3.61
#...
·
·
2
likes
|
|---|
|
Dear fellow astrophotographers, Over the last 15 days, I captured about 24.3 hours of good quality data on M31 organized in 9 panels that I stitched in a 3x3 mosaic. That is about 2.7 hours per panel, which is not that huge but not that bad either consider weather, moon, and limited night time constraints. So far, it is one of my sharpest image. Thousands of stars and nebulae pockets from an distant galaxy. These are RGB data shot with an ASI294MC pro. I plan to add data using my l-ultimate. You can find the image there : https://www.astrobin.com/48wv2a/B/ I am looking for further improvement suggestions. So far, my workflow was the following Preparation - for each panel, check individual subs for quality before caligbration, alignment and stacking in SIRIL. - for each panel, image solving, mosaic by coordinate and mosaic trim in pixinsight - photometricmosaic to stitch the panels into one large image Linear processing - dynamiccrop - image solving and photometric color calibration - DBE 2x (second with reduce tolereance) - noiseXterminator (OMG that's good !) Non-linear processing - Maskstretch (OMG !) - Starnet - starless : curves (colors, saturation), HDRmt, curves, colors saturation to pull out blue and limit yellows, BG correction with a mask (background neutralization, curve for darkening and color desaturation) - stars : deconvolution with deringing, increase saturation, reinforce stars in the middle of the galaxy by applying curves with a mask constructed with ACDNR on the starless (i.e. masks the stars that are not on the galaxy by extracting the corresponding mask from the starless with ACDNR), curve - pixel math - final touch with curves. I am using pixinsight since 48 days, minus 12 days of astro-free family holidays. I would greatly appreciate any inputs to get better. Many thanks & clear skies ! Patrice |
1.43
#...
·
·
3
likes
|
|---|
|
Hey! Nice work on that picture. There is not much to say to be honest, but here is my 2 cents anyway. Starnet is a good alternative, but StarXterminator is much more powerfull in my opinion. You think noiseXterminator is good ? Try BlurXeterminator and be done with that deconvolution BS. Its is CPU intensive, but if you have a good GPU you can use CUDA to do some GPU acceleration (this is the way!) Also, I dont know if you use photoshop, but splitting channel in pixinsight and recombining them in Photoshop give you way better control and color rendition. For stretching, get the GHS Process to your arsenal. its free. Also, there is a revised color mask script that work way better than the old one if you want to punch some color. Also i would only work on detail via my Luminance. My last step as always been to use Dark structure enhancement script. Thats it for now. I am on my phone but if you need link, I can send them when I get to my computer. Karl |
0.00
#...
·
·
1
like
|
|---|
|
Excellent work! I’ve never done a mosaic and this is inspiring me. +1 on the StarXTerminator and BlurXTerminator. Those two processes are worth their weight in gold. I have also been using the Generalized Hyperbolic Stretch process with I think great success. It allows wonderful details to be pulled out. Check out videos from Entering Into Space on his implementations of GHS in several of his images. https://youtube.com/@enteringintospace4685?si=LXD9mx0iJQNuOQPQ |
|
3.61
#...
·
|
|---|
|
Thank you very much, to you both. I installed GHS and will start looking into it. I was using something fairly similar in SIRIL but have to confess that I had limited success with it so far. I will dig in this time as this appears to be a fairly known tool in pixinsight. I already watched the lunatiko tutorial and discovered how to set the symetry point quite accurately and that we should proceed iteratively. Otherwise : - I do not use photoshop; I use GIMP with some success so far, even if that implies to mimic many PS tools. I like that because of get to understand what the PS tools are doing. - I do know neither the revised mask tool nor what you refer to as my Luminance tool. I am happy to dig in if you give me a pointer. - I did not use dark structure enhance on this image. However, I abused it on my Veil nebular images. I applied it quite early in the workflow. Is that bad ? - And yes : the terminators are apparently good. I now have noiseXterminator replacing EZ denoise. I will consider that at a later time. Not everything can be bought at the same time (even if the price is modest here in comparison of the rest of the equipment ! Thanks a lot ! Patrice |
1.51
#...
·
·
1
like
|
|---|
|
Very nice image! I would second (or third!) the comments recommending BlurXterminator, NoiseXterminator, and StarXterminator. Excellent tools. And I could not do without GHS for stretching. It is a game changer IMHO. One change you might consider in the future is to go to a monochrome camera. For practice, I have been processing monochrome data from the Moana Project since I can't image (with my color camera I might add) at our summer place. I have been very impressed with the quality of the monochrome RGB images, even when viewed at 100% (ASI1600MM). It would be hard to go back to OSC. George |
|
1.43
#...
