11 replies to this topic

[Lumenarian]

Hello,

 

I'm just starting my dive into astrophotography. I've noticed on the two images that I have completed, there seems to be a blue streak running across the image. In the one below you can see one running above V559 cas near the top of the image, and another fainter one a bit lower. The two images where I have seen it were taking with different optics. One was using a cheap zoom lens, and this one was taken using a William optics GT71. So I don't think the scope is the cause, and I'm assuming it is the camera since I am using an unmodified Canon T6i. I would also assume that whatever is causing it is also contributing to the blue areas around the couple of large stars. Does anyone know what this is and how I can preferably avoid it, or at least process it away?

 

 

[psuaero]

I believe the Canon T6i has a horizontal banding issue. It's not noticeable in every image but can be quite bad in some combinations of ISO/exposure. Read up on it... not every Canon DSLR model has it but I think some of the later models have it.

 

What are you using for processing? I know PI has a utility for it (Canon Banding Reduction). I think PS also has some tools to help minimize the issue.

[Lumenarian]

I believe the Canon T6i has a horizontal banding issue. It's not noticeable in every image but can be quite bad in some combinations of ISO/exposure. Read up on it... not every Canon DSLR model has it but I think some of the later models have it.

 

What are you using for processing? I know PI has a utility for it (Canon Banding Reduction). I think PS also has some tools to help minimize the issue.

Oh wow I had no idea, but you are absolutely correct. After you mentioned this I found a ton of examples (knowing what to search for really helps), and the script did a great job of reducing the banding. I'll have to incorporate that into the workflow from now on while the image is linear. Thank you so much for pointing this out to me.

[DubbelDerp]

I get that too with my Canon DSLR, but calibrating with darks usually takes care of it. Whatever is left over is wiped away by somewhat aggressive dithering. You didn't mention calibration in your post.. what kinds of frames are you using to calibrate your lights?

[psuaero]

Oh wow I had no idea, but you are absolutely correct. After you mentioned this I found a ton of examples (knowing what to search for really helps), and the script did a great job of reducing the banding. I'll have to incorporate that into the workflow from now on while the image is linear. Thank you so much for pointing this out to me.

I wasn't aware of it until a friend using a T6i or T7i ran into the issue. I have a much older T2i and that generation of sensors don't have the issue.

 

Don't let it stop you... just figure out the combination of settings that minimize it and then use the tools to minimize it.

Edited by psuaero, 14 October 2021 - 09:28 AM.

[Lumenarian]

I get that too with my Canon DSLR, but calibrating with darks usually takes care of it. Whatever is left over is wiped away by somewhat aggressive dithering. You didn't mention calibration in your post.. what kinds of frames are you using to calibrate your lights?

I am calibrating with darks, bias, and flats. I always forget to mention that.

[DubbelDerp]

I am calibrating with darks, bias, and flats. I always forget to mention that.

That's strange that there's still that much banding left over to show up so strongly in the finished image. One thing that might be worth checking is the camera temperature embedded in the RAW file. Programs such as EXIFTool will let you see this info. There might be too much difference in temperature between your lights and darks. Although this isn't sensor temperature, it's a good reference value to see how the temperature compares between the lights and darks.

 

If you're guiding, make sure you're dithering every few subs, too. Pixel rejection with dithering does a great job of removing these artifacts.

[Lumenarian]

That's strange that there's still that much banding left over to show up so strongly in the finished image. One thing that might be worth checking is the camera temperature embedded in the RAW file. Programs such as EXIFTool will let you see this info. There might be too much difference in temperature between your lights and darks. Although this isn't sensor temperature, it's a good reference value to see how the temperature compares between the lights and darks.

 

If you're guiding, make sure you're dithering every few subs, too. Pixel rejection with dithering does a great job of removing these artifacts.

I always take darks immediately following my lights, then bias, then flats. So the darks should be relatively close in temperature to what most of the light frames are. Unfortunately, I'm not guiding. I'm using a star adventurer 2i, and with my equipment, I'm basically limited by the weight capacity. I can't really add anything else until I get a better mount. With the supply situation, it seems like that likely won't be until well into next year. For now, I'll try using the banding reduction in PixInsight and see how that impacts my images until I can eventually get rid of it through equipment upgrades. Someday I'll be able to get a good monochrome astro cam which will make a world of difference.

[jonnybravo0311]

You're getting 60" unguided subs with a GT71 strapped to the SA? Even if you did use some plate solving to help, you still managed to find and frame what is essentially an invisible target. If your camera's not Ha-modded, that's a pretty good amount of detail you were able to get from the heart. I think you might have been a bit too aggressive with the stretch (you can see the de-noise artifacts)... but still... overall, quite an impressive result with your equipment. Well done!

[Lumenarian]

You're getting 60" unguided subs with a GT71 strapped to the SA? Even if you did use some plate solving to help, you still managed to find and frame what is essentially an invisible target. If your camera's not Ha-modded, that's a pretty good amount of detail you were able to get from the heart. I think you might have been a bit too aggressive with the stretch (you can see the de-noise artifacts)... but still... overall, quite an impressive result with your equipment. Well done!

I was quite impressed as well. I can pretty routinely get 1-minute unguided subs, but I didn't think I would be able to pull this much detail out with an unmodded entry-level DSLR. As for the de-noise artifacts, is that the blue rings around a couple of the stars? Would more data to increase the SNR help with that? Overall this was quite a small integration of just a bit more than an hour.

[jonnybravo0311]

I was quite impressed as well. I can pretty routinely get 1-minute unguided subs, but I didn't think I would be able to pull this much detail out with an unmodded entry-level DSLR. As for the de-noise artifacts, is that the blue rings around a couple of the stars? Would more data to increase the SNR help with that? Overall this was quite a small integration of just a bit more than an hour.

Here's what I meant by the image being stretched too far and the de-noise artifacts showing:

 

 

See those splotchy bits throughout this snippet of your image? Those are the de-noise artifacts. You can adjust the settings in the MMT to try and minimize this effect. Also, you stretched the image pretty hard (which I don't blame you for doing since you're trying to show the faint details of the Heart), which exacerbates the issue. Collecting more photons will increase your SNR, which means you don't have to stretch it as hard... so more data is a good thing

[Lumenarian]

Gotcha. The background is definitely mottled some. I figured that was due to the limited integration time. I had a second night to stack with it, but when doing the pre-processing, the whole night basically fell outside of the acceptable tolerance. I'll probably add more data to this stack over time since my imaging train won't be changing for a couple of years.