Some mono CCD data for you to to play with.

[MikeD]

Hi all.

Thought it might be interesting to see what different results people come up with, processing the same data.

Also to give people thinking of getting a CCD or one shot colour and DSLR imagers a go at some mono data.

So I have uploaded a folder with five 16bit fits files of M51 I took last year

as a zipped file( 18.6 MB):

http://homepage.ntlworld.com/lesley.deegan/m51%20raw%20cal.zip

The images are calibrated ,stacked and aligned, that is it.

The luminance is from 89 x 5 minute subs

the Red, green and Ha are from 9 x 5 minute subs

And the Blue is from 10 x 5 Minute subs.

There are some gradient's in some of the subs from Street lights around my house to deal with, mostly in the L and R chanels.

the image below is what I made of the data tonight.

Have fun.

Mike.

[SteveL]

oh you evil man.... thats going to take a few hours of my day

[Psychobilly]

Nice one Mike

Peter...

[FLO]

Good idea Mike

Very nice image!

[Aza]

Nice one Mike!

I am hoping to get a CCD, so will definitely have a crack at this

[MikeD]

Looking forward to the results guys.

Thought I should list the programs and processes I used in my edit.

I used Photoshop,Pixinsight and noiseware pro to process the data.

within photoshop I used Levels and curves to stretch all the data and to adjust the colour balance, masked layers to sharpen M51 without affecting the background.

the Ha data was added to the red channel ( after the RGB was processed) as a masked layer masking the ha data in the background to preserve the star colours.

after reconstructing the rgb with the new Ha/R channel the colour was readjusted using levels and curves.

then the ruffly processed L data was converted to RGB mode and added as a luminance layer and further processed with levels and curves while I could see the affect this processing had on the colour data, constantly messing with the RGB colour balance with levels and curves.

Noiseware pro was used on the background of the L data through a masked layer to preserve the galaxy detail and also on the rgb data to smooth out all of the rgb data before it was used to colour the L data.

Pixinsight's Dynamic background extraction tool was used on the RGB and L data to help get rid of a gradient in the data.

Hope that makes sense.

Mike.

[roundycat]

Mike, I hope you will not consider my comments out of place or take anything the wrong way but I was stuck for words (except for two, a phrase ending in .... me!) when I looked at the raw data. As for the result you achieved I was absolutely gobsmacked and that is no exaggeration.

I was initially puzzled by the background to all the images, it seemed to me to be blurry as if a smoothing filter had been applied. Had it? A quick comparison with some of my own recent M51 showed over twice the background noise but with a higher overall signal. Weird. The FITS header told a strange story about Bias subtraction, then Dark subtraction, then Dark-Bias then Flat. What was that all about? What did you use to calibrate and stack?

I did have a quick go at producing a finished picture but in truth it was such a pig's breakfast I would not even show it to my cats.

Truly, if I wore a hat I would take it off to you for your end result. I don't think I have enough life left now to get on top of processing the way you have done.

Dennis

[MartinB]

Mike, stupidly I had a quick look. I deconvolved the luminence with CCD sharp. sorted out the resulting star halos using a star layer from the undeconvolved image. Did an extra bit of sharpening with a high pass filter and had a very nice luminence. So far so good. Then did a quick and cheerful RGB and lobbed in the ha. Result - yuk!

So I reckon Steve was spot on saying it would take 5 hours of his day. Getting the colour data tweaked and getting the right blend for the Ha and then it still probably wont match your effort. Still I've saved the luminence and might just come back to the colour. On a rainy day!

It is a great image

[MikeD]

I was initially puzzled by the background to all the images, it seemed to me to be blurry as if a smoothing filter had been applied.

Dennis

Hi Dennis.

The data uploaded has had no smoothing applied, and was calibrated using maxim DL, 10 darks, 10 bias, and 10 flats with the following checked in the set calibration window: Calibrate bias, calibrate dark and calibrate flat and Bias subtract flats, thats what I do with all my data ,Is that not the right way to do it?.

It's intresting to hear my data seems miles off what is considerd normal

Thanks dennis for pointing this out if there is somthing I am doing wrong I am keen to find out what it is and put it right.

Mike.

[roundycat]

Mike, that makes it a bit clearer. I never calibrate my flats at the same time as the main image. I usually have all my night's image files in one folder by date and then associate a sub folder with it containing just the cal files I intend to use. All done in Win XP. I then set Maxim to calibrate dark, calibrate flat (using the relevant cal folder) and away I go. What threw me was the way the FITS header seemed to report what you were doing. It sounded like the bias was being taken out twice; put that down to my ignorance.

I certainly wouldn't say your data is miles off normal, only miles different from mine! I used to give advanced darkroom lessons for monochrome printing and would offer to print a student's negative. What a mistake. You simply cannot believe how comfortable you become with your own negatives. Everyone else's seem unprintable and when you try to print one you end up looking like an idiot.

I look forward to what others will make of your files, I'm sure the info is there, it's just a question of making it fit.

Dennis

[mostschaedel]

Hi Mike!

Thanks for your great raw images.

Tried the processing with Astroart and Photoshop.

Astroart was used to tune the histogramm of each channel, to

remove the gradient and to improve the star colors.

Photoshop was used to combine the L and HA with the RGB,

Tried to get a few of the many small galaxies and faint stars visible.

Its still not the optimum will try for a better version. Will take time.

[James]

I've been meaning to give your data a try for some time Mike but didn't feel I could do it justice. I'm not sure if I have after looking at your results and mostschaedel's above but all the same here's my effort... I ended up combining the Ha into the Red channel which in hindsight I think was a mistake - should have been in the luminance like you've done.

Processed in PixInsight, cropped, background neutralized, DBE'd (twice), colour calibrated, HDR Wavelets, curves and a little noise reduction.

James

[ollypenrice]

I'm running for cover...

Olly