·
·
1
like
|
|---|
Thank you very much, to you both. https://youtu.be/6Kg3jfaGgeM?feature=shared Thats the color mask script I was talking about. I tend to be very conservative with the dark structure enhance script. Many pass at very small increment always give me better result. To be honest, I never really worked with OSC camera. So debayering and stuff leave me clue less. That being said, RGB image is RGB image so my way would work too i guess. I would: -Dynamic Background Extraction -BXT -NXT -Stretch my RGB but not too much, just so my stars look great. -SXT, save my star image for later. -Extract luminance -Stretch my starless RGB to my taste. At this point I dont care about detail I want nice color -split RGB channel, and save a 16 bit tiff copy. I would now work with my LUM. Using all the detail enhancement tool from pixinsight to bring details in play (HDR, sharpening,Dark structure enhancement. you can even experiment with BXT with very small increment) -Using the tiff RGB channel I would use Gimp or PS to make the color punchy. -Once happy i would bring my new RGB file in PI and add the luminance to it. Final touch to the final image (ie: color mask script or game script and some color adjustment) and then add back the stars. This is what I meant by work on luminance for detail. Karl |
|
1.43
#...
·
·
1
like
|
|---|
|
And also, you can try thos Terminator tool 30 days without limitation for free I think. Give them a try but watch out. Its like drugs... you will want to pay for them lol Karl |
11.22
#...
·
·
2
likes
|
|---|
|
Hi, Nice work. I have no comments on your workflow. When I look at the full resolution image, I feel you applied too much smoothing (noise reduction) to the image. It looks super unnatural. Of course, when you tune down the strength of noise reduction, noise will appear. I think that just means you will need to add integration, or also tune down the strength of contrast/saturation boost. Just my 2 cents. |
|
3.61
#...
·
|
|---|
|
Many thanks for these precious comments. @Karl Theberge : thanks a lot for the explanation about the luminence. I was indeed working in a very similar way with GIMP before having access to pixinsight. I will get back to that in Pixinsight. Thanks for the workflow. I will try it that way. @Wei-Hao Wang : Thanks for this honest statment. I may have gone a bit enthusiastic with noiseXterminator and also with the color saturation. I agree that this might be a bit too much. I will take that into account. Thanks. @George Hatfield : yes, I know that mono is the Way. However, at this point, I fear a bit i) the price tag and ii) the hassle, especially with flats, as I am imaging with a newtonian. I feel I can progress a lot before the camera being the limiting factor. But yes, in the longer term, that's is one logical development. Those a great comments. I hope that I will be able to get back with a new version. Hopefully, the new version of get an astrometric solution. The latest one did not. Cheers, Patrice |
|
1.43
#...
·
·
1
like
|
|---|
|
Well Mono is expensive and a bit time consuming BUT it worth every dollar. The QHY533M is a really decent one. You dont need dark frameat all, you take only bias once in a while. I do mine at season change or when I change my imaging temperature. Flat using NINA is a breeze. 5 min and you get all the flat you need. I personnaly own a automatic flat panel and this with my rotator are the best gravy investments I did for my sanity. The full system with 7 slot filter wheel is not that expensive. The worste is the filter. You have to get it right the first shot or you live with regrets. I would go with astronomik LRGB+ 6Nm SHO filter as budget option. Otherwise, my 1st pick would be Antlia 3NM + LRGB. I personnaly own Baader SHO (garbage) and astronomik LRGB (really nice). I will never buy baader stuff ever again. Mono is the way and so are BXT, NXT and sXT 😅 |
|
0.00
#...
·
·
1
like
|
|---|
| @patrice_so I think you are doing the right thing by sticking with the equipment you have and learning. The one thing I wish I had done is spend more time with OSC in the beginning. I think it take a lot of other things out of the equation and allows you to focus on many other important things. Also, the stars aren't going anywhere....take your time an enjoy! |
|
1.43
#...
·
·
2
likes
|
|---|
The stars aren't going anywhere....take your time an enjoy! ⬆️⬆️ This ⬆️⬆️ Dont go sell your stuff and go crazy on mono. Same with PI stuff. All mono ppl say is the way.... same goes with equipment brand fanatic or old school ppl. There is also visual astronomer that prefer that to astro imaging, go figure. Karl |
