Cannot figure out these weird gradients

That is puzzling. could it be IFN? That is a fast scope…what Bortle? I searched AB for wide shots and didn’t see any with IFN, but every image has dark black point. -john

📷 Autosave.jpgHere is a reference image of stacked image without flats/darks/biases

Tony Gondola · May 7, 2026, 03:30 AM

Not sure why you’re doing 300 sec. subs with a 585 but that’s another thread. Since it won’t calibrate out I’d vote for dew or frosting on the sensor. Have you blinked your light frames to see if the effect comes in over time or if it’s static for the full session? Did you run the camera’s dew heater?

No its not the dew.
i have had dew before and the contrast is huge. and the dew spot shows up on the sensor itself.
This shadow only reveals itself after stacking whereas dew shows up on each individual frame

Looking forward to see more answers here.
On several occalsions, i’ve been struggling with circular artifacts i didnt understand, too, mostly with camera lenses. These sometimes included obvious iris blade artifacts, or even precisely separating left from right half of the image sensor. These artifacts seem to be reproducible, but i never could remove them by calibration or understand their cause.
One thing to also consider could be light leaking through the back of the reflector, but this one here is a refractor. I’ve been experiencing light leakage through the focuser, too.

I think, taking flat frames can not precisely reproduce the illumination conditions, especially with respect to angular distribution. I also couldnt get rid off my artifacts by taking sky flats instead of using a flat field box. Reducing inner reflections would then be one thing that could improve it.

Taking the dark / flatdark frames with identical exposures as the light / flat frames, and calibrating without bias frames?

Agreed - calibration issue most likely.

The reason you may not want to run 300s lum subs on a 585 is that the sensor has a very low well depth (electrons required to fully saturate a pixel), so 300s luminance subs are going to have completely blown out, bloated stars… With a 585m, lum filter and an f/4~f/5 scope I’d be running 60s lum subs at most… My IMX571 has significantly greater well depth than the 585, and with my 6” f/2.8 newt I still tend to limit lum subs to 60s to avoid completely burning out star cores.

I’m not sure if this is the issue but I remembered that Adam Block had a few videos on YT which covered something similar: https://www.youtube.com/watch?v=Kkg441UBNpo

In addition, something similar happened to me when my sensor iced up during a winter night and I had not turned on the internal heater. Don’t know if your camera has an internal heater though.

I believe you have two problems going on here. The first of all is definitely bad calibration. I am in agreement with Andrea that this could potentially be a light leak. You can test this by running fast exposures on a dark night and running a little flashlight around the imaging train. You can watch either the ADU count, or visually with a sharp stretch you will see where the light is getting in. Second, you can check for reflections. Something like this will throw off everything. I would check your dark frames, if they were taken while on the scope the light leak will be obvious.

The second issue I think you have clouds moving through your exposures. As they move through each subframe, the average of each pixel will bounce around and rejection will not handle it well. This will cause these fake light structures that look like IFN.

Start by checking your master dark. If confirmed, you can try restacking without darks. While some 585 models need darks, it’s not that bad without, by design there should be no amp glow to calibrate out. From there, chase your light leak. I would count this data as a loss, but you might be able to save some of it provided you can remove the cloudy subs and average it out with a few more nights exposures.

Alex Nicholas · May 7, 2026 at 06:50 AM

Agreed - calibration issue most likely.

The reason you may not want to run 300s lum subs on a 585 is that the sensor has a very low well depth (electrons required to fully saturate a pixel), so 300s luminance subs are going to have completely blown out, bloated stars… With a 585m, lum filter and an f/4~f/5 scope I’d be running 60s lum subs at most… My IMX571 has significantly greater well depth than the 585, and with my 6” f/2.8 newt I still tend to limit lum subs to 60s to avoid completely burning out star cores.

The 585 has a deeper well depth than the 571 when corrected for the difference in pixel-size.

Besides the possibility of a light leak, some considerations around the calibration. There seem multiple effects going on, potentially pointing to something with flat calibration as well as with dark subtraction. Perhaps it is worth checking the following options:

• Are the darks ‘fresh’? If they are older than a few months, perhaps shoot a new set of darks. Sensors will change over time.

• Are flats collected with the same Conversion Gain mode as the lights/darks? You can change the gain for flats without much penalty, but you don’t want to take flats in LCG mode and lights in HCG mode or vice versa.

• You may want to try to add a pedestal to your light images when calibrating them. Try a fixed value of 50 or 100 or so.

• Are all lights shot with the exact same settings (gain, offset, CG mode)?

Good luck trouble shooting and let us know what you find.

CS, Willem Jan.

+1 for light leaks… hard tiger to mate down…

Ciao

Lorenzo

Ok Everyone,
I think i have the solution.

For the past 7-9 nights, i have tried and tested some ideas and the ones that worked are

1- Adding an adapter ¾ Inch on top of the coma corrector, and flocking it seems to have done the job for the sky glow directly attacking the front element of the CC.

2- I have also found out that the stock focuser cannot handle the weight of the camera, cc and the filter wheel. So I have removed the FW for now, and have placed an order for the CYCK focuser.

Hopefully that will resolve all the issues.