Wit's end, need help

[Lokifish]

This will read like a mess but I really am that desperate at this point.

Everything I shoot with my 183C has an obvious magenta gradient except for subsecond exposure stuff. I've tried multiple settings in SharpCap and DSS, and hours upon hours in GIMP. I've tried having as much as 4 times the number of calibration files than I do subs. I even got to the point where I simply Googled "183C" and read every single forum post and review on this camera on the first 20 pages. Even done the same "183C" search on 10 or 20 different forums and read every post I could find. Have also done the same but with image searches. I have found not even a hint of anything like this anywhere.

So here's what I'm begging for. Please look at my data, process it and tell me if you see anything that looks like what I'm getting. If you can fix it, please tell me what steps and settings where used to do so. Please keep in mind I am literally a broke astronomer, so using pay software or replacing the camera is not really an option for me.

On another note. Altair's support forum reps have said they have no clue as to why it's there, but are sure it simply cannot be the camera. This is even with RGB color space images showing the same issue straight off the camera.

Link to my data
(776MB zip)



While not the same data as the zip, this is a prime example of the issue. I've eliminated sky gradients, OS, optics, filters, collimation, software versions, driver versions, color space, deBayer pattern, temperature, and so on as a possible cause. Have even tried every suggested and possible combination of settings in Sharpcap, DSS, Altair Capture I can find or think of.
30 subs
40 darks
20 flats
57 bias
Sharpcap, DSS, basic levels and satuaration adjustment in GIMP

[Lokifish]

I'll add this. This is typical result from the gradient map filter in GIMP "Astronomy" plugin. I have no idea how to utilize this, and documentation on the plugin is nonexistant.

[jmfloater]

There is other free software that may be easier and have better documentation. Iris comes to mind as well as Fitswork. I use PI but your budget is limited so maybe they may be a solution. Good luck!

[Lokifish]

Thanks for the suggestion but I have to be honest. I'm to the point where if I don't see proof that a piece of software or certain settings fixing it, I'm not putting in the effort. I know it sounds lazy, but I'm a couple hundred of hours into trying to fix with this since May. At one point even I went down the processing software lists on multiple sites and tried any free or trial software on it.

So while I do appreciate that you even replied, I'm at the point where I need a "here's how to fix this" for me to even consider continuing with astro/ap.


I'll also add this. Having a 183C owner chime in on this would be gold to me.

[jmfloater]

That's OK Lokifish, I understand your frustration, I just didn't know what you have tried to date. I'm not at all familiar with Gimp only what I read today on the internet. On a different forum someone was having issues with Gimp and gradients and another user replied that he switched to Iris to resolve the problem. Stay with it, somebody here should be able to help you.

[Gordon]

This was a really quick pass through StarTools.

Keeping in mind my 70 year old eyes don't work so well any more. I'm not seeing much of a gradient.

Your thoughts?

[JayTee]

Well, I gave it a go and ran your raw files (basically from scratch) through DSS 64-bit 4.2.1 using default settings and finally remembering to debayer the .fits light images so I'd wind up with a color image. I then took the DSS Autosave.tif and ran it through my normal PSCC (Photoshop CC) routine which included several stretches (both levels and curves), increase saturation, a color rebalance, noise removal + sharpening and there is a filter for PS called HLVG (hasta la vista green) and I used that filter at two different points in the process and that really got rid of the green side of the image. I think the magenta side is still there but it is hard to tell if it is gradient noise or signal.

So here's the result.

Cheers,
JT

[Lokifish]

This was a really quick pass through StarTools.

Keeping in mind my 70 year old eyes don't work so well any more. I'm not seeing much of a gradient.

Your thoughts?

The gradient is gone but looks like it was replaced with the ST mottling I see so often. At least it's good to know the gradient can be dealt with. Just need to find a free solution.

I think the magenta side is still there but it is hard to tell if it is gradient noise or signal.

That magenta you got is persistent in everything I shoot. Thanks for confirming it though.

[SkyHiker]

Yes there are huge green and purple gradients in your image. No idea why but the asymmetry suggests to me that it's more about ambient light conditions than the camera. I processed it first in DSS then in Gimp and came up with an image that I would like to show but unfortunately I don't seem to be able to insert attachments. Let me check with the mods what's wrong and add it later.

[SkyHiker]

Ah I had to click "Full editor". So this one should have an image attached.

[SkyHiker]

PS In DSS I first aligned the color curves and broadened them, moved them to the left then created a color curve and set the saturation to 15%. In Gimp I decomposed the image into RGB and used curves to remove the gradient - well not really, what I did was just removing the lowly lit parts, I wish they had a gradient option - for each layer individually then recombined them. I suppose I could probably have used the curves directly without decomposing and recomposing them, oh well.

I did not know there was an astronomy plugin. I started looking and found mostly negative comments. Do you have any positive experiences with it?

[Lokifish]

Yes there are huge green and purple gradients in your image. No idea why but the asymmetry suggests to me that it's more about ambient light conditions than the camera. I processed it first in DSS then in Gimp and came up with an image that I would like to show but unfortunately I don't seem to be able to insert attachments. Let me check with the mods what's wrong and add it later.

I eliminated the possibility of it being ambient conditions as it would require the exact same conditions were present for every point in the sky. Also, here's a stretched and saturation pushed stack of darks I did while trying to find the cause. The sad part is DSS doesn't do color dark reduction. If it did, then the gradient would be most likely be filtered out.

This is the same example where Altair said it must be a debayer issue, then a capture software issue, then a DSS issue, then a we have no clue but it can't be the camera issue.

I did not know there was an astronomy plugin. I started looking and found mostly negative comments. Do you have any positive experiences with it?

About the only thing I really use it for is the round stars filter. It helps me compensate for my mount's poor tracking/guiding.

[SkyHiker]

I eliminated the possibility of it being ambient conditions as it would require the exact same conditions were present for every point in the sky. Also, here's a stretched and saturation pushed stack of darks I did while trying to find the cause. The sad part is DSS doesn't do color dark reduction. If it did, then the gradient would be most likely be filtered out.

Darks only address dark current in the camera, not gradients due to ambient light. One way to find out if ambient light is the problem is to go to a dark site and see if that resolves it.

[Lokifish]

I eliminated the possibility of it being ambient conditions as it would require the exact same conditions were present for every point in the sky. Also, here's a stretched and saturation pushed stack of darks I did while trying to find the cause. The sad part is DSS doesn't do color dark reduction. If it did, then the gradient would be most likely be filtered out.

Darks only address dark current in the camera, not gradients due to ambient light. One way to find out if ambient light is the problem is to go to a dark site and see if that resolves it.

The problem is that gradient is not due to ambient light, the stretched dark stack proves that. If it were ambient light, it would not be a fixed gradient for every orientation and target/position in the sky. It would have also disappeared when I shot darks with the camera in a dark box to remove any chance it was an external cause.

As far as color vs mono dark subtraction, I had limited success when using a color live stack of what are normally darks as my dark subtraction in SharpCap. While it removes the gradient almost completely, it is overly aggressive when dealing with regular noise. Another issue with this method is you can only use SharpCap processed live stacks, and the dark stack has to be shot prior to the lights.