View Full Version here: : Startools / Pixinsight Comparison - Information sought
Outcast
29-07-2019, 08:57 PM
Okay, at the risk of starting an argument which is not my intent, I wonder if people with experience in perhaps both of these processing tools can comment on what they do or don't like about these tools.
I currently use Startools, I will readily admit I don't really understand it that well (I haven't yet really read the manual properly) so I tend to use it in 'auto mode' & I find it to be somewhat aggressive when it comes to processing... it's not that I hate it or anything, I just find that I'm relying on the auto functions & I find I have to do, redo, and redo things until I am remotely happy with the outcomes... What I did/do really like is the modest expenditure... I know many of you are using Startools to achieve amazing results so, clearly it is indeed possible & it is my knowledge & understanding that is holding me back...
I read from other peoples experiences with Pixinsight that it might be a better option & that once you learn how to use it may actually be more intuitive to use than Startools seemingly allowing greater control & flexibility... but, the price puts me off a little... If I go this path, I will still have to take the time to learn how to use it properly....
So, it is very clear to me that whatever I ultimately decide upon, I'm going to have to put in some considerable effort reading, understanding & learning how to use processing software to make the most of any data capture...
So.. do I take the plunge spend the money & immerse myself in learning how to use Pixinsight; I understand it's a different tool with a very different work flow than both Startools & for that matter other more conventional image processing software or, do I immerse myself in learning more about Startools & getting more competent with that??
If I stick with Startools, other than reading the manual & joining the forum.. can anyone recommend any video tutorials that will help me understand what I am doing please... conversely, if I go with Pixinsight, I will need similar... I am more of a tactile, observing learner rather than read a book/manual & figure it out type...
All opinions welcome but, if you are going to make assertions that one is 'better' than the other, please take the time to explain why you feel that way & if possible, back up your assertion with objective evidence please...
Thankyou in advance
xelasnave
29-07-2019, 09:09 PM
I have decided to become proficient with startools and maybe then but only then consider spending more money...I went thru a phase you describe with startools but after a while that changed...I use the bits I use and find that sufficient..I then use Photoshop..and gimp..each has a little something.
Read for sure but I am a trial and error type.
I do think startools has much more to offer than most get out of it but probably so does pixinsite.
Think of bike racing..start on a125 and when you can wring its neck then 250.. the reality is most folk can never wring the neck of a 125 but insist they need a bigger bike.
Alex
Startrek
29-07-2019, 09:51 PM
I’ve only used Startools so I can’t compare with anything else and the reason behind choosing it was it’s simplicity in structuring your workflow ( it only requires 5 or 6 modules to achieve a good result but has another 5 or 6 to use for more advanced processing) and the ability to go back on your work and compare outcomes
There are no histograms , levels and curves to adjust, with Startools you see your progress straight away in front of you and can change things easily
As a rank beginner I personally think Ivo has created a simple user friendly program for entry level astrophotographers who can build confidence the more they use it
Admittedly I found it a bit daunting at first but after a few weeks I started getting some good results
There are some good YouTube clips on Startools in particular the one on the Astro imaging Channel where they host Ivo Jager the creator of Startools. He spends nearly 2 hours on a processing workflow
The manual is ok but I found the various YouTube clips very helpful and provide great tips on your processing
The Startools forum is also a great resource and I have resolved many minor issues looking at the various posts
I will be using Startools for a while yet unless something else comes along which is more intuitive and produces better outcomes
Startrek
29-07-2019, 10:04 PM
For Startools to perform well it does need clean linear stacked raw data which has not been stretched or colour balanced / adjusted
I use a DSLR and stack my raw data (Cr files ) in Deep Sky Stacker and save the final stacked image as a fits file ( untouched ) This fits file is then loaded into Startools
In Startools I select option 1 on the left hand side to get processing underway
ChrisV
29-07-2019, 10:32 PM
I've used both Carlton. Started with Startools for about a year or so, then moved onto PI. I'm glad I did it that way even though it meant buying two lots of software. So I reckon startools is a good place to start.
Startools likes:
- it is "relatively" easy to get started with it. Although seems odd at first. I seem to remember Ivo has some good youtubes of how to use it.
- You can also make images pretty quickly compared to PI.
- Its cheap, and I used it with DSS to prepare the stacked linear images.
- The tracking feature was neat.
Dislikes:
- It was generally too aggressive, although I could switch to manual to tone it down. The colouring wasn’t very flexible
- It was hard to make subtle changes.
- there’s not a lot of info about how to use it – but I might be wrong on this now.
