>What is the SQM that you image from?
It should be around 20 I guess, I can see the milky way when conditions allow, but nearer the horizon it isn't very good. There are bunch of extremely bright streetlights in front of the house that are very annoying, and some towns within 10km radius. For Dutch conditions, I'm lucky though.
>So far I had no luck with using planetary stacking software for deepsky, but I have to try Autostakkert; seems to be a good alternative.
Be careful though, because it can be very tricky and is rather limited at the moment. Here are some tips I wrote down today; just a preliminary list, I might turn this into a small article for my website later on.
I'm assuming you are using planetary imaging camera's (where you can set brightness, gamma, gain, exposure, etc... )
- Make sure the camera it self is not operating in high speed mode.
- Always use a lot of 'brightness' when recording (I always maximize this actually, and get rid of the extra background brightness later on).
- Never use any gamma when recording (disable this, or set it to neutral - for ASI this is typically 50).
- Don't use gain above the point where read-noise is not getting lower anymore (about 25dB I think on the ASI1600, similarly for the ASI174 and Point Grey blackfly for example).
- There should be some (very limited) movement (dithering) in your images; I get this for free due to poor tracking.
- I use FireCapture - free, good capturing software. I record in SER format. I find this very easy to use and process. Don't use AVI format preferably.
- IF you use higher gain values, it typically doesn't matter at all if you use 16 (12) or 8-bit recordings, almost all of the extra bits are noise anyways. I pretty much always stick with 8-bit.
- Use the latest beta version of AS!2 (2.1 is terrible, 2.3 too, there are some bugs in the beta versions, but generally speaking they should work very well)
- Use a master dark at the very least (build it from a lot of dark frames inside AS!2, image calibration -> create master frame. The more darks you use, the better. But this is less critical if you have a bit of dithering in your light frames. Always try to use close to the same temperature and of course same exposure time and exact same other camera settings)
- Use column or row noise removal, but ONLY if the camera has that particular type of noise (and ONLY if the target is not too large, the image has to be rather 'flat' for it to work properly, no gradients, no object filling the entire FOV. You might tweak the settings a bit under advanced menu).
- Look at the light frames in AS!2 to see if calibration is done properly; you can increase the displaying brightness for example in the frame view in AS!2. This does NOT affect the image stack at all, it is just for inspection. Don't use the scaling option for this though, as this DOES mess with the actual data.
- Use surface alignment mode
- Place alignment anchor (to perform the coarse stabilization) onto a feature that is always visible (a bright star shoulw work, place it on top of the bright star, possibly you want to decrease the size of this alignment anchor - using numbers 1 to 9 on the keyboard, try with 3 or 4, but it depends on image scale, so simply try it out).
- (there should not be field rotation.. currently not handled properly. You CAN rotate frames - for example have it go from 0 degrees for the first frame to 0.2 degrees for the last frame, but near the edges it will introduce some artifacts. It does however work, so if you know what you are doing, and you know the actual field rotation, you can use this option)
- Choose expand for maximim image sizes (but less signal at the edges) or crop for minimum image size but with the same signal all over.
- Click analyze to run the image stabilization
- Always check after analyzing the images if the image stabilization worked. This is essential. Simply browse through the frames using the frame slider, there MUST be no jumps visible. If it does not look good, place the alignment anchor somewhere else and/or change the size.
- If coarse alignment is succesful, continue..
- Place an alignment point - on top of a NOT over-exposed part of the image, a not so bright star perhaps, or the deepsky feature itself.
- Use local quality estimator (default) to use the area of this particular AP to also judge the quality of the images (that is why this part should not be over-exposed)
- Decide how many frames you want to stack (typically a lot, but if many are blurred a lot, use less). You could combine multiple stack-sizes later on. Just experiment a bit and see what happens.
Then post-processing; you're on your own for now. This takes most of the time; stretching the stacks is important, keeping track of all that you do also. Only at the very end think about some (selective) sharpening or noise reduction, never before. Perhaps artificial flats to get rid of any left-over gradients (I always do this), etc... You can go crazy here, and it can really make or break the image. I typically work in Photoshop and just have some hours of fun with this. Occasionally I go back to the stack to see if messed up something or if - when I stack with some different parameters - I can get better results. There are many ways you deal with this, but essentially it comes down to:
- some kind of digital development to get both the brightest and darkest areas to still show details
- some further calibration (artificial flats - I often do this after some initial brightening so I can see what I do better.... )
- followed with perhaps some selective sharpening and at the very end perhaps some noise reduction
- sometimes I combine different stacks, which involves opening all stacks as layers in Photoshop, manually aligning them, and then changing the opacity of the stack layers in such a way that I get the lowest noise in the sky background. Typically I do this after the digital development stages.
Oh. And have fun of course. That is very important