📷 image.png
got this weird artifacts on integrated images come form oiii and sii not ha, just using wbpp, yes cosmetic correction, flats are ok, no drizziling, yes dithering every 3 subs, calibrating with bias+flats only, evertying the same, tmep, gain, offest and stuff. Any clue? Thank you for any help.
Weird artifacts in OIII and SII integrated images using WBPP Touptek ATR585M
uhmm single oiii frame… 📷 image.png
I saw that pattern near the end of processing an image from my Touptek 585m yesterday. A final treatment of noise reduction eliminated it completely. Sorry but since I was able to fix it I didn’t go back through the processing stack to see what was causing it. It has to be one of the steps that works the image in blocks, of course, many do. If I get a chance I’ll try and narrow it down.
Helpful Respectful Concise Engaging Supportive
Tony Gondola · Feb 17, 2026, 11:06 PM
I saw that pattern near the end of processing an image from my Touptek 585m yesterday. A final treatment of noise reduction eliminated it completely. Sorry but since I was able to fix it I didn’t go back through the processing stack to see what was causing it. It has to be one of the steps that works the image in blocks, of course, many do. If I get a chance I’ll try and narrow it down.
thank you i’ll try to process it but it’s unexpected
I have only seen artefacts like that when either:
Offset is too low and the signal floor is too low (ie - needs more sub-exposure time)
- You can check this by looking at the ADU count statistics for the raw, unprocessed subs.
Dithering AND Drizzle is used without ample integration time for it to correctly weight the levels for the background.
- You’re not drizzling, so this should not be the issue, but I would double check that drizzle is definitely off.
- The shape of the artefact will change depending on the mode you’re using (square, varshape etc.)
Well Written Helpful Concise Engaging Supportive
Alex Nicholas · Feb 18, 2026, 12:32 AM
Offset is too low and the signal floor is too low (ie - needs more sub-exposure time)
- You can check this by looking at the ADU count statistics for the raw, unprocessed subs.
I agree.
I guess the screenshots are STF stretches? Try to process those images as normal to see if the pattern pops up in the nonlinear state. Because of the very low noise floor of the IMX585, the STF stretch sometimes really exaggerates the stretch. The pattern should vanish if you stretch correctly and then apply noise reduction.
@Frank "Voloire" are you using the HDR mode?
Helpful
Look at the bias and see if you see the same pattern.
This is a well known problem. The Moire effect comes from offsets that are too low so that some of the calibrated data at the low end gets truncated at zero. Those zeros create the Moire pattern when the data is aligned. As some have pointed out, this can be caused by bias drift and because NB signals are typically fairly weak, that’s where the problem is most commonly encountered. It is easily fixed by simply adding a fixed offset during calibration—before image registration. It’s called an “Output Pedestal” in the PI ImageCalibration tool. A value between 30 and 70 ADU typically works well to resolve the problem.
John
Well Written Helpful Respectful Concise Engaging Supportive