No calibration frames for LRGB (bonehead mistake)

14 replies578 views
Rob Calfee avatar
Hi all,

I made a bonehead mistake because I was being squeezed by work, family, and nightfall and didn't take calibration frames before I switched cameras out for another target. 

Anyone ever do anything to compensate for the lack of calibration frames after the fact? Synthetic flats? Anything. I don't want to waste the data so I decided to make it a learning experience. 

Thanks,
Rob
Well Written Respectful Engaging
andrea tasselli avatar
Try to retake the flats as best as you can and cross the fingers.
Rob Calfee avatar
Yeah, I think after this imaging session I’ll try and put it back on and take the frames. Better than nothing I guess.
Lynn K avatar
Depending how the set up has changed, old calibration data may work. Older Bias and Darks at same exposure length and temp should be OK.  The flats may be an issue if filter dust has been rotated.  Any dust on the main optics will not come to focus.

Some software,such as MaximDL can compensate for Darks of different time lenghts.

It's a really a matter of how complex you imaging train is, such as filters ,flatteners and reducers that may be at a different rotation angle. It's  not the optical rotation that is a problem, it's the dust.

Darks and Bias can be taken latter. One just has to be able to duplicate the came camera temp, gain etc.

Just try and put the imaging train back the way it was and take flats. And hope no new dust has accumulated.  It is unlikely the old dust has fallen off.  Depending on how large the chip,
How dark the sky and fast the optics, flats may not be crucial and software flattening, gradient  techniques may be applied. 

That's my thoughts based on my experience.
Lynn K.
Helpful
Rob Calfee avatar
Lynn K:
Depending how the set up has changed, old calibration data may work. Older Bias and Darks at same exposure length and temp should be OK.  The flats may be an issue if filter dust has been rotated.  Any dust on the main optics will not come to focus.

Some software,such as MaximDL can compensate for Darks of different time lenghts.

It's a really a matter of how complex you imaging train is, such as filters ,flatteners and reducers that may be at a different rotation angle. It's  not the optical rotation that is a problem, it's the dust.

Darks and Bias can be taken latter. One just has to be able to duplicate the came camera temp, gain etc.

Just try and put the imaging train back the way it was and take flats. And hope no new dust has accumulated.  It is unlikely the old dust has fallen off.  Depending on how large the chip,
How dark the sky and fast the optics, flats may not be crucial and software flattening, gradient  techniques may be applied. 

That's my thoughts based on my experience.
Lynn K.

Thanks, Lynn.
Andy Wray avatar
Darks, bias and dark flats are no issue at all.  You can take them at any time, but preferably with the camera at the same temperature.  Flats, however, need to have the camera at the same angle and with the optics at the same focus point.  It's highly unlikely that dust will have moved on the sensor as it tends to stick where it is.  It's more about vignetting and the alignment of the optics.  So, if you can try and remember the angle that your camera was at and make sure the focus point is the same, then you can get a 90% shot at it.
Well Written Helpful Insightful Concise
andrea tasselli avatar
Or you can use an older flat (the most recent) and see how it goes. Incidentally the focus position isn't critical at all. Furthermore if you scope is a refractor is very likely that it is rotationally symmetric so it won't matter the rotation angle that much either. If you are not using a DSLR or the DSLR comes with a clip-on filter then most likely the dust stays in place if it is well attached to the front camera window. Less likely that the same will hold true for that of any filter system you might have but this is a lesser evil because is well further away from the sensor. Your best bet would a totally enclosed system with no moving parts at all, say an OSC camera sealed with a filter at the end of the 2" adaptor. Which is actually my case. I take flats every few months or so.
Andy Wray avatar
andrea tasselli:
I take flats every few months or so


Well that has taken me by surprise ... I had always assumed that vignetting (I have a Newtonian) would have altered with focus position and camera rotation.  I also have a focus tube that impedes the tube and affects the flats I believe (slightly).  I think I'll try your technique on some older images and see how that goes.  Sorry for any disinformation.
Respectful
andrea tasselli avatar
I don't think the OP is using a newtonian. Newtonian geometry is a bit a special case as the focuser might protude into the light path for very large changes of focus position but for normal variation of few mm it makes no difference. Same applies for rotation, newtonians are special cases as many layouts are not rotationally symmetric because of the position of secondary mirror in the light path. Having seen your pic of NGC6888 I'd consider yours a specimen of the latter sort.
David Nozadze avatar
Hi Rob!

If your camera does not have an amp glow, try stacking without darks and flats. Use only biases. Then use Pixinsight's Dynamic Background Extraction process. If you need to calibrate out the amp glow, then use the darks too. But I think you may get away without flats just fine. 

This image is processed only with biases and darks, as I had to remove the amp glow of my ASI294 MM Pro
https://www.astrobin.com/6irxc1/D/?nc=collection&nce=4766

This image is shot with ASI2600 MM Pro. So I only used biases
https://www.astrobin.com/2g9fs3/?nc=collection&nce=4769

No flats used in iether case. 

D
Helpful Supportive
Lynn K avatar
I would say focus position is relative to the scope, even with a refractor. My Tak FSQ106 has a very short focus travel. More likely I would be able to set the focus close. With the reducer at F3.64 , it will be more crucial.  My AP130GTX and TMB130SS uses the Starlight 3.5 " focuser. It has 100mm of travel and varies a lot depending on the train. The F ratio will also be a factor.
I agree that with a well corrected refractor, rotation is not crucial. If it is, the you may have bigger problems.

I also use the same calibration frames over  sereral sessions. I put a lot of effort in cleaning the filters from dust, so the dust donuts become less a factor.  Inow have permanet set up which makes all this much easier. However, when it did nightly set ups, I still used the same calibration frames. I know some imagers that set up nightly, but do not disassemble the imaging train, and one that built a special case to house the whole scope and imaging train.

Also the neccessity of flat calibration can vary with the object and post processing. The image in the above post has a rather dark even background and more easily flattened with software. But if you are doing nebula and attempting to bring out the faint outer nebulosity, you will need to stretch the data a great deal.  If the image has not been well flat calibrated, then you will be limited to how much the outer data can be stretched.
 
One of the pit falls of poor calibrated data is the imager will push the background black point too far, in order to cover up the noise and poor flat calibration.  Unlike galaxies, nebula is surrounded by dust and faint nebulosity. Not empty  black space. 

I realize this a bit off topic, but speaks to when flats are required. A lesson I had to learn.

Lynn K.
Helpful
Rob Calfee avatar
Thanks, everyone! Appreciate the help. I went back and put the camera and FW back on to the best alignment I could remember and then took the calibration frames. I'll post the results sometime this weekend.
Well Written Respectful
Rob Calfee avatar
Oh! And I'm going to try several strategies postulated here and compare them.
alesterre avatar
Another idea is to platesolve one of your ligth frames, that will give you rotation angle of the camera relative to the target before disassembling. You can try to replicate that angle with the camera reattached (platesolving on the same area of the sky to neutrilize inprecise polar alignment). Not sure if this can be accurate enough, but may be a good starting point. Then you can apply new flats and see if they fit perfectly. If not, adjust slightly until success (may be a bit tedious). If your mount was not moved since last session, I think you have good chances right away.
Helpful Concise
Ruediger avatar
If dust is not a big issue, you may also use an old Masterflat and gerate a “generic universal flat” by heavily blurring it and removing small structures. Best to be done in PI with MLT and only using the last wavelet layer. Works perfectly as long you do not have big and obvious dust donuts or other structures. At least you can remove gradients relatively reliably. An additional benefit: you have already a properly calibrated Masterflat you can use in WBPP.

CS
Ruediger
Helpful Concise