2022-05-29: APP 2.0.0-beta2 has been released !
Download links per platform:
Am I doing totally wrong with my Flats? ASI 294 MC Pro
Hello all, I think thats a neverending story in many discussion boards but I could not really find an answer to my issue: Flats and Flat-Calibration... I thought that had a good workflow but with the ASI294MC I see a weird behavior when doing flats. That was much easier with my DSLM 😉
At the moment I´m imaging with a L-eNhance Filter and I have a reddish rounded square around my light frames. I read somewhere that this uneven illumination is caused by the circumstance that the cooling pad is not enirely fitting the sensor and cooling should be moderate with a deltaT of <30°C. So, I try to not overdo it with the cooling and cool with a deltaT of 20-25°C at this time. Target T is 0°C what I can hold with a cooling power <50%. Thats the Background.
So, when I do my flats I use an iPad screen with a evenly illuminated, dimmed, 18% grey screen. I take flats with EKOS and set the ADU to 30k (tried 25k and 35k ADU as well, makes no difference). Also, I tried different exposure times (with dimming the screen less or more) between <10 sec and >30 sec - no signifcant difference in the outcome. The (stretched) flats show very clear the red ring now and look resonably okay to me, so I would await a nicely corrected image after calibration. I do the calibration with 50+ of each, darks, flats and darkflats.
After the calibration the images are not even at all. I more have the feeling it´s overcalibrated. The flats show a light outer area and darker inner area. The calibrated images become lighter (and redder) in the inner part now and the outer parts darker - something I would await but it looks overcalibrated to me and I´m having a hard time in areas with large amounts of faint nebulosity to work out what´s the light pollution and what not. I read that too bright or dark flats lead to under or overcorrection, respectively. But trying with 25k and 35k ADU flats is not changing the result.
So, am I doing something completely wrong or do I oversee anything? The current imaging train is a RedCat 51, L-eNhance dual narrowband filter and the ASI294MC Pro. The camera and scope are kept clean very well, so apart from the gradients, the flats are not showing any dust spots on the sensor because it´s clean. I would be happy about some info because I have the feeling that I could get much better results out of my data than I have done so far 😉
Happy about thoughts.
Here are some screenshots of what I tried to descibe:
Uncalibrated, stretched Light Frame
Master Flat (25k ADU but the others 30k and 35k ADU don´t really differ when max stretched, there is the red gradient around a darker center)
Calibrated Stack, all set on "Auto" (max stretched, instead of an even field I get the red, bright area in the middle and darker gradients around now)
i’ll follow along. I also have issues with the same camera and flat frame calib.
I had problems taking flats with a ASI1600MM-C Pro camera and ZWO Ha filter using an iPad as light source. It turns out that the Ha filter has an IR leak that makes the flats completely useless. I saw a similar sign of leak with a ZWO R filter, though MUCH less, but not with ZWO L, R and G filters nor with a Baader OIII filter.
When I used the white ceiling of one of the rooms in my house as flat panel during daytime I was able to take proper Ha and R flats and process my images without any artifacts. Not sure if the L-eNhance Filter suffers from a similar leak though but I'd still advice to try a different light source and see if that solves the issue.
That’s a very good point, Wouter! I just took some flats against the ceiling and bang! It looks immediately much better. Testing a stack right now and will report later 👍😀
Here´s a MasterFlat (30k ADU target given in EKOS, max stretch) from this afternoon that looks much better. A first test looked much more even and I´m stacking a rather large project now that drove me nuts last week to check and compare the outcome.
This is why I like the flat panels, they should be good for this and you can really standardize taking flats with it.
In New Zealand I had to be creative and ended up with flats taken like this (with extra white cloth over the lens): 🙂
Actually, that’s what I thought too and why I used the iPad as alternative flat panel. But if it is the case that the iPad has some wavelength in the spectrum that, at least my L-eNhance filter doesn’t like i.e. is letting through, then I ask myself if “real” flat panels maybe do the same... I was thinking to give the 60x60 Floalt panel from IKEA a try. It would be large enough for all my scopes incl. the EdgeHD 8.
