Blown out stars

Welcome back to the hobby! I started 10 years ago but still enjoy it. I want to give some advice though.

First I would say things have changed a lot os don't do anything until you see you need it. For example, a lot of people like to add autofocus but I found its not any better than a bahtinov mask and really more of an annoyance. It all depends on what your astro hobby is like. For example, do you have an observatory or do you travel to a dark site. These kind of conditions change everything. Recently I bought a new larger RC 8" and thought I would have more difficulty like people warn. In fact I had no problems at all and it worked beautifully - I did couple it will a nice long fl (400mm) 70mm guide scope though 😀

So to debug your present problem, I would re-stack the images without any calibration to see what you get. Also review your best and worst images to see what the individual images are like. The halos you have are quite extreme. It does not look normal to me. It looks like you either have fog or dew or damage to your image train -  could be main objective or in the camera - camera could have dew or even ice. I don't think the problem is a result of stacking. So probably need to clean everything including camera - and make sure the camera sensor enclosure is desiccated - just clean sensor with an air jet blower, no products unless you know what you are doing.

Good luck!

Thanks astrogee !

I looked at the optical train during those 3 nights (with a light), and could not see any dew or ice, but as I did not monitor this all night long, I am not 100% sure. 

Camera and Refractor are new, and I have a dew heater at the tip of the refractor along with one integrated by default within the ASI2600MC Duo (it's on). When live stacking with the ASIAIR, I don't get such Halos (see picture at the end, 55 min stacking)

Next steps:

    - Restacking with no calibration frames

    - Check best and worst frames

    - Restacking without the third night as quality seems off that night. I was trying to interpret the star analysis graph in this post but to no avail, https://www.astropixelprocessor.com/community/postid/31777/

Will circle back once this is done!

P.S: I have a ZWO EAF for the focus and it works rather well so far. I used to have a bahtinov mask on a Celestron Edge HD8 in the past, pretty neat as well.

About the star analysis. Normally I try to keep all criteria above 0.8, or 0.7 if I'm desperate to make the data count. The criteria are normalized so that all values are relative to the image with the best given criteria.

I see from your other post. There is a lot of bad images unfortunately. Your session 3 is bad but session 1 is no better. Dispersion, noise and sky background are all poor on session 1 - as are those of session 3. Additionally session 3 has bad star shape throughout. A good indicator of deteriorating conditions from cloud, fog or dew is the # of stars. This may decline but should be 1.0 all the time - looks ok in your sessions.

I've attached my results for my last outing, you can see the sky background and dispersion were not so good at teh start. I think there was 100% humidity that slowly cleared. However, most of the other criteria were pretty good. Only one was below 0.8 which I may have removed from the stack.

Thank you so much for the insight, this is very useful!

I am wondering why the noise across all my sessions is around 0.5. I thought the noise of the ASI2600MC was rather low, I’ll investigate this one.

On top of the ToDoList above, I’ll run a star analysis for individual sessions to see how it looks over the weekend.

Posted by: @gigaparsec

↑

I am wondering why the noise across all my sessions is around 0.5. I thought the noise of the ASI2600MC was rather low, I’ll investigate this one.

Yeah, you can see that at the beginning of session 2, you have the image with best noise so its value for noise is 1.0 - I would inspect this image closely to see if you can tell what's going on. Then things quickly deteriorate to 0.5 - visually compare this image to the first one. It's mysterious 😮 I would try taking 10-20 images without cooling and then start cooling and see what happens. All things being equal, it's supposed to get better - at least it should not get worse.

Oh, and it just occurred to me that the bad sky background and dispersion are probably due to the moon. In my example, I think the moon was setting - so moon and high humidity was a double shot against good imaging!

Update Following My Investigations

I’ve independently stacked the three sessions and conducted a detailed analysis, including star registration and normalization. Surprisingly, the analytical results were entirely different for each session (i.e. session 1 and 3 have the noise and SNR above 0.8-0.9)

To dig deeper, I examined every single frame individually to identify potential issues. Here are my findings:

• Session 1: Around 45% of the frames show a significant halo around Alnitak.
• Session 2: All frames (100%) exhibit a halo around Alnitak.
• Session 3: Approximately one-third of the frames display the same halo issue.

When I excluded the problematic frames, using the remaining 55% from Session 1 and two-thirds from Session 3, the results improved drastically. The difference was like night and day, which was reassuring!

Interestingly, the appearance of the halo seems random. It can occur at any point during the session, typically lasting for 4-6 consecutive frames before disappearing. I also expected the quality score to drop when a frame had a prominent halo, but that wasn’t consistently the case.

Based on this, I suspect two potential root causes:

1. Atmospheric light dispersion: Possibly due to thin clouds, only visible around very bright stars like Alnitak.
2. Dew or ice or humidity: However, this explanation seems unlikely since the halo appears, disappears, and reappears sporadically. I have currently selected 20% power in the AsiAir for the dew resistance, and will crank this up to 40% just in case. 

