A question of Green...

[Phil Leigh]

Hi there,
I'd like to know what - if anything - I'm doing wrong. Attached image shows a screen capture from Pi of 2 images; left is a stack of 36 Canon EOS 7d raw (CR2) light frames in Pi, following the Pi tutorial settings for debayering, registration and integration. On the right is a tiff file created by stacking exactly the same 36 images in DSS. No processing at all has been done to either image and no calibration files were used in either case. As you can see, the integration results are pretty similar except for the strange colour balance of the Pi stack... which seems very "green". The DSS image is colour accurate (this is not a real AP photo, it is shot indoors using a large format original from a book of Hubble photos!. White Balance was set to Daylight and no flash was used, ISO 100, F8, 1 second exposure)
Can anyone explain the reason why the Pi stack is so green and what I can do about it?
Many thanks
Phil

[ruediger]

No processing at all has been done to either image and no calibration files were used in either case.
[...]
Can anyone explain the reason why the Pi stack is so green and what I can do about it?

All looks normal to me. I guess you used ScreenTransferFunction with coupled channels. Uncouple the channels, apply "auto STF" and the colors should look more natural. After applying background neutralization and color calibration it should look normal even with coupled STF channels.

DSS on the other hand -- if I remember correctly -- does some kind of calibration during stacking, please check the setting of "background neutralization" on the "lights" tab.

Rüdiger

[Phil Leigh]

Rüdiger,
thank you. I did use STF with coupled RGB channels. I will try with uncoupled channels.
I think the major difference must be in the way the raw files are debayered...?
Attached are the two settings (DSS on the left, Pi on the right)
Do these look correct / should anything be changed?
thanks again
Phil

[ruediger]

The settings are OK, but the important DSS setting is in the dialog "Stacking Parameters". Is "Per Channel Background Calibration" or something similar activated? Try again with "No Background Calibration".

I used DSS (and IRIS) before switching to PixInsight and I've seen green (or red) backgrounds with these tools also, but especially with DSS I was never able to get good colors.

[Phil Leigh]

Yes - in my DSS the Per Channel RGB Calibration is set to "Middle"...
I can see how this would explain the difference.
Many thanks for your excellent help
regards
Phil

[tsaban]

... is a tiff file ....

Please note also that when you save in TIFF you also save a gamma value. This might have some influence too. TIFF is AFAIK not equal to FITS.

[NKV]

The settings are OK

I am disagree.

[ruediger]

The settings are OK

I am disagree.

Both settings do the same, namely calling dcraw with option "-k 0", i.e. setting the black point to zero, which is in fact "no black point correction".

[NKV]

Both settings do the same, namely calling dcraw with option "-k 0", i.e. setting the black point to zero, which is in fact "no black point correction".

Yes, you are right.

Checking the 'Set black point to 0' option is telling dcraw to use the raw data as is without subtracting the pedestal value.

[Phil Leigh]

Thanks for all the help and information...
Using auto-stretch on the Pi stack with RGB channels uncoupled does indeed get rid of the "green"...