To my understanding, deconvolution brings out more details by analyzing the star shape and how much it needs to be corrected to form a perfect circle, then by analyzing the image pixels and applying that same correction across the entire image, it brings out details that were previously a bit smeared. However, I sometimes have some issues with dark halos when deconvoluting.
I'm not very much into photography, more of a film-maker commercially but this guy is a reason why i started thinking photography is way more than posing and headshots. 🤍 Genuine admirer from India. Keep up that awesome work.
Nice to see you incorporating Pixinsight, quick tip - the StarAlignment process in Pix is a very easy way to align the channels, and will be more precise than manually aligning them in PS. Just select one of them as the reference view and apply it to the other
@@AstroBackyard I discovered this the other day when I watched Peter Zelinka’s video titled, “9. Create an H Alpha + RGB Image”. Watch the first 10 minutes to see Peter explain how to set a reference frame in DSS for different stacks. I tried it and was blown away! So simple and I had no idea it was there all along.
@@AstroBackyard I forgot to add colour correction - either use colorcalibration, photometiric color calibration (PCC) or a script like "autocolor". Note: the maker of the autocolor script does not recommend using it for SHO/HOO combinations. There's also the SCNR tool for removing (typically) the Green cast from RGGB based sensors. There's a trick where you can invert your image and run SCNR (Green channel) to remove purple stars. Nice 'n' easy.
@@davidleejenkins Sadly, he merges the color channels from a monochrome camera, which I don't have. I use a DSLR camera, and I am constantly having from making a HaRGB using a reference frame.
If you use the compute offset in DSS and stack your group tabs individually it will align your different filters for you. Your buddy Nico has a great video on it!!
I am glad I discovered your channel great little tutorial btw. I am new to astronomy and astrophotography. However I have professional background in seismic data analysis and imaging. Deconvolution, is a process that undoes some sort of filter operation on your data (seismic, image, acoustic, etc..) that in general has degraded your image in some fashion. Typically, introducing blurriness around sharp edges because of a variety of reasons. An every day example using sound is the effect of listening to someone with a pillow over their mouth. Remove the pillow and you can hear the frequencies of the persons voice absorbed by the pillow. Removing the pillow is a form of "deconvolution". Sharpens the sound of the voice by restoring the higher frequencies. When a filter is applied to any data, the process is called "convolution". Deconvolution or inverse filtering (the other name the process is known by) attempts in a statistical manner, to undo a filter process. In this case, an assumption is that the edges of objects in the image have been blurred by filtering (e.g. the atmosphere, optics, sensors of the camera etc..), and an inverse filter is designed (deconvolution) to, in this case, "sharpen" the edges of objects in the image (stars, transition in gas concentrations in the nebula etc..).
Fantastic info from you Trevor! I’m right now stacking my first 3 hours of Andromeda frames, taken with a DSLR and a Star Adventurer I got for christmas. All of this started after watching your videos about three years ago! Thank you for the inspiration and great teaching.
I just like how easily you teach people your processings,it was so useful and the way you show is so perfect and very understandable..and this picture is so good! ✨
Trevor, you should really denoise and do deconvolution in the linear stage. It is best to fix the image first before stretching. The simplist way to do this is for you to down darkarcons EZ processing suite. Many RU-vid videos on this. Keep up the great work!
Awesome video as always. Trevor, tip for starnet - if you have an nvivia graphics card you can speed up the starnet part drastically by offloading the processing to the GPU instead of the cpu. And the faster the card the better. You should look into it.
You can use Script -> Batch Processing -> WeightedBatchPreProcessing for a very DSS-like experience within Pixinsight with the benefit of better calibration settings and defaults, leading to a better final stack. The (small) added bonus of this is that your filter stacks are automatically star aligned to each other at the end.
Your video quality from your new office is wonderful. Having Ashley behind the camera makes everything SO pro-looking! Looks like I now have to buy Pixinsight!
Feedback. I have just used this idea to re-process some frames I captured on on 23 May with my elderly QSI 583, which has been modified to work with Canon lenses. I used a f 1.8 50mm Yongnuo lens to capture subs of eta Carina, which is well placed in the evening in the Southern Hemisphere during May. Mapping H Alpha to red and OIII to blue and green works really well. Have done the basic processing with Nebulosity version 4 - a programme I am growing to like more each day. Thanks for this great idea.
