Looking for suggestions on how to manage excessive red cast with modded DSLR

[cdavid]

Hi Harry,

I agree, I have shot images with a full spectrum modified canon T3i and a astro-modified Canon 1000D and both result in this peculiar red cast.  I think it is inherent to the modified DSLR's and the way the raw files are saved.  Although they are technically RAW .cr2 files i have read that canon has some proprietary process going on in the background.  In any event, your method has helped with many of my images but if there is a better way then I also would love to learn.

Does anyone have a specific answer as to why this red cast occurs....it seems all modified DSLR's have this problem when images are brought into pixinsight.

Carlos D.

[Carlos Milovic]

The red cast is normal. It is the result of an enhanced exposition of the reds, due the better transmition (both in efficiency and broadening of the spectral range).

If you have problems with CC due to noise, you may try to increase the exposure times. Also would help not using the structure detection, or fine tune it.

I've never had problems with CC. If LF helps, I would use it before BackgroundNeutralization, and then perform BN followed by CC as is intended.

[guest11353]

This appears to imply that LF is a good step :-) which is a relief because I can't see how that sort of red bias is "normal" ....to me it's outrageous and further..not common..there are images where it appears (M33,IC5146 etc) and others where it doesn't (M45) and so I will apply LF at the pre-processing stage - seems to make sense?
David

[pfile]

well M45 is a reflection nebula so there's not really any Ha and so there's no red cast to speak of. the strong red in a modified DSLR is usually the result of imaging an Ha-rich target.

more than likely the effects of BackgroundNeutralization and ColorCalibration will completely undo whatever effect the LinearFit had...

rob

[Alejandro Tombolini]

Hi, this video shows some examples that could help.

Saludos,
Alejandro.

[bhwolf]

Does anyone have a specific answer as to why this red cast occurs....it seems all modified DSLR's have this problem when images are brought into pixinsight.

Carlos D.

I have a modified Canon, with experience using it pre mod as well.  While images often do have a red cast, it seems about as expected with the increase sensitivity.  Is it possible the strong red component  is light pollution?  Are you using any filters? 

[cdavid]

Thanks for those videos and comments everyone!  Those videos were excellent!

I see now why I am running into this issue intermittently.....I have always just used background neutralization at its default values.....hence it removed the red cast on some but not all images....resetting the high point I see is critical.  This will essentially eliminate the need for the linear fit step.  Great to finally understand this!
Thanks
Carlos

[pfile]

ah - yeah it's important to tune the thresholds both in BN and CC in order to pick up the right pixels. i think both tools have a mode where they will output a mask of the pixels that were seen by the algorithm. you can use that to make sure the processes are actually doing something.

rob

[pdarmody]

Yes - thanks for the link to the video and all of the advice!  I have adjusted the levels in BN before, but not as scientifically as shown in the video.  I will use the readout info for the red channel on my BN previews to set the UL for the BN process.

[Harry page]

Hi

The way shown in the video is the correct way to use BN , But this still did not sort out

the problem Images I have seen 

Harry