- I had to write down a list of how to process, just to remember
The main reason I swapped was as follows. I started looking at some demonstrations of PI and thought it looked good. So I picked on one images that I had processed as well as I could in startools, then got a trial licence of PI and had a go with it - I learnt as much about it as I could beforehand. I pretty quickly found that it was better.
First the PI dislikes:
- It takes a long time to get to grips with it. I’m still a beginner after a year or so.
- Its way more expensive
- some say there’s not great resources. I don’t believe this
- its slower going processing images
The likes:
- It also has all the stuff to stack/integrate. Even though it takes a lot to learn it is really powerful. Has excellent weighting and ability to reject images based on well defined criteria. Plus other stuff that improves the initial linear stacked image
- the ability to make a wide range of really precise masks. These make it easy to target the object, background, stars, specific colours, or combinations of these.
- excellent filters for noise reduction, and the opposite for sharpening. Wavelets – you’ll get to love them.
- you can save different series of processes to do different things (calibrate/stack images, process OSC images, make masks, etc). So don’t need to remember everything – just open the process and off you go.
- although there’s a lot to learn, there are some really good resources out there
- its continually evolving and improving. The last release had some real game changers – improved selector for rejection and weighting images before stacking, neural network approach for star removal, etc.
Stop me now!
Outcast
29-07-2019, 11:15 PM
Thanks Guys...
Some very useful & great information in only a handful of posts thus far...
Thankyou Martin for the info on the Youtube videos by Ivo; I will go hunting those... I have some things I need to turn off in my DSLR but, yes, I use the raw, unstretched, unprocessed data in DSS & then the 'autosave' fits stack from DSS straight into Startools. Karlsberg sent me a link to a work flow for Startools which I use but, I'm still struggling to come to grips with it & understand what it does. I suspect I would have similar issues with PI too...
Alex, I've seen yours & and others results with Startools & they are impressive
Thankyou Chris, that is some great insight into both programs & on the strength of what you posted I'm now thinking I should persist with Startools, attempt to get to know it better & then decide if I want/need to move to PI... I still don't know how far I want to take this AP caper to be honest... I love the results & being able to share with others but, I still absolutely love the visual side of Astronomy.. AP is a side interest for me at present not, my main 'obsession' as such...
I'm interested in hearing others talk about their experiences but, even from a small number of responses it is easy to see there are pros & cons to both as well as no 'simple' approach to obtaining good results in the longer term.
Cheers
RyanJones
30-07-2019, 08:01 AM
I’m going to throw another cat amongst the pigeons in Photoshop. Carlton already knows this is the software I use so this is more for anybody else reading this topic but I’ll relate it back to my experience ( although limited ) with both startools and PI. The thing I love about Photoshop is that to use it and use it well for astroimaging you need to develop an understanding of what you are actually doing to the image. As you learn more and more about it, making changes becomes easier because you actually have an understanding of what is happening. From my experience ( again limited ) startools is very good at doing things for you and no doubt this suits many people on here very well and lots of people get very impressive results from it. Pixinsight on the other hand seems to be very technical but at the same time very powerful. Although it also does things for you, it demonstrates what it’s doing as it’s doing it. The ability to make the changes then apply them to the image sounds like a step too far for some but I like the way it works. For me I would say that Pixinsight is a great combination of PS and startools but it does come with a hearty price tag. I guess, in my humble opinion, it probably comes down to which way your brain works. All great software but one will suit others better than the rest.
casstony
30-07-2019, 09:13 AM
There's a free trial period for PI so you can get an idea if you like it or not.
My 2c:
I have a background in photography so have Photoshop as well as most astrophotography packages such as StarTools and PixInsight.
First up as an advanced photoshop user I assumed it would help me get into astrophotography but i quickly came to the conclusion it has no place in astrophotography except maybe to save a web suitable jpeg file, but not for anything else.
Something that people seem to miss with discussions such as this is they dont define what astrophotography means so the questions cant really be answered. To some it requires equipment capture and control capabilities, to others they mean something to prettify a single long exposure photo, to others it means video processing and others still being able to take hundreds of high resolution photos to align stack and stretch. Some people only expect/want/need some of these capabilities.
So in what context do you want to compare StarTools and Pixinsight???
For me ST is too limited and very niche. Its aim seems to be to produce pretty pictures rather than the most accurate. PI does everything pretty much in processing.
You need to comprehend when someone say "X does y" it typically means X is capable of being used to do Y by the dumb human using it, NOT that X is physic software that just automatically does Y. Try to comprehend (consult a dictionary) before commenting. If you the user don't understand the tools a piece of software provides, why and how and when to use them then the software is useless.
