A comparison between four different combinations of gain and shutter speed when imaging Jupiter.
Since I obtained my ZWO ASI224MC imaging camera earlier this year, I've always used a shutter speed of 10ms and a gain setting of 280 when imaging Jupiter based on what others have done. While those settings seemed to work well overall, I wanted to try faster and slower shutter speeds and corresponding higher and lower gain settings. The example above shows the four setting combinations I used. All images were produced by stacking 40% of the frames captured in a 90-second video in AutoStakkert 2 with wavelet sharpening (and RGB align and balance) in Registax using the same settings in all four cases. I tried to keep the histogram around 80% in the green channel, and so the gain settings were adjusted in order to keep the histogram levels even in all four cases. Note that data was taken through a Celestron 8" EdgeHD telescope and a Televue 2.5x Powermate.
I tried the following four setting combinations:
1) Shutter speed 5ms and gain of 340
2) Shutter speed 10ms and gain of 280
3) Shutter speed 20ms and gain of 235
4) Shutter speed 30ms and gain of 180
Since all four captures were 90 seconds, a faster shutter speed results in more frames, but each frame has more noise. Since the seeing was only average, I stacked about 40% of the frames. Despite the 5ms shutter speed giving more frames to stack (8504 vs 2929 at 30ms), the 5ms shutter speed image stack still had the most noise. On the other hand, the 30ms shutter speed stack blurred details the most out of the four setting combinations.
The bottom line is that shutter speeds of 10-20 ms seemed to produce the best overall image, in-line with what I was doing previously and what I had seen others do. The 20 ms shutter speed seemed to give similar results, but variations in seeing may show more differences. In general, faster shutter speeds should produce a better final result when seeing is poor, so your mileage may vary depending on conditions.
You can watch a video discussing the comparison below, which includes a sample of the raw data.
Comments