RAW files - too bright?

[twobleak]

Hi All

I'm still rather new to imaging so I wanted to get some advice, if possible...Attached is a RAW file from my DSLR shot at ISO 640, 25 seconds and I think at f4. Is this just too bright to begin with? I'm trying to find a good starting point...No moon but under Bortle 5 skies. Thanks

L_0005_ISO640_25s__NA.CR3

[powerlord]

looks fine to me - histrogram should be off the left, and about 1/3rd up ideally. looks about spot on.

stretched, i can even see a bit of nebulosity there around Sadr is it ?

[twobleak]

Excellent, thank you. Yes, it is indeed the region around Sadr. I stacked 100 of these, 20 darks and 20 bias and like you say, I can see some faint nebulosity in there but my final image is really not all that. It wasn't even an hour of data so perhaps that isn't long enough or my post-processing is at the bottom of a very steep curve!

[vlaiv]

2 hours ago, twobleak said:

Excellent, thank you. Yes, it is indeed the region around Sadr. I stacked 100 of these, 20 darks and 20 bias and like you say, I can see some faint nebulosity in there but my final image is really not all that. It wasn't even an hour of data so perhaps that isn't long enough or my post-processing is at the bottom of a very steep curve!

If you can't see anything in stack, that is ok. Stacks have very high dynamic range and you need to stretch your data quite a bit to really show what lies in the murk.

[powerlord]

well, feel free to post the stacked fits here and myself or others can have a look if it helps. An hour of data should certainly get some decent data. I started with DSLRs though mine were both modified. But here's an early one from an hour on rosette for an example.

[twobleak]

I've attached the FITS file from DeepSkyStacker, if anybody wanted to take a look...I had several tries at stretching the image myself but it did not turn out very well.

sadr.FTS

[vlaiv]

This one was a bit challenge to get working.

Image was saved as 32bit integer in float point format - which means that most of the image was negative numbers in range of ~2,000,000,000.

Gimp could not recognize it and it displayed completely grey image. I wonder if other software had similar issues and that is why you thought that image is blank / dark.

In any case - loaded it up in ImageJ - I scaled image to 0-1 range manually. Did background removal as gradient is very strong. There is vignetting in the image - flats should help with that, but I guess you did not take any.

I binned image x4 to bring up SNR and help expose faint signal. Above is the result after composing and stretching in Gimp. I did not do any sharpening or noise reduction.

Data is good (apart from missing calibration frames) and focus is excellent - stars are tiny.

[twobleak]

Wow that was quick! Thank you. I wasn't sure if I should save that as 16bit or 32bit - only two options I can see in DSS when saving as a FITS file. Would it make much difference if I posted the 16bit output from DSS?

It wasn't that the image was dark, more that my attempts at stretching in PS were all looking rather ugly. I couldn't get anything close to what you posted above. 

I'm glad there is something in the data though and that I'm not doing too much wrong on that side of things.

edit: no flat frames - not got to grips with those yet.

Edited November 29, 2021 by twobleak

[powerlord]

I don't think that's stacked right then, as I stetched the single sub just posted before and got more out of that.

- just around butterly nebula below - this is from one sub, not the stacked fits.

[vlaiv]

5 minutes ago, twobleak said:

 I wasn't sure if I should save that as 16bit or 32bit - only two options I can see in DSS when saving as a FITS file. Would it make much difference if I posted the 16bit output from DSS?

Always use 32bit precision - but use 32bit float precision.

I think that there are two options for 32bit - 32bit int and 32bit float. Second one should be used.

 

[twobleak]

8 minutes ago, powerlord said:

I don't think that's stacked right then, as I stetched the single sub just posted before and got more out of that.

 

 

What could I have done wrong? I did not change much in DSS, just went with the recommended settings.

[vlaiv]

5 minutes ago, powerlord said:

I don't think that's stacked right then, as I stetched the single sub just posted before and got more out of that.

This is red channel only stretched like crazy (vignetting really clips more than just shows in corners):

Here is reference image of that region rotated and scaled to roughly show the same FOV/scale:

I see a lot of captured detail, and keep in mind that this is camera is not modded for Ha part of spectrum.

[vlaiv]

3 minutes ago, twobleak said:

What could I have done wrong? I did not change much in DSS, just went with the recommended settings.

If you worry about stacking in DSS - just stack with simple average method (nothing fancy) - that maybe won't produce best results, but will always work (you can't mess up parameters there):

So no background calibration and average stacking mode.