Horrible Vignetting with ASI6200 and AT130EDT

The small reducer does of course create vignetting, but the flats should correct it so it looks like an issue with your flats.
To me it looks like they are rotated 180 degrees, but it's unlikely that's happened unless you used different software for lights and flats?

How do you take flats?
Maybe you could post a single unprocessed light and flat?

You should not do darks on this camera, it hurts more than it helps so just use bias and flats.

Hi,

Checkout this thread:

https://www.cloudynights.com/topic/652016-es-7x-focal-reducerflattener-really-bad-imaging-patternvignetting/

doesn't offer a solution but the pattern is very similar to yours and it was the reducer/flattener.

Cheers,
Dimitris

Rob Calfee:

The small reducer does of course create vignetting, but the flats should correct it so it looks like an issue with your flats.
To me it looks like they are rotated 180 degrees, but it's unlikely that's happened unless you used different software for lights and flats?

How do you take flats?
Maybe you could post a single unprocessed light and flat?

You should not do darks on this camera, it hurts more than it helps so just use bias and flats.

Thank you. Here is a flat I used. Noted about the darks. Can we post in FITs here?

I still have the scope up the rig still in the same config. Should I try taking flats again?

andrea tasselli:
Hi there,

There may be a question on how you take flats but I have taken flats with 30% light loss at the edges and still come through clean enough. Mind you, if you really lose a LOT of light at the edges you should definitely revise your imaging training, possibly ditching your current flatter as it is also a reducer and that can cause a lot of grief in flat-fielding in FF. Or ditch FF and go with smaller field and the existing reducer.  No easy way around with this sort of things. By the look of your posted flat field it seems to be the case. To advise more I'd need the original files.

Hi Andrea,

Noted on the imaging train. I'll do some experiments removing the FF/reducer. Here are the original files. Thank you so much for your help. Learning something new every hour I'd say.

@andrea tasselli

Here is the Flat: https://www.dropbox.com/s/09o6fdkq31x1kgc/Flat_80ms_Bin1_gain0_20201227-022756_-10.5C_0006.fit?dl=0.
Light: https://www.dropbox.com/s/ler5cu2xmz48vud/Light_M42_180s_Bin1_gain0_20201226-213609_-10C_0001.fit?dl=0

Appreciate the help!

Cheers,
Rob

John Hayes:
Rob,
Getting flat calibration to work properly can be a challenge and there is more to it than simply getting the math right.  First, it's important to understand why you are calibrating in the first place.  Flats correct for cos^4 radiometric irradiance fall-off, mechanical vignetting, which includes light loss due to dust motes, small optical imperfections (such as pits and bubbles) and edges that protrude into the light path, and PRNU, which is caused by the pixel to pixel variation in responsively of your sensor.  All of these factors are multiplicative to the final signal.  Flat correction will not correct for anything that is additive to the signal, which includes things like stray light.  Stray light in the form of light leaks due to poor baffling or other sources is a killer for flat data.   Since flats are correcting for multiplicative errors, the final calibrated result is computed through the process of division.  That requires proper removal of bias or offset levels in your data before the division.  It also means that as optical vignetting increases, small errors in the flat data can translate into very large swings in the output.  In general, it is best to start with a good optical system before chasing after other "down-stream" errors.  My advice when there's a problem with flat calibration is to first start with the optical system and get it tuned up to minimize vignetting.  Sometimes that requires a careful look at the optical layout and that may result in remounting the components with new, bigger spacers.  If you've got a reducer with a small diameter, that's not good--especially if it may be blocking some of the marginal rays, which would cause that component to become the system stop.  That would be pretty unusual, but it's common to introduce secondary components that are way too small to handle the full field of the sensor--and that's never good.

Anyway, my recommendation is that with the amount of vignetting you have is to FIRST fix the optics, then figure out what else might be wrong in the way that your calibration is being done.   When you have it right, your flats should be smoothly varying without obvious sharp shadows.  Sharp drop offs and distinct edges in your flat data are a bright red flag that something is wrong with the optical system.

John

John,

Thanks! Heading out to the garage now to examine the image.

Cheers,
Rob

Could you upload a bias file too?

Could you upload a bias file too?

Sorry @Xplode, I thought Andrea was asking. Here ya go: https://www.dropbox.com/s/xnf9pek9kyijqvt/Bias_1ms_Bin1_gain0_20201227-012124_-10.1C_0001.fit?dl=0

Rob Calfee:

Could you upload a bias file too?

Sorry @Xplode, I thought Andrea was asking. Here ya go: https://www.dropbox.com/s/xnf9pek9kyijqvt/Bias_1ms_Bin1_gain0_20201227-012124_-10.1C_0001.fit?dl=0

I was asking indeed but I can go without bias or dark. The flat is wrong as it overcorrects the light frame. I might have t experiment to" tune it in" in order to maximise the available frame.

Additionally, I'd like to add that the field flattener/reducer does a very poor job. There is plenty of chromatic coma and lateral colour in the image. I'd venture to suggest you use a flattener of the right size and bin 2x2 your images if you need to chase after faint target. You just not going to need those tiny pixels anyways, not in a OSC (I know, I use them).

Here is the first serious stab. Got the corrected flat in place and performed a bit of PI wizardry on it to remove gradient and minimise the vignetting (not gone but reduced). Got the colour right (I had the wrong info on the fits header, the focal length was wrong and the coordinated of the image centre wrong as well). Imaged scaled down 2x.

I seem to be trailing John Hayes in these flat-related forum posts!

He is correct, your flat is questionable (is that completely dark on the edges? That should not be happening), however the biggest issue, sadly, could be unavoidable with that reducer. Although I use an SCT, I've had the same problems and that ringing is caused by a focal reducer without high-quality optics. And the fun part? It can't be corrected well with flats... The reason is complicated and boring, but very stubborn. Essentially, the light occasionally bounces off of the sensor and other camera parts, bounces off of your concave glass element (the focal reducer), and is focused in a ring back at the sensor. Sometimes it bounces multiple times, leading to multiple rings. Flats don't correct this well, because the amount of "flux" that participates in this effect depends on how much light hits the glass and where it's distributed throughout the image! This means, using a standard flat (uniform light throughout the sensor, with high intensity) doesn't properly reproduce the effects for calibration. I spent hours on flats fixing this issue, with no relief, ever (even going as far as taking outer-space flats near the target with 5-min exposures untracked!) Nothing works.

The solution? Throw the reducer in the trash, or get very good with PI's DBE process (see a very aggressive usage of this on an image that suffered from this problem both here and here. Although I threw my entire soul at it with aggressive DBE parameters, you can still see the ring if you look closely on a bright screen. I knew it was the reducer because this didn't happen without it in the train, on the same scope.

I wanted to nab the problem once and for all, so I bought a $400 Starizona SCT corrector/reducer for the 8se so I could fully dismiss the possibility of ever solving it and just return it, however to my surprise, the results were incredibly good. There is ringing in the lights and flats, however the flats calibrated it out beautifully because internal reflections were minimized in this particular reducer/corrector (as high-quality optical elements account for it). I ended up keeping it and it's very delightful not to wrestle with that anymore when I use the 8se (I use an EdgeHD when imaging at prime focus). I've been imaging at prime focus for awhile so I don't have any final results with it here on AB, but they are incredibly promising.

You can see the same effect on my setup with the cheap antares focal reducer on my site here. I'll hopefully have a post up soon showing the difference with the high-quality glass too.