[Nebulosity] "dark file not right color type" error

[troypiggo]

Posted this on the Stark Labs mailing list, but might post here too. I've since realised that bias frames are already included in darks, so no need to comment on that. It's more about the error message I'm trying to solve:

Doing some testing last night, and I'm new to image capture using
Nebulosity. I've used it successfully to process images using CR2
RAW files captured with EOS Utility and my 40D, but haven't used it
much to do the actual image capture.

Took some 8min FITS light frames, 8min FITS dark frames at same ISO.
For the flats and bias frames I wasn't sure how to do that in
Nebulosity, so took those using EOS Utility again, but they're CR2
RAW files.

Aligned and combined the darks, flats, and bias frames using "save
stack", "none/fixed", "std dev 1.25" and saved them to "master
dark", "master flat", and "master bias" FITS files.

Now when I try to preprocess the B&W/RAW images, and select the
master darks etc and click OK, I get an error message "Error: Dark
file not of right color type".

At first I thought this may be because the lights and flats I took
were using a Ha filter, so heavily red channel although there is
some blue and green data there I think. But retried without the
flats and just used the darks, but still got the same error.

Any ideas?

[bmitchell82]

they are different types of Raw files troy, CR2 is a propritory raw from cannon, and hence the size of the image is different. Taking flats and bias in nebulosity is exactly the same as any other, put the time and iso in hit capture! i use nebulosity all the time and it works flawlessly for what i need it to do. though i do do my stacking in DSS just for convienience!

[troypiggo]

I think you misunderstood/misread my post.

I tried it with just the darks, which were taken in Neb as FITS files, so they were taken identical to the lights. Still same error message.

Anyway, I know CR2 is different to FITS, but you can align and combine CR2 files in Neb but the resulting "master" is a FITS file anyway. Pixel size would be the same too. Wouldn't that be acceptable?

(The filesize of the image will be different, but the pixel size should be the same, no?)

And the error message isn't about size mismatches, it's about "not right color type".

[bmitchell82]

mmm i don't think i misread your post, but to clarify

your taking lights and darks exactly the same (just putting the tube cap on so to speak)? they will work, theres no possible reason for it not ot work.

you mention that your taking the flats and bias though EOS utils, do you try and add them to the fit files generated from nebulosity? if so irrispective of you saving your CR2 stack as a Fit file they colour information is different.! just the same as your 16 and 32 bit TIFF files. look the same but are different.!

[troypiggo]

:lol: I know it should work, but it didn't. That's why I asked the question.

Anyway, I figured it out I think. At least just using darks and lights works. It was because I was choosing "Preprocess B&W/RAW Images" instead of "Preprocess Color Images". I had assumed FITS files were classed as B&W/RAW. So that side of things is sorted.



Yeah, guess so. Last night was only testing some other things, so not too worried about the data. Next time I'll be capturing all images with Nebulosity so different file types shouldn't be an issue.

Thanks mate.

[troypiggo]

Ok, sorted out the flats problem now too thanks to Brendan's hints above. The FITS files captured by Nebulosity were classed as color, not RAW. The master FITS flat file, because created from CR2 RAW files, was still RAW. This is what was confusing me - didn't realise FITS files can be different.

So the solution was simply to demosaic the flats master and use that. All good now. Thanks Brendan.

[bmitchell82]

happy days chief! :) now back to study i go! bloody civ engineering

[troypiggo]

Really? I'm a Structural Engineer - BE(Civil) ;)