SOLVED: Driver update changed the default offset for the camera from 0 to 8 - lead to darks not letting flats calibrate correctly.
Hi all,
I have just started working with narrowband filters on the ASI294MM. I am having a problem where my flats are over-compensating for some artefacts on the sensor instead of properly calibrating them out.
You can see in one of the original subs (img 3) there is some uneven atrefacts, especially down the bottom left/right. In the master light (img 2) you can see that this has been made lighter, but much too light! You can see what the master flat (img 1) looks like, with correctly shows the uneven surface.
I am using the WBPP script in Pixinsight, and have tried tweaking all kinds of settings with no luck in getting the calibration correct.
Any help would be GREATLY appreciated.
Last edited by BlakPhoenix; 08-03-2021 at 09:38 AM.
try DSS instead and see if that renders the same result. Pixinsight is not as good at stacking as the other stacking programs. It does everything else really well.
You need to scale your flats with a bias frame of equal duration taken at the same temperature. This has been discussed many times before so a search on the forums should give you all the info you're looking for.
I use them all and DSS asks for them all, despite what gets said from time to time.
Lights .... obvious but shoot to a consistent temperature
Darks..... same time and temp as lights with cap on , I use 30
Flats..... I use one of the light boxes exfo sells here on the forum... really well made and adjustable more info later.
Dark Flats ..... same length as Flats with cap on
Bias .... fastest exposure camera can do with cap on
Things I’ve found useful
If doing darks during the daylight hours check absolutely no light leaks get in, the tiniest amount will cause a problem
Flats , the light boxes are a treat, because some color filters have different responses, you can adjust the intensity to allow the same exposure for all filters, in my case as I have a shutter I use 3 seconds, and adjust the intensity to keep histogram in same place, I’m a big fan of keeping things as equal as possible this way I can use the same dark flats for all the filters.
DSS creators made it so all could be used, I do and don’t have too many issues other than the odd cosmetic pixel getting through.
Make sure your flats are well into the histogram, so space at either end,
As with all these things consistency is the key, same temperature, gain, etc for everything, don’t mix and match stuff,
CHECK THE FITS HEADER ON YOUR CALIBRATION FILES perhaps gain offset settings are different, some capture programs require you to set them each run (QHY does)
If you’re doing Tshirt Flats they will likely be different as you can’t control the light well, if this is the case it’s seperate dark flats for each filter, don’t mix and match.
Try the Deepsky stacker program, it’s free and I’ve never needed anything else.
This is using darks and flat darks too. The flats are 30 seconds long, flat darks are also 30 seconds. The light frames are 240 seconds and the darks are also 240 seconds. All images taken at -10C, 120 gain & 2x2 binning on the 249MM. The flats are taken using a flat pannel at night, so no t-shirt method or other.
This is all done in NINA, with the exact same settings between image types. I will try DSS and report back. Hoping it's just a weird setting in PIX.
Gave this a go in DSS and am seeing the same issue (see below image). I also took a screenshot of the histogram, I believe this is exposed correctly for a flat?
Gave this a go in DSS and am seeing the same issue (see below image). I also took a screenshot of the histogram, I believe this is exposed correctly for a flat?
Could you please dropbox one light sub, your master flat + dark flat, dark (if any) and bias? Will have a look. (all FITS format)
Send stuff off for a second opinion always a good idea
The only other thing that comes to mind, is maybe it iced up sometime during the process and it’s given you a false reading around the edges. Maybe you can run the camera for a while and check , reshoot the calibration then whiz a torch down the tube at the end.
Ok, I figured it out. I knew I had to be doing something wrong somewhere along the way. Turns out the darks / flat darks when I took the darks I had an older version of the camera's driver installed that defaulted to 0 offset, where as these new images used an updated driver that defaults to an offset of 8!
Thanks Alchemy for your suggestion to recheck all my calibration files. I looked into the headers of each and it was staring right at me!
Just goes to show that I need to check all my setting after every software change. I thought the issue was because of the new narrowband filters, but instead it was all software related!
Took some new darks/flat darks overnight and boom, perfectly calibrated frames!
Thanks to everyone for their help and big thank you to multiweb for your preparedness to even look into my fits/exifs.
Assuming the 294MM behaves the same way as the 294MC, don't use bias frames. They don't behave as expected.
With my 294MC, I used lights and darks at matching temp, exposure, gain and offset, then for flats I used a dim light panel to get ten second exposure times for a good flat frame, and ten second dark-flats at matching temperature, gain and offset.
If you do not apply a master dark as well as the master flat, the dark current (Which is relatively even) will be accentuated in the corners where vignetting is being corrected by the master flat. Every time I use a flat/master flat on a light frame without a master dark I get something similar to what you posted in the first post.
