Using Maxim DL master calibration frames in PI

[dayers]

Hi all,

I recently purchased one of Adam Block's new tutorials applying his Photoshop (mostly) workflow on some M1 data.  He also sends along all of the data used in the tutorial - many hours of Ha and RGB exposures, along with master bias, dark and flat files. I can't resist watching the tutorial and then attempting to process the data in PI, but I may be cutting off more than I can chew.

The problem, as some of you may have guessed by now, is that the calibration frames were generated by Maxim DL. I have read several older threads dealing with this problem, but I'm still confused about what I need to do to convert from Maxim's 32-bit floating to 16 bit for use in the BatchPreprocessing script. At least, that is the way the problem appears at the moment.

Just to see what would happen, I loaded the three masters and a few lights into the script and ran it. It churned away for a minute or three and gave me some calibrated files. They came out black with a few stars and cannot be stretched.

If you were me (but had more smarts), where would you go from here? Do you think that the BatchFormatConversion script could be used on the master calibration files to make them acceptable to PI?

I also might do my homework on the ImageCalibration tool and use that after the format conversion, but I would really like to be able to stuff them into BatchPreprocessing. I'm lazy, and I have had such good results from it when using my own image data.

Any help, shoulder shrugs, or guffaws will be appreciated.

Dave

[Nocturnal]

I would not bother trying to re-stack masters from maxim using PS. My recommendation is to get your hands on his unprocessed stack and then try to follow along in PI. Back in the day I learned a lot from Ron Wodaski's PhotoShop processes when I 'ported' those to PI. Of course you will run into masking based steps that are hard to transfer but you will still benefit from an extra understanding of post processing I'm sure.

Actually I would recommend first taking the unprocessed stack and post-processing it yourself in PI. Then compare that against his result. I think processing classes a a lot of fun but when professional grade equipment is used to acquire the raw data the relevance for us mere mortals is limited. The uninitiated will see that no matter how hard they try their images will never look as good as those used in the class. The -real- trick is to take mediocre data from beginning astrophotographers and optimizing results during post processing. *That* takes skill.

[pfile]

set the FITS reader to "ask on out-of-range values"

when you open the fits file you should see the min and max values. see if they fall into the range 0.0-65535.0 . if so, set the low and high values to 0.0 and 65535.0. then click ok on the dialog box and the file will open. save it out as a f32 file and it should work when read in by ImageCalibration.

it's not foolproof but i have successfully used masters generated in MaximDL from Lightbuckets.com this way.

rob

[dayers]

Sander and Rob,

Gee, don't you guys ever leave the forum to go have a beer or something? Thanks for your fast response. You gave me some hope that I just may be able to pull this off.

Sander, I'm with you all the way. I sort of turn a blind eye when Adam grabs his paint brush and lasso tool and all that stuff. Been there, done that. I do not have that artistic touch. I like the PI philosophy much better.

Rob, I assume you are talking about Format Explorer > FITS > Edit Preferences dialog. Right?

Dave

[Nocturnal]

Hi Dave,

I've been horribly absent from the forums and astro photography in general. I hope to pick things up again this winter. I lost my motivation but things are coming back to me again.

[pfile]

oh i just came back from vacation and so the best thing to do seemed to be to check the PI forums!

rob

[dayers]

Very glad that you are getting back into the swing of things, Sander. I value your contributions to the conversations on this forum.

Dave

[dayers]

set the FITS reader to "ask on out-of-range values"

when you open the fits file you should see the min and max values. see if they fall into the range 0.0-65535.0 . if so, set the low and high values to 0.0 and 65535.0. then click ok on the dialog box and the file will open. save it out as a f32 file and it should work when read in by ImageCalibration.

it's not foolproof but i have successfully used masters generated in MaximDL from Lightbuckets.com this way.

rob

That did it. Thanks, Rob. The calibrated files look great, so far.

Dave

[pfile]

that's good!

this procedure is a little tedious. i think mike schuster outlined a method (in another thread) to fix such files with pixelmath. however, if you only have a few masters like this i guess it's okay to go thru the hassle one time.

rob

[dayers]

this procedure is a little tedious.

rob

Well, I have 15 more flat masters to wade through, but that won't be too bad. Too cold to image tonight, so I am glad to stay inside. I'll let you all know how it goes. I will be using BatchPreprocessing script, calibrate only. I'm like a kid in a candy store--I've never been near such quality data. FWHMs down near 3px for some of the chrominence lights) and some Ha frames even better.

My plan is to produce a synthetic L from R+G+B, touch it up with Ha, deconvolve and denoise it, then add some Ha to R to produce HaRGB to combine with the synthetic L. This is essentially what Adam does in his tutorial with his PS magic touch. I don't really expect to equal his beautiful results, but I'll give it my best shot and maybe learn something.

Thanks for your help,

Dave