Every piece of imaging software I've seen produce stunning photos in the hands of people who understand them well. You also need to comprehend you are not stuck with using just one piece of software to go from input data to output image. You move data to software where you can best utilise it. The software is irrelevant, its the user putting in the effort.
For lazy people PixInsight will never work because it does require thinking and effort. To them StarTools is a more simplistic approach while people willing to put in the effort can dive deep with both packages .
I would say ST is easier to get started with but its java foundation is unstable and its processing approach tends to be unique so its harder to adapt tutorials and techniques you find for other software since the language is harder to translate (if even possible).
PI on the other hand has pretty much everything in there and often several times in several ways. Comprehending its components is its strngth and weakness. Weakness because it makes it harder to dive into and learn. But strength because as you learn new techniques you find PI already has the capabilities there waiting for you. This makes it software that grows as you grow. The complex terminolgy is often universal to other programs as well meaning its easier to migrate from another package or find a tutorial from something else and put it to use in PI.
I'll reiterate my start, it depends what context of astrophotography you are talking about. The way I work and capture PI is perfect for my widefield but poor for planets (it takes me considerable work in PI to match what AutoStakert produces simply.) Based on your way of working you may feel exactly opposite. And of course AstroArt, Nebulosity, AstroPixelProcessor and DSS etc are all very capable to users whose workflow gels well plus other programs offer other features users need making those programs more suitable in thos situations. For MY situation its PI as gold standard and AstroPixelProcessor a close second .
Outcast
30-07-2019, 01:17 PM
Cheers Steve,
you raise some very valid points to which I will answer thus:
Planetary & Lunar
1. Generally happy with planetary/lunar imaging & processing. Equipment wise I generally use a ZWO ASI224MC or my Canon 1100d at prime focus.
2. I have a Bhatinov mask for both my scopes.
3. I shoot video for planetary & lunar, use Autostakkert, Registax & some final tweaking in Affinity photo & providing the footage is good, my results are generally pretty good. Always searching for improvement though.
4. Still learning the in's & out's of both Autostakkert & Registax but, don't generally find them too hard to use.
Deep Sky
1. Very new to this & producing some mixed results.
2. Now using predominantly an 80mm Triplet on a Celestron CG5 Adv GT EQ mount, unguided at this stage imaging with an unmodded Canon 1100d at Prime focus.
3. Getting my PA's down to under 10 arc seconds in both axis & working to get them as tight as possible.
4. Still learning the in's and outs of my equipment & imaging routine but, I think generally getting better in this department though, haven't got past 90 second exposures as yet.
5. Processing is through stacking in DSS; learning more & more about that program, starting to understand it better & recently learned how to do group stacks with different exposure lengths, ISO's and associated darks.
6. Haven't started doing flats as yet
7. Image processing is then done with Startools at present & this is where I feel a little lost & unhappy. I'll be completely honest & say that I don't fully understand what it is doing at each step, I don't feel as though I have any real control & I have been struggling to find good information to allow me to fully understand it & gain proper control over what I am doing. It always seems to be very aggressive on it's stretches & it takes me a number of attempts to try & tone it down to an acceptable level by trying various things I don't understand properly I just don't feel if it is possible even to gain proper control of what it is doing.
8. I am a bit lazy though... it's not that I don't want to learn new things it's just I find it easy to make excuses to not learn new things or at least not learn them properly. I have read that Pixinsight is very difficult to learn but, very powerful once you have learnt how to use it & as you describe, grows as you grow...
9. Whilst the price is substantially higher than startools that's not the main 'put off' factor... Just how hard is this program to learn?? How good are the tutorials, guides, etc... does this make it easier to learn to use if one bothers to take time to learn through the tutorials, etc.. or, because I can be a little lazy, will it just frustrate me more at a higher price?
Hope that makes it clearer what I am trying to establish with my question & helps to make it easier to answer my questions... if not, let me know & I'll try and add more detail..
I'm not looking for a quick fix persay (although, that would indeed be nice) I'm just looking for something that with a reasonable amount of effort I can become proficient at using & feel like I have control & some sense of knowing what I am actually doing/achieving at each step.
Cheers
LewisM
30-07-2019, 01:44 PM
I have both. I have not used Startools since 2013 by the looks of things.
I can present a fairly completed image in PixInsight from the compiled data in about 15 minutes. I can't do the same in Startools. I also get VASTLY different images between Startools and PI, with PI being much closer to what we see from the professionals just using the PI defaults.