Additionally, I noticed an odd behavior during the final frames of Session 2. As the sun began rising, the background brightness and dispersion decreased significantly, yet SNR, noise, and quality metrics spiked. I’m unsure why this happens, though I recall reading something in APP (Astro Pixel Processor) about this behavior when hovering over one of its options. It seems expected in some situations.

Next steps: I plan to stack Sessions 1 and 2 and will share the results here once done.

Calibration Frames: I’m encountering the warning, “We have detected a possible sensor offset issue between the MasterBias and the MasterDark.” I found a post by Mabula mentioning that this is normal with newer CMOS cameras. I’m currently using Darks, Flats, and Bias frames. Would switching to Dark-Flats be better?

Sebastian

A couple of suggetions.

Firstly, from what you have said I would think that thin cloud is the most likely culprit. These clouds can be difficult to see. I now use an AllSkyCam to continually monitor the sky when I shoot. I have it set to take a 10s exposure every 2mins, and I can monitor it with any web browser in the house. Mine uses an ASI662MC and one of those cheap ZWO cs lenses, athough I think you can use just about any old ASI camera you might have lying around. The point is that I find it can easily see thin cloud that my aging eyeballs might not see. By the way - the AllSkyCam software can do a lot more than the simple real time monitoring that I use it for. Runs on a Raspberry Pi.

Secondly. When I first got my 2600MC I was still using bias frames, and this did not work well - I got overcorrecting flats. I gather the sensor in this camera is not linear at short exposure times. Others have reported the same thing, although mysteriously it doesn't seem to affect everyone. Anyway - I switched to using dark flats (aiming for around 2s for both flats and dark flats) and the issue went away. I would certainly recommend dark flats for this camera, and now I use dark flats for all my cameras.

JC

Posted by: @gigaparsec

↑

Calibration Frames: I’m encountering the warning, “We have detected a possible sensor offset issue between the MasterBias and the MasterDark.” I found a post by Mabula mentioning that this is normal with newer CMOS cameras. I’m currently using Darks, Flats, and Bias frames. Would switching to Dark-Flats be better?

The camera you have shouldn't have much noise or artefacts such as amp glow.  You probably don't need dark frames at all.  Dark frames do add noise but if you have things such as amp glow then the noise increase is worth it to remove the issues.

See what happens when you don't use dark frames.  Does it make a visible difference?  It should make processing quicker without them.

Posted by: @gigaparsec

↑

Calibration Frames: I’m encountering the warning, “We have detected a possible sensor offset issue between the MasterBias and the MasterDark.” I found a post by Mabula mentioning that this is normal with newer CMOS cameras. I’m currently using Darks, Flats, and Bias frames. Would switching to Dark-Flats be better?

Yes I use dark flats now and they are easy to acquire right after doing regular (light) flats if you do it at the end of your session. 

The offset warning is interesting because you should establish an offset for your camera. The ZWO drivers will use different offset for different gain but I find its best to take all images, whether lights or calibration with the same offset, no matter what the gain. So find the worst case offset and use that. The worst case offset will be needed to offset the worst case noise in the image to rise above the 0 value from the camera. The point being that you are trying to calibrate out noise but this cannot be done properly if your noise is blocking at the bottom of the histogram. 

Thank you so much for all the valuable feedback. It’s been incredibly helpful!

Here’s a recap of the adjustments made:

• Removed all frames with halos
• Replaced bias frames with dark flats

Remarks:

• I tried processing without the darks, but APP gave a warning, so I included them anyway
• With the ASIAIR Plus, the offset is fixed at 50 ADU and can’t be adjusted
• There are fewer frames now (34 x 300s and 132 x 120s), resulting in:
  • More noise and less detail,
  • But much better control over the very bright stars. I can now see the Flame Nebula and IC 434 behind Alnitak more clearly
  • I will need to work better on the remaining halos as I've probably been too aggressive with the cursors

Overall, I’m pleased with the results, though I’ll need more hours of data to regain the detail I initially achieved. Considering this is my first attempt at astrophotography and post-processing after a 10-year break—and not the easiest target—I’ve learned a lot, so it’s been worthwhile.

Next steps:
I have 70 x 300s frames of Ha-OIII data, so my plan is to work on an HaRGB integration next.

Here are the results:

Hi Sebastien @gigaparsec,

@astrogee, @connor231 and @wongataa, thank you very much for assisting Sebastien with his data issues. I have not much more to add from reading the whole topic. I only have a comment about using darks.

Darks (or bias) are needed to subtract the sensor offset from the lights to be able to have the flats work as expected. That is the reason of the warning when you don't use darks. Just make sure that you supply enough darks (at least 40) then the added noise from darks will be negible and you don't have the risk of flats not working optimally 😉

Mabula

@mabula-admin thanks a lot for the clarification. 

Here is my HaRGB version. It’s a bit oversaturated and I am still learning the whys and wherefores but Alnitak is more under control and I really like being able to play with the 4 channels.

Hi Sebastien @gigaparsec,

Excellent, that is looking nicely already ! The colors are nice in the stars and the nebula as well 😊 

Compared to the image posted in the first post if this topic, the improvement is big, great work!

Mabula