Log in

View Full Version here: : Jupiter 10_09_09, Playing around with processing


desler
12-09-2009, 11:06 PM
Had a bit of a play the other night in probably 5/10 seeing.

Processed by using virtual dub, RGB split, Registax 5 and Astra Image, Spent a bit of time playing with Deconvolution.

Comments welcome.


Darren

Thinking a DMK may be the go, trouble is Justifying it, especially with the Banker saying, "Not Jupiter again!"

Take it easy.

rick01
13-09-2009, 01:57 AM
Nice detail with SPC900 :). You could capture shorter avis, cause of fast Jupiter rotation. I usually capturing around 90-120 seconds in 5 meter focal lenght.

In astra image I always using Maximum Entrophy Deconvolution filter. Looks like the best solution for my QHY5 mono. Which one you used? Did you do some sharpening in Registax?

Cheers, Roman

Harpspitfire
13-09-2009, 06:12 AM
interesting, since you images FAR exceed mine, i have some others questions- what were your capture fsp and total frames?-- one other thing- what does it look like without using RGB split?-- i never could any better with deconvolve in AI for some reason

Quark
13-09-2009, 10:20 AM
Hi Darren,

The elongation of the moon and its shadow would seem to indicate that your capture times were a bit too long.

Have never had to use RGB spit as I have a DMK, if it does the job I assume it does ie; split your colour data from the ToUcam back into R, G & B channels, then it looks like when they were recombined in Astra Image that the channels were not quite aligned correctly. When you recombine in Astra Image it pays to use the zoom function, maybe twice, to make it easier to align.

Initially align the three channels on a Jupiter feature, not the moon.
Finish processing and save that image, then reload the three channels and recombine them again but his time use the moon as the alignment feature. Finish processing and save the second version and then open both of them in Photo Shop and copy then move the aligned moon onto the image that was aligned with a feature on Jupiter.

I recenty asked this question of how to fix the image with the missaligned Jup moons, I think on the Tips & Techniques forum and Mike gave a detailed descrpition of how he does it, which I found worked very well.

Hope this is of some assistance.
Regards
Trevor

desler
13-09-2009, 12:55 PM
Hey guys. Thanks for the tips.


I'm using the toucam through a 5 x powermate to increase resolution. I'm normally limiting my captures to around 100 seconds at 10fps, which results in 1000 bmp's. I try to keep the exposure meter at about the 220 mark.

For some reason, registax 5 doesn't open if you put in more frames than that, I get an error message, "path cannot be found".

I got to this processing path by following mikes planteray imaging page, and have modified it. I split the AVI's using Virtual dub, I'm not using PCFE to centre the images, then RGB split to get the images into Red, Blue and Green.

I then process each channel in Registax, using wavelets to sharpen.

From there I open each channel in AI and use Maximum Entropy Deconvultion. I then recombine the RGB image and run either further Maximum Entropy or just adjust brightness and Gamma.

I haven't played with adjusting the edging of the planet or sharpness of the moon as per Mike's instructions. From your comments Trevor, that looks like the next step in this ongoing saga.

Thanks for the feedback guys.

Darren

Quark
13-09-2009, 02:23 PM
Hi again Darren,

This is the link to Mikes instructions regarding moon alignment.
http://www.iceinspace.com.au/forum/showthread.php?t=49665

Cheers
Trevor