Yes, PI is much more expensive, but I have found it EASIER in the long run. It seems daunting at first, but it's actually not, especially if you read the Light Vortex tutorials.
I now do all my processing in PI - stacking through to completed image, and only put the final image into PS to tweak saturation a little and save as web ready.
Outcast
30-07-2019, 03:59 PM
Thanks Lewis, appreciate that perspective...
as you say, things may often look daunting at first but, as you gain understanding they become easier.
I am starting to see that I need to have a good look at some of these Light Vortex tutorials, see if they make sense to me & then if/when I think I have some understanding.. jump into the 45 day trial on PI
Cheers
LewisM
30-07-2019, 05:34 PM
I am still a little confused by the pixelmath routines, but I find if I use the LV formulae with different filter bandpass, then the results work. Makes integrating NB into RGB EASY.
irwjager
02-08-2019, 12:37 AM
First off, I'm the author of StarTools, so take anything I say with a chunk of salt. :)
That also means I'm not qualified to directly answer your question about likes/dislikes, except that I can tell you that I started StarTools in part precisely because, having a signal processing background, I disliked PI's IMO outdated approach to signal processing/fidelity.
Upfront, also, I can offer you, or anyone else reading this thread, a personalised workflow if you are having trouble with StarTools, would like a second pair of eyes, or are a PI user and don't believe the outrageous claims below. Whatever your reason, I've found this the easiest way to get people closer to an "a-ha" moment - whatever that may be.
Secondly, I can offer anyone here a preview of StarTools 1.5 (unreleased), which is a fairly major release, pushing signal fidelity limits again, improving NB features, as well as making ST slightly easier to get to grips with (happy to expand on that).
With that out of the way, the following is not meant as a marketing exercise; for that you can go to the website (which I won't link here either for that same reason).
So here goes;
StarTools is a wholly different beast to PI, owing to the vastly different signal processing architectures; PI uses the tried and tested filter-upon-filter linear approach, whereas StarTools performs time-independent processing designed to make the most of every last morsel of signal.
For example, ST will - if current prototyping efforts continue to be successful - soon even take into account your flats and darks during all your post-processing. That is how focused ST is on signal evolution and noise tracking. It wants (and needs) to know everything at all times - nothing is wasted. The more inputs the better.
The time-independent processing ripples throughout the entire application and touches virtually every aspect of the application; the interface, your workflow and - the reason for all this - the results. It is also an unapologetic memory and CPU hog because of this (though optimisation efforts are always under way).
Yes, StarTools has an enthusiastic following amongst beginners; it is easy to get reasonable results, quick, form even mediocre datasets.
However, the most common misconception about ST, is that this is because "guardrails" were put in. This is not the case. It is entirely because the engine mandates the mathematics behind transformations remains sound throughout time (e.g. as you process); the signal path needs to stay reversable so that processed signal can be traced back all the way to its initial state (and forward again). The latter is the core premise behind the Tracking feature. The time-independent flexibility of the sequence of operations, combined with the absolute need for congruent mathematics, makes it much harder to make a catastrophic mistake, perform a nonsensical operation or "overcook" an image.
There is no undo history or need for "process containers", because processing isn't linear. In StarTools, if you need to tweak something, you re-do it, because, apparently, it wasn't right the first time. "Just-another-tweak" syndrome is not a thing in StarTools, neither is related artifact or noise grain exacerbation. Closure on an image is much easier to achieve.
In fact, the time-independent nature of processing makes it preferable to perform certain operations at a much later stage than traditional processing normally allows for. For example, in ST, it is preferable to apply deconvolution after stretching, while color calibration is preferably applied towards the end. In both cases, the result remains mathematically correct (the operations are "simply" back-propagated and the results "simply" forward-propagated), but the postponement has yielded further valuable information about signal evolution that is used to achieve superior results.
The engine borrows concepts from how artificial neural nets work, are structured and learn (happy to expand for whomever is interested!).
All this, then, is the reason why people with more advanced knowledge of signal processing and the way commonly used algorithms work, have similarly flocked to StarTools. From a mathematical/signal processing standpoint, StarTools is objectively better. No discussion. Secretly I would love lots of discussion, of course, to prove that point! ;)
So this is where the PI bashing is going to start, right? Nope.
If you find yourself at an intermediate stage, besides the top notch pre-processing, going PI is not a bad move for educational purposes alone. Precisely because of that, for years now, I even have a direct link to the PI trial in a sticky on the StarTools forum (http://forum.startools.org/viewtopic.php?f=7&t=447). There is also a handy translation guide there between PI and ST terminology for people moving between the two applications.
