Embossed circles in long integration stacks? - Forums

Topic starter

Jul 7, 2024

Hi All,

Sorry if this has been posted before. My searching skills were not good enough to find a post about this. Recently for the last couple of projects I've been working on in UV/IR cut, I've been seeing these odd embossed circles only really late into the integration. The bottom left 7 o clock one is always there. The 8/9 o clock and the 2 o clock ones are fainter but usually there as well. They have been extremely annoying as they are rather faint but RangeSelection still picks up on them when I'm trying to pick up on faint objects. I probably have to scrap the projects (or post them then revisit). I can't confirm if this is always happening since some of my other projects are shorter but it seems like anything over 5 hours in UVIR has it.

Some Examples please excuse the funny gradients, I used Gradient correction before cropping sometimes.

UVIR Cut 184x180s

UVIR Cut 198x180s

UVIR Cut 65x300s last night

UVIR Cut 8x300s (not present)

UVIR Cut 208x30s (not present but there is a star halo)

I HIGHLY suspect that it's my UV/IR cut filter (which I only use because it has the same thickness as my other filters so I don't have to adjust the backspacing in my wheel). From some tests on super bright stars, the halo from it is way more defined than my other filters. However, I would like to learn about what else it could possibly be. My other filters SiiHb, SiiOiii, HaOiii, Quadband do not seem to cause this problem. I also have not changed up my imaging train, rotated it, or opened it in a while.

Equipment in the imaging train:
Skywatcher Esprit 120ED
Skywatcher 1x Flattener
ZWO EFW 5x2in
ASKAR Backfocus adjuster M54
SvBony UV/IR Cut
SvBony Sv220
Antlia Quadband
Antlia SiiHb ALP-T
Altair SiiOiii
ZWO ASI2600MC Duo

Thanks for reading and any advice is welcome!

Helpful Engaging

Eddie Pons

Jul 7, 2024

Did you retake flats? I've had this issue with mis-matched flats (that I thought would work, but if things move ever so slightly, you can get this effect.)

Concise

Quinn Groessl

IOTD Staff

Jul 7, 2024

To me it looks like something is up with your flats.

Joel Lee

Topic starter

Jul 7, 2024 · Edited: Jul 7, 2024

Eddie Pons:
Did you retake flats? I've had this issue with mis-matched flats (that I thought would work, but if things move ever so slightly, you can get this effect.)

I take flats almost everyday. Good idea, I’ll double check them shortly. It could be ASTAP doing some funny things with calibration since I have no clue how they work. I load up all the flats, it replaces them with master flats then calibrates but I don’t know if it matches day by day or nearest day to nearest day.

Joel Lee

Topic starter

Jul 7, 2024

Here are two of the more recent flats. Maybe it's something to do with when I take the flats? I take flats in the morning after the sequence is run. However, I swap around filters during the night and usually end on the HaOiii one recently. One is a sky flat and the other is a flat panel (drawing panel) + diffusion sock. Both from ASIFits viewer saved autostretch.

Quinn Groessl

IOTD Staff

Jul 7, 2024

Joel Lee:
However, I swap around filters during the night

If your filters aren't oriented the same, the circles won't be in the same place in your flats vs your lights. With a filter drawer you pretty much need to take flats before each filter change otherwise things aren't going to line up to calibrate out properly.

Well Written Helpful Insightful Concise

Rafał Szwejkowski

IOTD Staff

Jul 7, 2024 · Edited: Jul 7, 2024

This happens when a speck of dust moves between taking lights and taking darks, or if your filter wheel doesn't return the filter to the exact position every time. One way to figure out the problem is to not move the filter wheel at all during a session(necessitating using just one filter all night). OTOH cleaning the filters will temporarily solve the issue, unfortunately the specs they will come back sometimes so it's not possible to avoid them every time completely.

In this case, since the damaged areas of the stack are in an "empty" space I'd judiciously use Clone Stamp to repair them in a starless image and preserve the image. No need to scrap the projects.

PS. There's one more possiblity, that of a light leak, when dust specks are not illuminated quite the same way in lights/flats and the calibration is just a bit off, leading to the embossed look.

Helpful

DiscoDuck

IOTD Staff

Jul 7, 2024

You can also set your filter wheel in the driver settings to rotate in one direction only. Should increase the chance of it returning to the same position each time that filter is selected (i.e. "overshoot" will always be in the same direction)

Well Written Helpful Insightful Concise

Joel Lee

Topic starter

Jul 7, 2024

Quinn Groessl:
Joel Lee:
However, I swap around filters during the night

If your filters aren't oriented the same, the circles won't be in the same place in your flats vs your lights. With a filter drawer you pretty much need to take flats before each filter change otherwise things aren't going to line up to calibrate out properly.

You can also set your filter wheel in the driver settings to rotate in one direction only. Should increase the chance of it returning to the same position each time that filter is selected (i.e. "overshoot" will always be in the same direction)

I use the ZWO EFW with the one direction setting on. Does it still apply that I should take the flats before each filter switch?

Rafał Szwejkowski:
This happens when a speck of dust moves between taking lights and taking darks, or if your filter wheel doesn't return the filter to the exact position every time. One way to figure out the problem is to not move the filter wheel at all during a session(necessitating using just one filter all night). OTOH cleaning the filters will temporarily solve the issue, unfortunately the specs they will come back sometimes so it's not possible to avoid them every time completely.

In this case, since the damaged areas of the stack are in an "empty" space I'd judiciously use Clone Stamp to repair them in a starless image and preserve the image. No need to scrap the projects.

PS. There's one more possiblity, that of a light leak, when dust specks are not illuminated quite the same way in lights/flats and the calibration is just a bit off, leading to the embossed look.

