3 ms·

How long are digital exposures? It seems like most of the interference could be removed with a bit of signal processing.

• terramex 6y ago
  If you take long exposure of dim object, bright satellite passing in front of it will saturate signal and there will be nothing to recover. To avoid that you need to track all the satellites and split one long exposure into multiple shorter ones - but that multiplies read noise accordingly.
  • system2 6y ago
    This is why I take 500+ images and process them. I don't think the moving satellites would cause any harm to long exposure deep space photography. I agree if you take 30 pictures, it could be a problem.
    • terramex 6y ago
      > This is why I take 500+ images and process them. If you do 500 exposures you get 500x read noise of the sensor mixed with signal. Modern CCD and CMOS sensor have low read noise (compared to thermal and shot noise) but not zero. For scientific observations of faint objects long exposure times on order of hours are necessary, for amateur astrophotography you can just go with multiple exposures.
      • mkj 6y ago
        Thanks for the explanation, I didn't realise there was a per-exposure noise component.
• antognini 6y ago
  A deep image will typically consist of many 20-minute exposures. You can't have the exposure be too short because the CCD read-out process introduces noise, so you want as few read-outs as possible. But if the image is exposed for too long parts of it may saturate and start to ruin the rest of the image. (Plus you may have the whole thing ruined by a cosmic ray or a satellite or something.) Something around twenty minutes is generally the sweet spot. Throwing out an image or two that had a satellite or cosmic ray pass through out of a few dozen isn't so bad. But if every single one of the images has a satellite trail you may be out of luck.