CBR to SuperBias before using it?

[Alejandro Tombolini]

Hi

I have a SuperBias made with 300 bias where it desappeared all horizonal lines exept banding.

Should I applied CanonBandingReduction to SuperBias before using it? Which is the advice?

Saludos. Alejandro.

[Alejandro Tombolini]

Maybe another question before. Is it possible to apply CBR before debayer?

[Carlos Milovic]

Hi Alejandro
If you use the SuperBias process that is in the development section, you may get rid of those horizontal lines by dropping several wavelet layers on the large scale model. The result will be much cleaner and smooth than this 300 frames integration, even with only 60 frames.

I would not use CBR before debayerization...

[Alejandro Tombolini]

Hola Carlos,

Thanks for the answer. I finilly found the module!!!. 

Teach me a little more. I have selecte as Master Bias Target Frame my actual MasterBias.fit (undebayer) and applied global.
It generates a MasterBias_superbias that look like the attached image. (Image1)
Is this the one I have to use to calibrate the .CR2 subs with ImageCalibration tool?

I debayer it to see how it looks and it appears as the second attache image (Image2), is it ok?

Saludos, Alejandro.

[Carlos Milovic]

Image1 looks fine for such zoom factor.
Actually, it has no sense to debayer bias (or dark) frames, since they are not color correlated... it is information at a pixel level, that has nothing to do with the bayer matrix.

BTW, check the convention used for the coordinates. If you are using bottom-left instead of top-left, debayerization of frames fails.

[astropixel]

This is a question about SuperBias.
Excuse me for jumping in here.

I have not tried SuperBias because it has been difficult to acquire good data since the middle of last year. However, I now have a good data set and will try the SuperBias method.

I think a section in the DSLR_RAW workflow would be helpful. Perhaps others are using the method anyway.

Is there a concise description of the process that can be adapted for inclusion in the DSLR_RAW workflow?

Carlos, in an earlier post you mentioned extending the process to darks and flats. Is that still under development?

Many thanks

ap.

[Carlos Milovic]

BTW, I sent the current code and full documentation to Juan... so it may be included in the standard releases soon.
SuperBias is performed over the master bias, creating a new master. If you see vertical patterns, then it is a good idea to use this process. In my new T2i, I can't see any vertical lines, even with a master made from 2000 frames. In this case, individual variations are much important than vertical patterns.

Vicent is working on new methods to optimizate the dark's scalation, but they are highly complex to fine tune, due the non linearity of pixels. Also we'll start some new techniques applied to flat calibration... Anyway, both fields still need a lot of work, so we don't have anything to show at the moment.

[astropixel]

Thanks Carlos. I will try it out. At this stage I will point it out as an option in the DSLR_RAW workflow.

Regards,

ap.

[Enzo De Bernardini]

BTW, I sent the current code and full documentation to Juan... so it may be included in the standard releases soon.

Cool!, I wait for 32 bits version of this module!

Thanks team!