I'm getting an odd, wavy pattern in my new CCD47 images. These bands show up in my flat and light frames, but not the bias or dark, so I've ruled out the cables/etc. and narrowed it to the optics. It seems to show up whether or not I use filters (LRGB or empty), so it seems to be the sensor. Temperature doesn't seem to have an effect, either. I've read up a bit on etaloning and other optical issues, but this looks different than the other images I've seen posted. I'm at a loss. I've attached an example. Any thoughts?
To me, it looks like an etaloning effect combined with something else, possibly the light source is modulated or there is IR getting in. Some questions: 1. Can you describe your optical setup? a) e.g. specific make/model and optical parameters of the equipment - eg telescope, any focal reducer/field flatenner ? b) what is the light source for flats (eg is it a flat panel, make/model, power source, is it dimmable/PWM modulated)? 2. Any infrared light sources in the observatory (eg security camera)? 3. Have you tried doing a white t-shirt flat or twilight flat or white paper stack flat with daylight or halogen/incandescent bulb? 4. Filters - I know you mentioned using an empty slot, however you say it shows up with filters a) What exact filters do you have? b) are they edge blackened? c) for the RGB, are they IR rejecting? d) are you sure the right filter slot is being selected? 5. Have you tested it with MaxIm LT/MaxIm DL?
Justin, nothing to worry about. Attached is a flat with my Aluma47-10. Looks the same as yours. I think it's the structure of the chip itself is what we are seeing, anyway it calibrates out.
These are back-illuminated sensors. They are initially built the usual way, adding layers of semiconductor, insulator, and conductors... and then then flipped over and bonded to a temporary base wafer. The sensor is then ground down, polished, and chemically etched until it is just 10 - 20 microns thick. When done it's removed from the base wafer and mounted into the final package. It's a difficult process. If the sensor is too thick it won't be sensitive enough, and if it's too thin... well you have no sensor left! Yields are lower, which is why these sensors are so expensive. Inevitably the thickness of the sensor isn't perfectly uniform, and the polishing marks become visible when you take a flat. As long as they calibrate out, all is normal.