I follow your channel for a long time and just coincidentally saw your subscriber count. Damn, Trevor! Good job and you deserve it :). Keep going and break through that million subs border. Make astrophotography famous :D
I think I will try giving your data a download over the 3 day weekend for memorial day in the 🇺🇸 and following along. My biggest struggle is astrophotography is the post processing. Thank you for the lessons, very awesome! Clear skies
I’m an Astronomy and Astrophysics major at Florida Tech, I’ve had a telescope but never did much more than peak at the moon and our Jovian pair, but recently I’ve been obsessed, especially after using the telescopes at Roque de los Muchachos :) I have my own post processor I made in python but I definitely will try this and compare :)
Thank you very much for the video. So much work in it. I can feel how you want to share the enthusiasm. I am so much in love with the H-Alpha stretched image. I feel there is so much depth in it. It’s like showing through a window, a landscape, so much in dimensions, 3D. I downloaded the H-alpha tif file and ran a stretch and i’m so happy you shared it. Thank you very much
Absolutely leave the stars as they are! If humanity survives this decade, someday someone lost inside that nebula might turn to your picture for navigation and say: oh, good, this is exactly what I'm seeing here! Loved your ninja trick! You should absolutely post such details for whoever is skilled enough at astrophotography to get to use them! I mean: look what happened to the Rocketdyne engine building techniques!
Just bought my first telescope and have caught the bug. Thanks for sharing your knowledge. I’m so excited. I mean it’s not that I don’t think earth is nice. I just enjoy seeing things on my own. Kinda like you did with Jupiter. I will see Jupiter in around a week. I cannot wait. Feels like I’m waiting in the line at Disneyland as a child and I 40.
Blown away with that image Trevor! The star separation looks to make things a lot easier. Iv been putting off using PixInsight for a while but its becoming more tempting! Great Video, Learned a lot!
Thanks a bunch ! You inspired a very technical person, who usually was initially running a 12mm F2 AF lens ~ mostly for vlogging and other bokeh/travel shots I feel my lens & wake-up times where usually designed for night-time photography and generally night time work. Astro & Night-Time photography both are very complex topics and deep rabbit holes. Yet as someone who's passion is to learn new complex things (hardware-tech & music), Astrophotography seems the next logical rabbit hole to dive into Learned a lot from you towards colour mapping, bit from the chemistry side and general good practices. Soo a general thanks is more than deserved ! Once borders and travel ease's, i'd like to figure out a good middle step between technical tinkering ~ and lightweight travel-movability * towards sky-photography (into work-vlog) 😃 Soo for now, you'll remain a valuable learning channel & inspiration to get into this deep topic. Keep on the good work ✨ * Might need some advice for lightweight gear towards sky-tracking ~ or well build gimbals would just do it on fixed coordinates
Hey Trevor, nice tutorial. With the HDRMultiscaleTransform tool, try ticking the boxes for "To Lightness" (preserves the colour of the image) and "Lightness mask" (applies the algorithm to the brightest areas of the image to give improved dynamic range). That way you won't get that dulling of the image you were experiencing. I often find 6 to be a bit aggressive and often set to 7 or 8 depending on the result. (Can also apply a Range Selection mask first to the image to more isolate the effects of the HDR tool to the area you want and help exclude some stars). Great final image by the way.
Thanks that was very helpfull. I've just redone the edit of recent picture taken with this method, and the result is astonishing !!!! Thanks a lot !!!!!
Certainly, an incredible image very well done. Unfortunately, I had no freakin clue what you were talking about. Not your fault btw. Fortunately, the final image says it all. I'm a total noob with astrophotography and just recently purchased my first telescope. But I have subbed to your channel and hope to learn as I find all of it extremely interesting. Thank you!
Deep sky stacker will auto align images if you set a frame as the reference, I just pick the one with the highest score. Right click then select it as the reference.
I believe if you applied the star layer as a mask in pixinsight and then inverted the mask, then the morphological transformation should only affect the stars so next time you can try that
Hey Trevor, amazing shot and thank you so much for sharing your data files! One question: how did you add the grayscale images from one PS project to the other in the color channels? When I copy and paste the layers from the other project they stay grayscale.. Do I have to copy and paste the layer into the channel or a copy of the pixels? Thanks a lot!
Just a comment, since I stumbled over this. Starnet does not actually produce a star mask, it extracts the actual stars. Which is why this whole approach works (or you would need to use a star mask to extract the stars yourself, which I suspect would not work super-well?). Anyway, thanks for the walkthrough, I was always curious how you managed to tease out this much detail!
Really nice video, and an amazing final image. The more time I spend on pixinsight the more I love it. The star alignment features are very good too. It can stack too but is so slow I just use dss! I am very jealous of your camera and filters ;)... and mount (did you use the eq8rh for this), I hope by the end of the year I will have an qhy268m or the 2600mm ... but for now I make do with a modifed canon 5dmk2, and I am happy! Thanks for all your content Trevor, you and a few other creators were whyI got into this awesome hobby!
Trevor your video production is brilliant I always watch your videos whether it’s relevant to me or not. One point on aligning the Oiii and Ha, I load the reference file used for my Ha stacking to my Oiii before registering which aligns the Ha Oiii with each other.. It goes without saying I untick the Ha reference file as I don’t want to include it with my Oiii. Once again Superb video👍
Loving your tutorials Trevor, as I follow through step by step on my own data - You've got me hooked on Pixinsight in under 5 minutes! - When it comes to aligning images, I tend to do this in DSS - I group each filter set of lights (as you did the Per night groups) then once all registered, I choose the highest scored light from all the filters by right clicking and selecting it as the reference frame. That way all my data is aligned before it ever gets into editing. Is there a reason you would recommend doing this differently or is it just a case of personal preference? Thanks again for your amazing and inspiring channel!
