I’m trying to capture high resolution digital photos and videos from my video synth output by pointing my mirrorless digital camera at my small CRT television.
However, I encounter severe aliasing (moire) everywhere; in the electronic viewfinder, on the camera’s back LCD, and when reviewing the images on the computer afterwards. The heavy moire destroys the color and creates interference patterns that change with how I zoom in on the screen.
Does anyone have any ideas on how to deal with this?
Use a LED screen as output instead of a CRT perhaps?
Experiment with capture resolution and frame rate on the camera - the higher the better
This has sometimes got round the vertical scrolling issues when capturing from a crt (at least for me)
I’ve generally had good results with a HD LED TV (which internally handled upscaling from component PAL, although I’ve sometimes experienced some interference… mostly what look like very faint un-synced horizontal ramps… although in my experience it’s been very rare - maybe 10 out of 700 or so captures
This is good info, thanks!
Seeing how the moire might be difficult to overcome with a CRT, I will investigate LED screens.
It’s funny how I can’t really see how much of the moire is actually “baked” into the image and how much is generated by viewing it at a certain magnification on my computer screen. In any case the pattern and severity seems to change whether I zoom at 6%, 12%, etc.
Set your camera to the highest resolution available.
Use manual focus. Slightly defocusing the image can help cure the moire disease.
Set exposure time to at least two video frames. E.g. for interlaced PAL, set your camera exposure time to 1/12 second. Obviously you’ll need a still image on the CRT. If it’s moving, you’ll have problems.
Ultimately I would recommend directly capturing the video to the computer rather than optically rescanning a CRT. Rephotography is really only justified when the display can’t be captured electronically, e.g. vector monitors.
Also, digital viewers definitely can cause moires. The only way to know what you’re really getting is to view the image at 1:1 pixel-per-pixel.