i managed to finish off the 6:06 am raw mode 90 sec 450 frame video.
Split into r/g/b avis via aviraw,
virtualdub into bmps
ppmcentre to centre the images and to crop to 400x400, but no quality estimate necessary due to great seeing
registax with each colour
turn auto processing off
align, and then using framelist, click on the newly positioned #1 frame and align again.
limited each to around a quality of 88%.
create a reference frame of aprox 10% of the # of frames left after limiting
move slider3 to 5.8 and then press continue. (this 2 lots of alignment, create reference and slight wavelets creates an reference image that should be slightly better than your best individual frame. the program then goes through each frame and will try and alter it to match this "ideal" image.
optimise (not optimise and stack) with resample set to mitchell if wanting to upsize it a bit
goto stack tab
click stach graph and move line down to limit any frames that could not be optimised well
stack with histostretch not checked
goto wavelets and move slider 6 to the right, then 5, then 4, then 3 ro your liking. I have three saved setting soft medium and hard
save as a tiff file
in astral image
open up each of the 3 tiffs r/g/b
convert each to greyscale and then rename each to r,g,b
do a quick rgb combine and align the colours if necessary, zoom in for fine control to use as a reference and rename to combine1
i have used a LR deconvolution of 3 x 1.3 on each of the three frames. As my image scale is around 1000x, the little preview window is misleading to how each frame will turn out, so trial and error here, sometime ini ME, i have used 10x2. There is no set rules here. Jupiter has been a ME planet for ages for me, but with these images only a light 3x1.3 LR deconvolution sharpened up some of the cloud formations
combine and align if necessary (settings from before should still be there)
unsharp mask: i have showed all three with a value of 2.0
median 3x3 then softens slightly the unsharp mask.
if using evaluation version of astral image, then do a screen dump (shift-Prnt Scrn)
into photoshop for pasting the screen dump and arrange image(s) as necessary
My temperatures of my mirror and ambient were within .2 degrees of each other. There was a hole in the jetstream over the state and i used the moons of jupiter to focus. I did star tests at around 500x (12.5mm in the 5x powermate) for around 1/2 hr at 11.30pm before i went to bed. I did a quick check star test before i started imaging at 5:15am.
Conclusions: When working with raw mode, it is preferable to leave the histo stretch option unchecked whilst stacking. Astral image will do a histo stretch automatically, but it seems it is less severe and the slight burn out hass all but disappeared.
Feel free to play with any of these images and i have a 80mbyte zipped up file with the 450 each of r/g/b bmps, many tiffs from different resamples and histo-stretch on/off. I can burn to cd and send to you, or if 80 mbytes is ok for you to download, i will put on my webserver later today
I am sure i have forgotten something, but will add it as we go along.
The detail in those shots is superb and the colours very true and natural.
You've also done a great job with the processing. Not too harsh, which is very tempting with wavelets but you didn't need to push so hard with the detail already there in the raw frames.
Well done again, and thanks for including the thorough detail of your entire process.
Simply stunning – what a beautiful sequence! You are cranking out some amazing images of late. Thanks also for taking the time to document your processing procedures – a very comprehensive write up indeed.
yeh matt, i basically chose a wavelet seeting that was starting to show clearer detail and then once at the deconvolution stage chose a white cloud formation in the brown bands and then made sure it was getting sharper with each tweak.
well i am ticking off all the paramaters ie cooling collimation, processing, camera modes ie raw and 900nc. I hope to be fully armed going into mays opposition of jupiter. There will be moon transits, great red spots etc along the way that given the way the seeing is so fickle, you want to make sure you take advantage of those opportunities as they come
spec-bluddy-tacular DP! the festooning and white swirls in the main bands are amazingly detailed... I feel slack and slip-shod in my approach too after reading your exhaustive checklist
I feel slack and slip-shod in my approach too after reading your exhaustive checklist
thanks rob,
but as asi and i were spamming on messanger last night, anyones approach to set up and say processing will always sound complicated to someone that has not done it that way before. eg raw mode for the toucam sounds hard, but it is easy to swap over in to that mode - i am used to it. registax / astral image used to scare / frustrate me, but through patient explanations from bird on why registax does what it does (thanks mate), i now feel i am coming to terms with it. there will be things on the sct's that need precise attention to detail before imaging at high mags etc that would seem mind boggling to me.
There have been great tips from ice and rumples along the way on wavelet settings and histograms that initially sounded prescriptive, but using those as a starter, i now know pretty well what will do what and hence present am image the way i prefer and can tweak it.
I was able to write that "process" off the top of my head as i have done it soooooo many times, but i would have to be honest, the most time consuming / frustrating process is still aligning the finder and initially finding the planet when working at 1000x +. All the other things are now part of the fun. I am comfortable with them.
We will all continue to learn from each other on a technique or piece of software or camera setting. It is part of the fun!!
David what a fantastic result.
You've really got the technique down to a T.
Just a quick question.
Is it true that you can only perform a LR Deconv on greyscale images only, not colour. I have a trial version of Astra Image and was wondering if colour LR was disabled.
You can only do deconvolution on greyscale images. If you bring a colour image into Astra Image, the first thing you do is "Process->Split into colour planes". You then work on each R/G/B image, and then recombine them using "Process->Recombine into RGB".
the festooning and white swirls in the main bands are amazingly detailed... I feel slack and slip-shod in my approach too after reading your exhaustive checklist 