Hi, thanks for the vid! I'm an inexperienced Pixinsight too, mostly following along someone's tutorial without knowing what I'm doing 🤣 but I suggest you try the Weighted Batch Processing script, be curious to know what you think.
Hello, I'm new to astrophotography. I've been following for quite some time but I have recently picked up a Sky-watcher EQM-35 Go-to mount, Sky-watcher Evostar 72 APO refractor and a Canon EF 75-300mm lens. I'll be using an older Canon EOS Rebel T2i for images and I'm having a hard time deciding on processing software. I would like to ask if Adobe Photoshop Elements 2021 & Premiere Elements 2021 would be a good choice? If not, do you have any other recommendations? I'm trying to stay away from subscriptions and just purchase a standalone version.
can you add some free RAW photos for us to practise.I dont have a astrophotography camera .I would buy soon but before that I want to get acquainted with the image processing part.Thanks
Out of curiosity - any reason you don't use the reference frame feature in DSS? That way you just need to RGB combine in PS and everything will be aligned for you. Great tutorial! Always helpful to see your workflow.
nice tutorial i just download pixinsight to and i use your data to practics, but when i doing auto stretch the image to prevew like you and is more brighter then how its look like in your video, why that? thanks
23:55 re the star colours - could you take a single frame of each of RGB, turn them into a single file, apply the star mask, clip that to the star layer and apply one of the colour blending modes? That should give the natural colour, whilst noise shouldn't matter as you're only after colour, not detail, in very discreet parts of the picture. Re-focussing probably isn't necessary either as only the colour would be coming through. Or am I way off?
Hey I'm trying to follow along but having some trouble aligning the layers in photoshop. I have linked both the HA layers and the OIII layers, But when I try and drag the starmask to align them I can never get them to align. It seems that the HA and OIII are rotated slightly off from each other. What am I doing wrong?
Quick question: I have a Celestron PowerSeeker 127EQ, I don’t have any of those special cameras, I use my iPhone, am I still able to get a result with these programs? I want to make sure I’m able to start deep sky imaging with what I have
I’m confused. You take monochrome images and then add color? Are the color choices arbitrary, or is there data in the images that tells what colors things actually are?
Perhaps you know this but I thought I would mention that DSS can align your images for you by loading them all at once (though processing them one at a time) and choosing a reference file, if I remember correctly. Anyone, please feel free to correct me if I made a mistake. It's been a while.
Awesome I just got a star adventurer 2I cause of your videos I do wonder when you did the kit lens video with nico and used the ball head with the dslr and 70 to 300mm lens did you have much issues with tracking accuracy since Im not sure using the dec bracket and counterweight makes much sense when my camera and lens weighs only 2.5 pounds
Nice to see you getting a bit more technical. I will come back to this when I start using pixinsight and ps. Currently I pay about 50 quid for star tools which is great but very much the budget option. It would be interesting to see what you could so with that but I appreciate you are perhaps beyond using cheap software. P.S. I feel like you are starting to strike the right balance with the adverts. Not too long but still gives the company paying you the exposure.
I wouldn't consider Startools poorer for being cheaper. Less artistic, more scientific in its treatment of the data. Well written, good UI, uses the gpu, much more modern and up to date.
@@philleng480 OK fair point. I've had really good results with star tools. I have not used pixinsight or any photoshop beyond lightroon so have no idea if they're any better.
Many thanks for this, it's very helpful! But I feel like I missed something--how did you get your starless TIFFs into the colour channels within the new Photoshop image? Every time I try to do it, the result is a blank screen.
I've used it a few times. Feel like it's better for brighter targets. Did an amazing job on my Orion nebula image but not so well when I shot the elephants trunk. It's worth a try and it does the stacking and processing in one application. Any free software is worth trying out
@@philleng480 if it matters it's good to know that dark flats replace biases in calibration math. Just drop your df's into Siril's biases folder... assumes you have darks to take bias out of the lights.
By stacking you’re creating an artificial image not as it’s truly seen common sense tells me this am I my wrong? Pretty much just getting pretty pictures ,by doing the stacking?
Hey I’m pretty stuck with my telescope and I’m wondering if u can help. So the problem Is, I have a Sky-Watcher skyliner 150P and the lense/eyepiece thing that came with it wasn’t very good and I was wondering what good lense I should get with it and if a Barlow lense would be good.
Nice video, thanks. Why do you actually need to load data from different nights into different groups in DSS? Why not just load all Ha from both nights into one group and stack them ?
