My primary use of MaxIm over the years has been to operate the camera for imaging. The past has included a vast array of SBIG cameras and using ACP to acquire the images. My processing has been with another software for a good number of years now so now that I'm imaging with a SBIG CMOS camera things have changed dramatically. Processing these images has proven a bit of a challenge. Not calibration or stacking into master frames but rather the color combination. Either I'm missing an important step or something has gone wrong. So I figured since I have licensed copies of MaxIm it would be worthy to see what results I can get in creating the basic RGB image. The steps for this are long forgotten by me I'm afraid so looking for the documentation that describes the step by step process to calibrate and create these master frames and then create the RGB combined image. Does MaxIm do something with this CMOS data that other programs don't? Any pointing to the step by step process that MaxIm used to process these images would be appreciated. Can you take master frames created by another program and combine them in MaxIm is they are .fits files? Or is there something special MaxIm does with the AC4040BSI data that other programs don't/can't?
A simple Ha image of the Ring with 30 3 minute frames taken with a 24" f/7 RC at SRO. While not extremely deep, not sure how this compares to xx exposure w/CCD, it shows a good deal of detail. I am still trying to get better sky flats and upped my flat exposure time from .01 seconds to 3 to see if some of the oddness of the flat disappears. There are clearly 4 quadrants noticed in all exposures, a column on the left side that looks like a chain with a few missing links apparent in all exposures. The flats don't seem to eliminate these affects. I've been using -10 setpoint for the cooler on all frames. What steps will ensure the quadrant will blend into the background? My standard process involves the calibration of all data, alignment of the frames, combining to make masters, then stretching the data. W/ PI I usually do a Masked Stretch at 0.125, then use histogram to bring the dark point up then use curves for the rest. As I said before I use MaxIm mainly for image acquisition through ACP/Scheduler. Suggestions very much welcomed...
We have a new GSENSE Calibration tool, but it is still experimental. For the short term, I recommend the following: Use a large amplitude dither Run the Band Reduction filter on the image Rotate 90 degrees and repeat Band Reduction Rotate back to original location Stack with Sigma Clip or SD Mask The Band Reduction can be done to each individual frames via Batch Process. To make this easier we will be adding the ability for the Batch Process command to remove vertical banding also.
Thanks Doug. I think I now see and understand that the data from this camera should only be processed via MaxIm? In the mean time is there and reason not to take the raw image data with High Gain StackPro? And will High Gain flats work with this data. I mean the flats would be 12 bit while the image data is 16 bit. I looked back in ACP Scheduler and my ROMode file had High Gain StackPro for Normal, Fast, and Flat. My goal ADU for the flats was 1500and in the begining I had set the exposure to 0.01s but after the results with the flats later changed to 0.3s to see if there would be a difference. Is that an issue? I'll admit this move from the STX-16803 to AC4040BSI is major. Being a remote system doesn't help much.... What is this "large amplitude dither"? at this point I have no dither set. I haven't used MaxIm to process data in a very ling time. It's been used for imaging with ACP. So if I'm following this thread I need to set dithering in MaxIm for imaging (suggested value) then calibrate this data in MaxIm 6.50 or above, and continue with the additional steps listed. Just to be sure this dithering is during image acquisition correct? Considering the system: 24" f/7 RC and AC4040BSI What value should the dithering be set to?
You can use whatever software you like to process the images... it's just that we've been developing tools to support these cameras. I do strongly recommend High Gain Stack Pro. I currently have one in my observatory and it's the mode I use. If you take an exposure that's too short for High Gain Stack Pro, it just does a single subexposure, which ends up being identical to regular High Gain. When taking flats you should generate matching flat-darks. Once the flats are dark calibrated those flats will work fine with the longer exposures. I always recommend using dither, in all deep sky imaging situations with all cameras. Dither, in conjunction with Sigma Clip or SD Mask stacking, really helps remove any particularly nasty hot pixels, columns, etc. A normal dither amplitude might be 3 pixels. I'm talking bigger numbers, like 20 to 50. As wide as or wider than the cross pattern.
So I am using High Gain StackPro and should leave the settings in the Read out Mode file in ACP to all High Gain StackPro. I can and have had flats taken at 0.01 and should revert back to that exposure? I've also had auto dark set for the flats, good, bad or indifferent? As for dithering I need to measure the pixel width of those dark bands and use that amount?
Best practice is to use separate flat-darks for the flats, because that allows you to stack the flat-darks. Otherwise only one dark is subtracted, which inherently increases noise compared to combining multiple darks. As for dither, it's a matter of smoothing things out a bit more. Probably 20 pixels will do the job. You might be able to use the typical 3 if you also use the Band Reduction.
Note that Band Reduction is under Filter. Many folks are just learning this feature is there to help with CMOS camera banding effects.
Thanks guys. This is a whole new process for me after 25+ years with SBIG CCD cameras. So at what point should Band Reduction be used? Is it used on darks and calibrated single images?
Band reduction will get rid of the variation in horizontal lines. Done on single images. Order of operations - you'll have to experiment. ADU level is what matters. There is no way for us to guess what your exposure time will be - eg we don't have a "formula" to punch in flat panel wavelengths and intensity, filter transmission at wavelengths, QE of the sensor less the chamber window transmission, and your optical configuration...
The manual says to use the fastest exposure for flats. Initially I used 0.01s because that was the fastest I believe but the banding issue was there so then I tried 0.3s which had pretty much the same results. All were set for an ADU of 1,500 and were started in ACP when the sky brightness allowed either timing to be done. The flats on average are at/near 1,5oo ADU. I was checking to see if there would be a difference in banding and I think they look pretty much the same. The time for exposures to begin are determined by ACP taking exposures every 15 seconds and checking the ADU. Dusk flats star central frames with the darkest filters to the lightest. As the sky brightness changes ACP starts accordingly taking the requested frames. Dawn reverses the process staring with LRGB and the narrow-band if requested.
