Depth of Field Processing Report

Bolts_Tweed (Mark) · #1 17-11-2009, 05:26 PM

Gday - firstly disregard the image it is just a crop of M42 that needs more dynamic range masks to reduce the core (at present they range from 30 min to 5 min and some shorter ones will come this weekend).

I have been playing with Ken Crawfords (http://www.imagingdeepsky.com/) processing techniques for while including Multi Layer Deconvolution Blending, High pass Filters and Selective Smoothing for Depth of Field Processing etc and thought that passing on some thoughts on the methods may be appropriate. I dont pass myself off as an expert - far from it - but if someone else is interested in the methods and hasnt found them yet it may be worthwhile.

The first image is a basic out of the camera shot with some curves and levels while the second has some of Kens methods applied (with some of my own modifications). The idea was to make the circular opening appear in the foreground and the hollow to receed from it. I particularly wanted the wispy grey clouds top left and the curling wave like shape at the top of the neb to look in the foreground. Unfortunately the low res 8 bit image doesnt show it as well but in the original tiffs it looks like you can stick your hand in the opening.

Bolts_Tweed (Mark) · #2 17-11-2009, 05:31 PM

Damn - wasnt ready to post that but my computer has a mind of its own.

Anyway while watching TV with my wife I noticed he is exactly right. The eye is drawn to sharper objects and thinks they are in the foreground while smoother less sharp objects appear in the background.

I am just amazed at teh simplicity of the concept and the results it produces - While the processed image isnt great - Hopefully it will get better over the coming weeks - the original image just appears well flat.

Anyway if anyone has any ideas or exetnsions on the depth of field processing - I am all ears.

Mark Bolton

multiweb (Marc) · #3 17-11-2009, 06:31 PM

That's pretty cool. It does add some depth to the picture. Thanks for the link.

Ghost_Returns (Richard) · #4 19-11-2009, 03:16 PM

The link is dead

Bolts_Tweed (Mark) · #5 19-11-2009, 03:31 PM

Thx mate try this one

www.imagingdeepsky.com

I just typed it and it worked and i'll try the hyperlink when I shut this message down. Alternatively google Ken Crawford Astrophotography. This guy knows his stuff.

Mark

troypiggo (Troy) · #6 19-11-2009, 03:32 PM

I think the issue with your post #1 above is that the closing bracket is included in the hyperlink. Just remove that.

Good stuff, by the way.

Bolts_Tweed (Mark) · #7 19-11-2009, 03:34 PM

Yeah it worked.

Look under tutorials. A bit of a pain because they are movies but it explains the process really well.

Im trying to write a PS actions file to perform a easier version of the process. Will post it uder the actions topic in this forum if I have any luck with it.

Mark

Bolts_Tweed (Mark) · #8 19-11-2009, 03:36 PM

Thx Troy

Im leaving for the observatory in a hour or so - so i'll stop bombarding the actions post for a while - unless of course its cloudy then i'll have to do something

Mark

troypiggo (Troy) · #9 19-11-2009, 04:00 PM

I'm coming out to Leyburn this Sat night. You going to give me nice weather?

You going to be there? I'll have to introduce myself.

bloodhound31 · #10 19-11-2009, 04:32 PM

Thanks for passing that on Troy. You don't think it is starting to look a bit painted? To me, M42 can be processed to death and whilst I realise there is so much going on, so bright and such a dynamic range, this can tempt the processor into giving it a solid look when we all know it is gas. (If you know what I mean.)

Other than that, wow, a lot of data there.

Baz.

troypiggo (Troy) · #11 19-11-2009, 04:42 PM

I get the impression that the above is more a "proof of concept", ie showing what the technique of selective sharpening gives the effect of. See what you mean about "painted" look, but don't think that's a result of the process discussed above, but more of whatever other processing Mark has done.

bloodhound31 · #12 19-11-2009, 04:48 PM

Quote:

Originally Posted by troypiggo

I get the impression that the above is more a "proof of concept", ie showing what the technique of selective sharpening gives the effect of. See what you mean about "painted" look, but don't think that's a result of the process discussed above, but more of whatever other processing Mark has done.

Agreed mate. Fair point too. I guess I divulged a little there from the OP point.

Sowwy...

bloodhound31 · #13 19-11-2009, 04:50 PM

Double twit me. I didn't realise Mark was the OP. I thought it was you Troy....

goofed..

troypiggo (Troy) · #14 19-11-2009, 05:04 PM

hehe, no worries.

Bolts_Tweed (Mark) · #15 19-11-2009, 05:05 PM

Yeah I agree - the painted look comes from a range of dynamic masks (about 15 to 20 from memory) trying to reduce the core - 30 min for faint brown outlying dusk down to 5 min so far. I'd have to use pixinsight or similar for a HDR processing tool if I was working up a concentrated M42 image. As I said at the start - forget the image - its rubbish - (would have been happy with it back in the film days but). This forms a small part of a mosaic of the sword to the belt and in that context it works without having an over exposed M42 white blob in the wide field. Even colour saturation is based on the mosaic for uniformity.

yeah its the depth of field concept that impresses mne. the idea of sharpness at the front and softness at the rear with the degree of softness increasing the further you go back (and vice versa) - a lot of astroimages are just compeltly 2D with even sharpening (or deconvolution) over the whole field and while that was ok I think Crawford is starting to take it to the next level - Just wish I had his ability. The HHead is another example i would luv to see him work up. The faint brown dust underneath it should be at the front, B33 next going back obscuring the HII region at the rear.

This is a bad example I agree and the low res makes it even more so but if you look at Ken Crawfords stuff you'll see what I mean. I am particularly impressed with the large blue planetary he has processed for DOF.

I'll be out there Troy so come up and say hello

M

Bolts_Tweed (Mark) · #16 19-11-2009, 05:32 PM

Just re read my last post and thought it might have sounded a bit harsh - trust me its not meant to be. Thats the only thing wrong with forums - you cant hear tones of voice

I make my living as an academic subject to peer review on work and publications and trust me some of those (trying to think of a nice word) gentlemen make suggestions re astrophotos a breeze.

What has happened but is I have been inspired to maybe have a crack at a really deep M42 image this summer at a longer focal length than the TV101 - there has been a lot of nice data published here lately and it inspires to have a real crack at old faithful the way I used to. I had better pay for Pixinsight but by 30 day trial has expired.

Im going bush to take some images - have fun guys

M