Major issues with ASI533MC Flats

[XCalRocketMan]

Hey all you ASI533 users,

I just started to test my new ASI533MC and I have an issue with flat correction. The image below shows the final output from Astro Pixel Processor (APP) of 60 lights, 20 flats, 20 flatdarks. The center image is the master flat file produced and the associated histogram. I used a LCD Sketch pad as my light panel. The final image is the same set of lights produced without flats.

It appears that my light panel produces a significant green peak and negligible red. The flats and dark flats were produced using NINA's Flat generator and I set the ADU to about 45% of max, which turns out to be about 26,000ADU. Since the exposure time was pretty short (0.53 seconds) and I know flicker with these panels is sometimes an issue, I increased the panel brightness and added multiple layers of paper to get the exposure times to about 3.8 seconds. Did not solve the problem!

Anyone have this issue with this camera? It looks like maybe the flat is too hot and maybe I need to really scale down the brightness to get the green peak below 40-50% ADU.

Tonight I am going to retake some new flats getting the green peak under control but I thought I post this in case others have had this issue and have a solution.

[Juno16]

Hey all you ASI533 users,

I just started to test my new ASI533MC and I have an issue with flat correction. The image below shows the final output from Astro Pixel Processor (APP) of 60 lights, 20 flats, 20 flatdarks. The center image is the master flat file produced and the associated histogram. I used a LCD Sketch pad as my light panel. The final image is the same set of lights produced without flats.

It appears that my light panel produces a significant green peak and negligible red. The flats and dark flats were produced using NINA's Flat generator and I set the ADU to about 45% of max, which turns out to be about 26,000ADU. Since the exposure time was pretty short (0.53 seconds) and I know flicker with these panels is sometimes an issue, I increased the panel brightness and added multiple layers of paper to get the exposure times to about 3.8 seconds. Did not solve the problem!

Anyone have this issue with this camera? It looks like maybe the flat is too hot and maybe I need to really scale down the brightness to get the green peak below 40-50% ADU.

Tonight I am going to retake some new flats getting the green peak under control but I thought I post this in case others have had this issue and have a solution.

Wow! No issues that I am aware of. I've been using this camera for just over a year and my calibrated masterlights are as flat as a pancake.

This is a recent master flat with a uv/ir filter that has been debayered and STF in PI along with the histogram.

I use an old ipad to make my flats with the Lightbox app and N.I.N.A. Flats wizard. I use 50% adu target in the Flats Wizard and usually I aim for 2-5 second exposures.

[XCalRocketMan]

Wow! No issues that I am aware of. I've been using this camera for just over a year and my calibrated masterlights are as flat as a pancake.

This is a recent master flat with a uv/ir filter that has been debayered and STF in PI along with the histogram.





I use an old ipad to make my flats with the Lightbox app and N.I.N.A. Flats wizard. I use 50% adu target in the Flats Wizard and usually I aim for 2-5 second exposures.

Very strange. My histogram in PI looks very much like yours, except the red is the least dominant peak.

[Juno16]

Very strange. My histogram in PI looks very much like yours, except the red is the least dominant peak.

Is there anything that I can share with you to help you get to the bottom of this?

I know that are many folks here that also use the 533. Maybe they can share any experiences.

[XCalRocketMan]

Well, it looks like user error may be the root cause of this problem. A few months ago I started experimenting with my ASI2600 by not using Darks in the calibration process - just flats and darkflats - since the sensor was so clean of any noise. My success led me to do the same with the ASI533. Alas, the ASI533 seems to require darks in the calibration process. When I tested it last night, taking 30 x 30sec lights of M42, and flats of various exposures (yielding 25%, 30% and 35% of max ADU, the issue remained in all cases. But when I decided to take darks and apply them the image was perfect! Not sure exactly why the end result was so bad without the darks - seems like the flats should still have worked, but I think I'm going back to using darks on all my cameras regardless of the level of thermal noise.

[Juno16]

Well, it looks like user error may be the root cause of this problem. A few months ago I started experimenting with my ASI2600 by not using Darks in the calibration process - just flats and darkflats - since the sensor was so clean of any noise. My success led me to do the same with the ASI533. Alas, the ASI533 seems to require darks in the calibration process. When I tested it last night, taking 30 x 30sec lights of M42, and flats of various exposures (yielding 25%, 30% and 35% of max ADU, the issue remained in all cases. But when I decided to take darks and apply them the image was perfect! Not sure exactly why the end result was so bad without the darks - seems like the flats should still have worked, but I think I'm going back to using darks on all my cameras regardless of the level of thermal noise.

Great to hear that you have gotten to the bottom of the issue Mickey!

Actually, I have done several "quick processes" before and really did not have much issues using no calibration files at all. Of course, these were just quick peek processes.

I do the same. Darks, flats, and darkflats. Works well without issues.

Good news!

[UlteriorModem]

These LED light boxes have kind of narrow band output. As you noticed yours is skewed in the green spectrum.

I see this with both my homebrew light box and a sketch pad. It is evident when looking at the histogram while taking them. You see two distinct peaks as opposed to three. Ultimatly it would be ideal to have a broad spectrum LED light source but it is not really practical. It is just the nature of LED's in general. I suppose one could build a broad spectrum light box with an array of different colored LED's but that would be a real pain to figure out let alone build. By far my best results is the tried and true T shirt method, though that is not always practical either.

While typically not a problem since you should not debayer the flats or master flat, some cameras also have regions of the spectrum they are particularly blind to and this can sometimes lead to errant flat data.

Also in your original images it almost looks like the background neutralization (if any) went awry.

Yea not much help really just some food for thought.