Have had a couple of breakthroughs lately with my planetary imaging. With a little bit of good seeing tonight, here are my best Jupiter's yet.
First breakthrough was finding the right settings in Envisage to eliminate onion rings. Up until a couple of nights ago, all my Jupiter images (and my Saturn ones, for that matter) have had a bad case of the onion rings which meant I could not 'push' the processing very far. There are some advanced settings in Envisage that allow you to increase the gamma. With this done, the onion rings are gone!
Second breakthrough is a DIY SCT cooler. I used this tonight for the first time. Had it in for about two hours before observing. Having a nicely cooled scope leads me on to the third breakthrough.....
Collimation. Because my scope has never really been cooled down properly before, collimation has been very difficult. Tube currents have made the out of focus rings very 'shaky' and difficult to see properly and, though I have been pretty close, I have not been able to 'critically' collimate my SCT before tonight. I was able to use my 6mm eyepiece to collimate on Zubenelgenubi and was rewarded with the best views I've had of Jupiter. I think I still have a little way to go with collimation, but having a cooled scope is going to help.
Seeing tonight was about 5-6, possibly going to 7 at times. All images were taken at 0.25 sec exposure, 100% gain, offset of 55, gamma was 1000 (as per the setting in Envisage), saturation was bumped up to 1450 from a default setting of 1300. Images were captured in Envisage as BMPs. Each run was 3 minutes.
The first run was started at 20:50, the second at 20:56, the third at 21:19, and the last one at 21:26.
Processing was done in Registax - nothing fancy, just a straight stack and wavelets. Now that I have some decent data I can try some other techniques such as RGB splitting and deconvolution.
Good onya Adam. Great to see you succeeding with the LPI. It certainly can be a challenge. I'd be interested to see how your images would look with heavier waveletting. Do you remember the settings you were using?
Paul, the wavelet settings I used were 4-30, 5-40, 6-50. I tried not to hit the wavelets too hard for this intitial run.
I will reprocess no.3 & 4 using ppmcentre and possibly some deconvolution. I might even try an RGB split and see what comes of it. Where do these planetary guys get the time for all the processing?
I didn't mention in my first post that I was using my 10" LX200GPS and 2.5x Powermate (but now I have!). Also, my dew heater was on quite aggressively, as it was a bit dewy.
Thanks, Eric. By the way, fantastic images you've been posting. I'm glad you've been getting some clear skies.
Here are images 3 and 4 processed through ppmcentre - cut to 300x300 and qestimator used. In Registax, I used a 95% quality rating (had used 90% for the earlier post). Image 3 is 106 stacked images out of 408 captured, image 4 is 103 images of 387. Wavelets were 4 - 30, 5 - 40, 6 - 50. Even though these were the same settings I used earlier, I felt that more detail was brought out because of the higher quality images.
Thanks Mal! Still only have 2 Jump Start batteries! I run the Cooler from the same battery I use to run the telescope - when I am ready to observe I take out the cooler and plug in the telescope.
Last night was very nice and clear - had a look at a few nice galaxies as well as Jupiter - too bad tonight is looking like a complete wash out - oh well, more time to process, I guess.
Here are the separate R,G,B results for the third image. I split the 106 best frames with my own VB application. Stacked each separately in Registax. Wavelets were 4- 30, 5 - 40, 6 - 50. Next step is to combine them (possibly in AIP4WIN).
Separate R,G, and B files for 4th run. Same process as for the third image except this is the best 103 frames of 387 captured, split with RGBSplit, stacked in Registax, wavelets at 4 - 30, 5 - 40, 6 - 50. I'm quite happy with the detail that has come out in each channel.
By the way, the RGB files for this and the last image are in the original orientation as captured.
Adam, these are fantastic and I reckon the best Jupiter images I've seen taken with the LPI. Excellent work!!! Love the DIY cat-cooler too. I recently bought a commercial version.
The two images attached are the RGB combined images from run 3 and run 4. I like these images but I don't think the RGB processing has given up any more details. These were RGB combined in Paint Shop Pro.
Just having a play with RGB combine. I created a 'pseudo' Luminance layer and combined with the RGB files. Highlights the detail a little better. Anyway, here they are. The left image is the best colour process I have managed (ie no RGB split), the right image is the LRGB combine. Have done this for run 3 and run 4.
Still playing around with the images from Friday night.
This time did a VC deconvolution in AIP4WIN on each of the RGB channels and recombined. Gaussian PSF - 2 pixels and 16 iterations. Recombined in AIP4WIN now that I have figured out how to do it!
After lots of playing around with LR deconvolution I have settled on VC deconvolution as best for these images. Here are the 'final' versions of the 3rd and 4th runs from Friday night.
Workflow for each run was:
- capture in Envisage,
- ppmcentre for centering and reduction to 300x300 and image quality estimation,
-images 95% and better quality (106 frames of 408 for run 3, 103 frames of 387 for run 4) were split into R,G, and B channels using my own VB app,
-each channel stacked in Registax and wavelets applied (3 - 30, 4 - 40, 5 - 50),
-each stacked colour channel converted to greyscale and Van Cittert deconvolution applied in AIP4WIN (Gaussian PSF, 2 pixels, 16 iterations),
- colour channels recombined in AIP4WIN,
- into Registax for a tweak to the RGB alignment as it was slightly out (and I don't know how to do it in AIP4WIN, yet ),
- a very slight unsharp mask and small adjustment to brightness in Paint Shop Pro,
- borders and text added for presentation in Paint Shop Pro,
- and adapted for Ice In Space by hitting it with a big hammer.
The question I wanted to answer for myself with all this processing was whether it was worth all the extra effort to split into RGB and use deconvolution. I suspected that, with my smaller image scale, there would not be much of a difference with the extra processing. To my eye, at least, I think the extra processing has brought out more details - maybe only 5-10% more, but, more nonetheless. Especially for images taken on nights of less than perfect seeing (as these were) I think the extra effort is worth it.
Thanks Lester, it's only taken me 12 months to get this far!
Even though most of the planetary imaging talk on IIS is about ToUCams and NexImages, the basic concepts of gain and gamma and post processing are the same and I have been able to apply the lessons learned here to the LPI (very slowly and randomly!). Not to mention the inspiration of seeing all these great images posted on IIS - including yours.
![[1ponders]'s Avatar](../vbiis/customavatars/avatar45_9.gif){kind=avatar}
