Hello. I recently noticed that my calibration frames have become very different from the ones I used to make a few years ago. For example, I am attaching master images from 2016 and recent ones. I also attach new single and dark images. I don't like the increase in std dev from 6 to 14 on master bias images and from 17 to 77 on master dark images. Is this due to camera sensor degradation or could there be some other reason? Power supply is used original, driver version is on the screenshot. The firmware version is 2.63 Thanks for your attention and your possible advice.
I had a fast look. Serial number of camera? Do you have original images from 2016? What software was used back in 2016 to acquire the images? What software was used back in 2016 to stack them to make the master darks/bias frames? I'm looking at the NINA images, and wonder if they were read out fast, which can introduce noise. e.g. I'm assuming you were using Cedric Ragenaud's ASCOM driver, or maybe something you used. How about trying good old CCDOps 5.66 to acquire the bias and darks today, and compare?
Looking at the image, I can see evidence of pattern noise. That is typically caused by electrical noise on the power supply. It can be caused by crappy external power to the camera, but it can also be caused by a failure in the camera's internal power supply.
Hello. Thank you for your quick response. 83F010957. The camera was purchased new in Optcorp. I'll try to find it, but probably not. I keep master frames only. I don't remember exactly, but it was probably a Maxim DL 5.24 I think the 0.01 sec exposure listed in the FITS header is probably the wrong value, because the software cannot force the camera to make the exposure time less than the minimum possible for that product, which is 0.1 sec, if I'm not mistaken. I do not use the ASCOM driver, it is not installed on Windows 7 64. Okay, it's not difficult, I don't think I've seen the ability to do bias imaging in this program, but I'll take a look. Yes, thanks, I noticed that right away too. Do you recommend checking the supply voltage waveform with an oscilloscope? Hmm, that wouldn't be good. Thank you again.
Are you using the power brick that came with it? Do you have a different power source you can try? Power down all other equipment and take a bias frame - maybe something in the vicinity is generating a lot of electrical noise.
Yes, it is original PS Yes, but with another jack, so it will take some time. Ok, thanks. It is good idea. P.S. By the way, so how do you do bias imaging using CCDOps5 (I have v5.59)? Grab - 0.1s - Dark only? Or something else? Thanks. P.P.S. My version CCODOps has Xtra menu with many functions. Perhaps I could use some of these?
Here are the single bias (dark 0.1s) and dark 30s. Both are at -10C with a different power supply. I don't think there is a pattern here, but std dev is still times higher than it was in 2016.
Yes, I agree with you. And that surprised me. And I discovered maybe just something interesting, or maybe something important. If I use CCDOps - the minimum exposure will be 0.09s, even if I manually specify less. And with that exposure the image comes out great. But if I use other programs (Maxim DL, NINA) where I can just specify that I want to shoot BIAS, without specifying the exposure time, the frame turns out uneven, with obvious darkening in the upper part. And that's how dark looks with exposure time less than 0.03s. Probably, the new versions of the driver contain a bug and incorrectly specify the minimum possible exposure for this camera. It turns out that to shoot bias-images correctly it is necessary to choose not "bias" but "dark" with exposure 0.09s
Hello! Sorry for the silence, I've been busy for a while. I downloaded the trial version of Maxim DL 6 and tried to check with this program. It turned out that when you specify to do BIAS, you end up with a good even frame, without noticeable artifacts. The user is not required to specify the exposure time manually. It turns out that at least this program works with the camera correctly, while other programs require not only to specify "make BIAS", but also to specify the exposure time. I am not sure if all STF-8300m camera owners are aware of this feature, nor do I know how other SBIG cameras behave. It seems to me that it would be correct at the driver level to make it impossible to use exposure times less than the camera specification, and if this is not possible, to warn users to pay attention to this feature of camera usage. Thank you.
By definition, a BIAS frame is a zero second dark exposure used to sample the inherent instantaneous bias charge present in the CCD sensor array.
No, it is not "zero" second, it have to be _minimal_ exposure for this camera. For STF-8300 it is 0.09s. But not every software know it.
