mostly vertically unblurred.
-- Estimated Horizontal Tracking Error: ~0.5 pixel.
I see that the total sync track accumulated misalign is probably approximately ~0.5 pixel of horizontal tracking error margin. Very good for a manually propelled pursuit camera.
Now, what I have not really seen often, is extremely high resolution focussed on all those motionblurred pixels, to see details that are normally below human-visibility noisefloor because they're so tiny! Macro for pursuits is usually overkill, but they do definitely capture details that are fascinating to inspect closely.
______________
Interesting catch I observe:
Now, since you captured so many megapixels with these pursuits, revealing ever tinier motion artifacts that are not visible to human eye (unless you have really good vision). Zooming closer to your images, I notice some interseting digital stepping artifact (see the faint vertical bands)
- vertical-bands.jpg (380.01 KiB) Viewed 23979 times
Very subtle. I wonder if this is some kind of ultrahighfrequency LCD PWM that is not 100% DC-filtered. It could also be a very subtle (nearly invisible) digital camera sensor sampling artifact, but I know some monitors PWM-free backlights have some ultra-faint PWM from insufficient capacitor filtering (etc). This would not be visible to most human eyes, though. If you used a motionspeed of 960pps, it looks like a possible 1000Hz PWM that was (imperfectly) capacitor filtered into PWM-freeness:
