I have a STC-7 (w/SC@ & AO-8a). A year or so ago I was having trouble with vertical banding. I don't recall what I did, but it disappeared. Well now it's back. I'm using Medium Stack pro, Bin1, Cooler set @ -10. 300 Sec exposure for NB, 300 Sec dark frames, Flat Field frames. I have tried processing with & with Bias frames and it makes no difference. I have made all new calibration files ( 36 frames of each )in the last week. I have no clue on what to look for.
Wow, whatever that band is, it's about 1/4 of an ADU on the raw data. It appears to be something that doesn't subtract perfectly for whatever reason. I'm going to have to assume this is something the sensor itself is doing, because that doesn't correlate with anything that happens the camera electronics. Currently, the only thing I can suggest is making dark frames under the exact same conditions, adjacent in time to when you take the light frames. Maybe that will help. There is something coming in MaxIm DL 6.40 that might also help. We have added a "band reduction" command, which is useful for removing the horizontal banding noise that appears on some CMOS cameras. I just used it on vertical banding by rotating the image 90 degrees, performing band reduction, and then rotating it back. Yeah it's not intended for that purpose but it seems to have done a pretty good job.
Thank you Doug for your attention on this. If I need to take darks in the same imaging run, how many darks frames are required to get a decent subtraction? "There is something coming in MaxIm DL 6.40 that might also help. We have added a "band reduction" command, which is useful for removing the horizontal banding noise that appears on some CMOS cameras. I just used it on vertical banding by rotating the image 90 degrees, performing band reduction, and then rotating it back." Can this be done on the final combined image, or does this need to be done on individual images pre-combine? "Wow, whatever that band is, it's about 1/4 of an ADU on the raw data. It appears to be something that doesn't subtract perfectly for whatever reason. I'm going to have to assume this is something the sensor itself is doing, because that doesn't correlate with anything that happens the camera electronics." Would this be something that could be solved by doing something like an "Electronics tuneup" on the camera?
It could be done either on the individual (calibrated) images, or on the final image... basically you will have to experiment to figure out what works best. Please note that it's easy to do hundreds of images in a batch - that's exactly what the Batch Process feature is for. As for the camera, I'm afraid there's nothing to tune up. All CMOS sensors have some quirks when you push them hard enough. CCDs have always had the advantage that every pixel goes through exactly the same processing, so their image quality is hard to beat. I think that's something that's been lost along the way. The CMOS sensors are great in other ways. Nothing is perfect; everything is a compromise.
I just looked back a little ways, and in August, I was shooting in "High Gain Stack Pro" Bin1 and Bin2 and there was no banding. The images posted above are in "Medium Gain Stack Pro". I will next try a sample High Gain Stack Pro. But naturally, the 10 day forecast is for clouds. Will post the results when I get some.
The sensor will indeed perform differently at different gain levels. I'd recommend using High Gain Stack Pro because the High Gain part gives you the best SNR but limited dynamic range, while the Stack Pro part gives you high dynamic range.
I looked on your web page for the current Maxim version number, but I did not see it. I the 6.40 currently available now?
Thank You Doug. Banding update: I got a successful test using "High Gain Stack Pro" on both Bin1 & 2. I did make up a new calibration library for the test. I didn't change any procedural steps during the test. I might retry "Medium Gain Stack Pro" to see what I get.