The flat panel should be very evenly lit and have a broad spectrum and it would be great if you can adjust the brightness of it. I use SGP because that has a very nice flat-wizard, it treats all data as 16-bit no matter what camera you use so I always know I should up the brightness a bit (or lower) to get to 30.000 ADU (50% of the range) and an exposure of a second or so. I have no experience with the IKEA panels, they are cheaper for sure so it might be worth just trying it out. I only have experience with a "astro" panel, which is probably very similar, just with a higher price. 😉
Okay, it is a while ago but instead open a new post I try to continue on this here:
I thought I managed my Flats in the meantime wuite well but that is maybe true for my 72ED refractor but now I see new problems with my EdgeHD 8... Attached I show you the masterflat (shot with ASI294MC-Pro, monitor-flat, 25k ADU (with Ekos), exposure time 11.4s, dark-flats were made too) and the integration result that was calibrated with darks and the flats (incl. darkflats). Both are maximum stretched. The resulting integration has some very well visible reddish and greenish halos what I belive comes from something I do wrong in the flat calibration. I can´t figure out if this is over- or under-correction? The masterflat is not overexposed and I thought that 25-30k ADU is a pretty good value for my camera. Also, it makes no difference if I do monitor-flats or t-shirt flats (I got a tracing tablet for my smaller scope but don´t have one in larger size for the EdgeHD yet) - the result is the same. This haloes are almost impossible to edit out either in APP (Light pollution removal) nor in PixInsights DBE. I´m a huge fan of how APP is calibrating my images so far but this time I´m a bit at the edge of my knowledge (again ;-))
Happy about every suggestion on what else to try.
Okay, I just found in a german astro-forum that I maybe have a problem that is caused by the use of some layers of paper to dim down my light source... As white paper apparently tends to let throug more blue than red and green (an effect caused by the whitening) it might be that my flats are not well enough ballanced. Meaning that if I aim at 25 ADU only my green channel reaches that level roughly but blue is too high and red is too low (if the mean ADU is somewhere in the middle). When neutralizing the image in APP then red (and green) are shifted towards blue and thus the mean ADU is too high. So, even if colours should not play a role in the flat-calibration, that´s too strong ofset is maybe the cause the halo that I see because red ist too low and blue is too high? A soultion would be to lover the target ADU in Ekos so that blue isn´t too high or to get a better colour ballance in my light source (I still hesitate to invest in one of these totally overpriced flat field boxes 😉).
I just wanted to type it might have something to do with the source or an internal reflection even. 11 seconds exposure is also quite long, shouldn't matter too much I think but I usually aim towards 1-2 seconds. Being able to adjust the brightness is nice for that. A paper might indeed function like a filter, when I didn't have any equipment when I was in NZ I used very even, white bedsheets and a table with a sheet that reflected the sunlight. 🙂 That worked out actually, but was a pretty bright source too. It's all about the need of a very even light source that is diffuse and that can be tricky.
@vincent-mod yes, I usually aim for 3-5 sec too but in this particular case I did not see any difference in the results (I do use dark flats always too as I cannot make bias frames with the ASI294MC. I at least excluded a light leak as well. Yesterday I tested it with sealed filter drawer and covered OAG in a dark room. However, yes a even light source is the next step. I ordered a A3-sized dimmable tracing tablet now to test that - many people report good results with those. Do you think, in case the channels are still separated as shown above, lowering the target ADU low enough that the blue channel is ending up at about 2/3 of the histogram would help too? Then the Red and green are even lower but when neutralizing it the whole thing should not be too bright... 🤔
Yes, the only goal for a flat is to not have clipped data basically on either the left or right. Apart from the evenness and having the same focus as the lights.