Thanks for the suggestion, I'll give the clone stamp a shot on some of them. I'll also try out the one filter all night into flats. I did that last night but forgot to take the flats.

For light leaks, would that be the case in the EFW even if the other filters do not exhibit this? Would that be because there are no specks on those filters?

Quinn Groessl

IOTD Staff

Jul 7, 2024

Joel Lee:
I use the ZWO EFW with the one direction setting on. Does it still apply that I should take the flats before each filter switch?

In that case no. End of the night should work just fine. I thought I read filter drawer somewhere.

Markice Stephenson

Jul 8, 2024

Joel Lee:
Here are two of the more recent flats. Maybe it's something to do with when I take the flats? I take flats in the morning after the sequence is run. However, I swap around filters during the night and usually end on the HaOiii one recently. One is a sky flat and the other is a flat panel (drawing panel) + diffusion sock. Both from ASIFits viewer saved autostretch.

It could be the case that your panel isn't sufficiently illuminating the sensor or that there is variation due to flickering. When I use my panel at the lowest setting, there is flickering and it results in poor correction. Some dust motes are removed, but a lot is left over. Try brightening your panel and see if that may work. if you need longer times, try to add multiple shirts with the panel rather than lowering the panel light itself.

Some other things that have caused this kind of result for me include: old dark library. Fixed by taking new darks.
Taking flats in the non-linearity range of the imx492 sensor. Fixed with longer flats.
Outside stray light. Fixed with dew shield + reviewing debayered LIGHT frames to remove subs impacted by power line/powerline shadows, building reflections, direct light sources. I also don't take flats outside due to stray lights. I break down and take the flats inside (careful, to avoid dust moving around). Have not had calibration issues since.

There's a chance the uv/ir filter is problematic, but I think the reason you're seeing an issue with the UV/IR and not the narrowband filters is because the narrowband flats are typically longer and won't have the issues caused by the tracing panel or non-linearity issue (which I'm not sure is a thing on the 2600mc)

Hopefully in checking for any of these issues, you might find what's plaguing your system

Helpful

Joel Lee

Topic starter

Jul 8, 2024

Joel Lee:
Here are two of the more recent flats. Maybe it's something to do with when I take the flats? I take flats in the morning after the sequence is run. However, I swap around filters during the night and usually end on the HaOiii one recently. One is a sky flat and the other is a flat panel (drawing panel) + diffusion sock. Both from ASIFits viewer saved autostretch.

It could be the case that your panel isn't sufficiently illuminating the sensor or that there is variation due to flickering. When I use my panel at the lowest setting, there is flickering and it results in poor correction. Some dust motes are removed, but a lot is left over. Try brightening your panel and see if that may work. if you need longer times, try to add multiple shirts with the panel rather than lowering the panel light itself.

Some other things that have caused this kind of result for me include: old dark library. Fixed by taking new darks.
Taking flats in the non-linearity range of the imx492 sensor. Fixed with longer flats.
Outside stray light. Fixed with dew shield + reviewing debayered LIGHT frames to remove subs impacted by power line/powerline shadows, building reflections, direct light sources. I also don't take flats outside due to stray lights. I break down and take the flats inside (careful, to avoid dust moving around). Have not had calibration issues since.

There's a chance the uv/ir filter is problematic, but I think the reason you're seeing an issue with the UV/IR and not the narrowband filters is because the narrowband flats are typically longer and won't have the issues caused by the tracing panel or non-linearity issue (which I'm not sure is a thing on the 2600mc)

Hopefully in checking for any of these issues, you might find what's plaguing your system

That's a good suggestion Markice. I'll try slowing down the flats. From sky flats they are super short (0.06s) for this UVIR filter.

For the other two suggestions, there are numerous stray light sources in the local area but I think the Esprit prevents them from happening as the embossing comes from multiple object locations but are in the same spots. I should still try some flats inside.

For the dark library that's likely a good idea as well, my dark library was built almost 9 months ago now.

Tomorrow morning I'll try out everyone's ideas and post an update when I get the chance. Tonight it's imaging only UVIR (no filter wheel movement) and same target (ghost). Then in the morning, longer flat times and some outdoor darks to check for light leaks (it's pretty bright by 6am here).

Respectful Supportive

andrea tasselli

Frequent contributor

Jul 8, 2024

It's the FW not returning the filters in the same exact position and nothing else. Avoid that and most likely you're sorted. Just take the flats after finishing yuor exposures (length does NOT matter. contrary to what some folks here say) and before switching to a different filter.

Joel Lee

Topic starter

Jul 9, 2024

Hi everyone,

This morning, I got the flats taken for this morning after a whole night of UVIR filter images (no wheel movement since it's the first position). To help it out, I cleared out all the flats from ASTAP's flat area and loaded it up with today's flats (0.5s now instead of 0.06s). Here are the results. All images are either ASIAIR auto stretched or STF screenshotted (SPCC).

Stacked Flat 20 frames:

Dark Frame Cap on and out in the morning sun 1x300s:

Last night's stack 51x300s:

Stacked with the previous night's subframes only using today's flats 115x300s:

From these, it looks like it was an issue of older flats lying around and getting used but only for UVIR. The flat from this morning was 10x longer and did not exhibit any blobs. The dark is fully dark for 5 minutes. The stacked image from today and combining yesterday's subframes and today's had no embossed circles. As a result, it had to be a flat from one of the previous nights. I'll need to double check the other UVIR stacks using the flat from today but I suspect that it may be the case.

Lesson learned: periodically remove/update the flats used and check those for weird artifacts like that for this refractor. Looks like I'll have to go hunt down those older flats and make a note of it.

Thanks everyone for the input and advice here! I'll go stack the other images with the newer flats and try to find the guilty flat.

Helpful