Final mars image from Snake Valley

[bird]

I tried lots of different tricks today to see if I could increase the signal/noise in my raw frames before processing,and finally found something that seems to work a treat.

I ran all the frames through an "optimal estimation" filter which comes as part of the netpbm package - google for it if you are interested, and not afraid of the command-line... Don't panic - Windows commandline versions are available :-)

As a result the input frames to registax were smoother and I could get more mileage out of them. This should be my "final" image (is there such a thing?) from the Snake Valley camp.

This time there's three images, Paul (aka Rumples) mentioned that the earlier ones were a bit dark, and now I know why. This time there should be something for everyone :-)

Interesting sidenote: I measured the size of mars in an image viewer, and it came to 150 pixels - corresponds to an 8x effective barlow in my scope, so image brightness would have been around f/48.

EDIT: I've added another image as well, taken at 3am when the seeing was at its best even though mars was dropping lower in the sky.

regards, Bird

[asimov]

I know I'm repeating myself here Anthony but I don't know what else to say! WOW!!! I prefer the darker one, great stuff!!:prey2:

[janoskiss]

That is just fantastic. Thank you for sharing your talents with the rest of us. I am proud for just having been there and witnessed you collecting the photons for these images. And for being able to pilot your spaceship for a few minutes.

[janoskiss]

BTW. I hope you keep your raw data. You never know when and what new techniques may come up for reprocessing. If lack storage space is a problem, I'm sure we could organise something.

[bird]

Now Steve... remember we agreed not to mention that?

Asi - the leftmost one looks best on some LCD displays, and the rightmost one looks best on most CRTs and some other LCD's.

Bird

[bird]

Storage...hmmm, with all the copies I'm making as I process it's up to about 80Gb from that night alone. By the time I chuck away the redundant bits and compress the rest I might fit it onto a handful of DVD's.

I never throw any raw data away - I still have all the raw data from when I started in 2003 :-)

Bird

[Wombat_In_Space]

All I can say is wow:eyepop::eyepop:

[[1ponders]]

Truely sensational bird :clap:

[iceman]

Excellent image Anthony, i'll have to check out this netpbm thingomajigo.

Nice image scale too, how much did you upsize?

[Robert_T]

What can one say! The detail's so fine you can see the eyelashes on the "eye of mars" in this one. Well done Bird.
Robert

[bird]

resized by 150 percent. It makes it easier for registax to align the frames. Normally I downsize again at the end of the processing but this time the data was good enough that I didn't have to.

Bird

[iceman]

What's the command line version of this called? (for windows)

Thanks

[bird]

It's part of the pnmnlfilt program (PNM non-linear filter). I ran the frames through ppmcentre first just to cut down the size and do the centering stuff, then I ran them through a script that did these steps for each frame:

"fitstopnm" (converts my FITS format to PNM internal format)
"pnmnlfilt 1.3 1" (optimal estimation filter)
"pamscale 1.5 -filter=mitchell" ( 150% resizing. and smoothing)
"pnmtofits" (rewrite the files back as FITS format)

These programs are meant to work as a pipeline, so you chain them together:

fitstopnm infile | pnmnlfilt | pamscale | pnmtofits > outfile

If you want to see a UNIX-style shellscript for this then I've put the script I used on my website here:

http://www.acquerra.com.au/astro/software/fitsnlfilt

When I run this script it excpects the alpha and radius parameters on the commandline, that's what the $1 and $2 are - the first and second parameters.

The next step is to load all the frames into registax and processed from there.

If your data is in BMP format then you'd use the "bmptoppm" and "ppmtobmp" converters in the first&last steps instead.

The netpbm package has been around forever (I remember using it in about 1990) and has been ported to just about every platform. It's a standard part of any linux installation.

Here's the website: http://netpbm.sf.net/

and here's the documentation:

http://netpbm.sourceforge.net/doc/

Bird

[davidpretorius]

Well done mate,

As Steve H mentioned I am getting a big buzz out of being part of the process. Just watching and sharing in the excitement of it all.

I reprocessed my mars again and am getting a much better handle on the processing thanks to you.

"All good things come to those that wait" not sure about "Those that stay up all night and then drive to canberra", but.....

[xstream]

Truely magnificent mate.

As others have said; Just being able to witness and learn from you your art was an honour.

[iceman]

Anthony,

When I try to run it against some bmp's, it says:

How do I run it against the bmp files that i've previously run through your ppmcentre?

Thanks

[bird]

:-) Well, I *could* say "read the docs" for netpbm... I've expanded my reply above with some more info that should help.

But you should read the docs *anyway* or you might have trouble getting anywhere :-)

The short answer is that you have to build a pipeline with these pieces:

- convert your format (BMP) to one of the PBM/PPM/PGM formats with bmptoppm
- run tools that read/write the PPM format
- convert the format back to BMP when you write the file with ppmtobmp.

Bird

[iceman]

Thanks Anthony, I could've read the docs but your explanation above is heaps better ;)

I'll see how it goes with my latest avi.

[bird]

Here's another tip - because the netpbm tools can't show you what they are doing it's hard to know what values to use for some of them.

So, I download and use the GIMP for that - it has some of the same filtering options and it's an image viewer/editor similar to photoshop etc. I load one of my frames into the GIMP, play with the filters until I'm happy with the result and then use those settings in the script to process all the frames.

In this case under the "filtering->enhance->NL filter" you find the equivalent of the pnmnlfilt program.

regards, Bird

[bird]

update: Added a second image to the original post, I think it's a small improvement over the previous one but by now I'm not sure anymore... I've spent almost the whole of the time since I got back staring at mars images :-)

Bird