Getting to grips with the mathematics, controls and considerations behind the algorithms and filters in isolation - as PI offers - is something very useful. I'd even go as far as saying that, what StarTools does, cannot fully be appreciated without a more fundamental understanding of what ST exactly tries to improve on.
Finally, for all this blabber and how things "are and should be" according to Ivo Jager, at the end of the day you should use the tool that inspires you the most and makes you the most happy. All other things be damned (but I implore you to keep an open mind and keep learning!).
Clear skies!
PS ST is fully native code compiled from C for each platform, and is not Java (ew!) :)
Ivo......great post mate, really impressed with your openness.
I’m a ST licence holder and have used it more or less exclusively on my Astro journey. I have been very happy with results I get from it and the feedback I get on my pictures on this forum show others are happy with how it handles data too.
Recently I have had reason to try out PI as I have purchased some narrowband filters and when first put through ST I was pretty disappointed with the outcome. So off on a very steep learning curve I went, immersed myself in PI and followed every workflow I could get my hands on to process this narrowband data during my 45 day trial period. I probably tackled it 15 different ways using as much of the collective knowledge on the web as I could.
End result was a realisation the reason I couldn’t get a good result from ST processing was my data was crap.....PI couldn’t do anything good with it either.
I’ll be sticking with ST for its ease of use, processing abilities and amazing developer support (as seen by Ivo’s posts here)......it hasn’t let me down yet.
Look forward to seeing the new release and will happily be a tester for you if you need one Ivo.
Ivo
I can come across as a PI fanboy but I do have a ST licence (not sure of its terms or limits, licencing is a big bugbear for me. If I buy a piece of software I expect to "own" it forever including upgrades etc as it grows, not to be bled for more money just for maybe a feature I will use rarely. I don't like subscriptions especially as I can't guarentee I will have money at a point in the future when the sub is due and I have never had a credit card in my life, never will. I want to buy once and be done with it with no worries about bills or software stopping working etc. I dont recall off hand any issues with ST along these lines , but licencing is a major factor in my buying decision and I am happy to throw away software that changes its licencing style. There is no software that is so essential and unique I cant live without it. I rarely touch my old photoshop. end licence rant. sorry ivo that was a general rant not ST specific.
That said, I havent abandoned StarTools, its still part of my toolset. As you mentioned the "A-ha moment" I havent experienced it with StarTools, Nebulosity or AstroArt. ST does have some impressive tools but maybe I am an old fogey but the linear workflow of PI is something I am comfortable with and for myself I have a detailed step process noted down and as I learn new techniques I find it easy to add or remove steps to suit. It also helps me understand the what/when/how/why of each step beyond just adjusting a slider. I do dabble at times in more scientific data measurement which PI allows me to explore and learn. The pretty picture end result means little to me.
Sometimes I process data to produce a picture, sometimes to measure a change I've made in my capture hardware, sometimes its just for fun. I dont know what specifically ST could have that will help me try it more often than I do. It'd just end up becoming like the other software if I did.
When I started AP I was annoyed at the way each piece of software seemed to only support certain input files and they all seemed to expect everyone was capturing a particular way so for me the first hurdle, since I capture with DSLR if the software couldn't take my RAW files to work with that was a huge negative for me. AstroPixelProcessor is the new "entry" level program and it only supports some RAW formats so that excludes people immediately from using the gear they have. I hate software authors trying to dictate you should buy specific hardware rather than providing an accessible path into the hobby with whatever people have on hand. Sure using a mobile phone to capture subs is not optimal but nor is it impossible. ST approach maybe more a younger generational thing like "apps" where instructions and control dont exist and a fast gratification of a result is all that matters. On occassions where I've tried my hardest with ST to process the same data I have with PI I have been unable to get better than about 90% as good as the PI result. So there's definitely a problem with my understanding of the tools and documentation etc. APP on the other hand is 99% as good with its semi automatic defaults which surprised me greatly and I use it often since I have that confidence in its algorithms. Confidence in the process is important for me. Knowing that what I see on my camera screen is very dark and full of noise but knowing that all goes in my workflow helps me be confident to let my setup take subs and that I'm not wasting time. Early on in my AP learning I was wrongly focused on what was in the camera especially with planetary videos being fuzzy and watery but then I was introduced to wavelets and dread to think of the videos I threw away that could have been of use.
I'm still waiting for that perfect software that can do everything astrophotography related regardless of subject, wide field, galaxy, nebula, comet, planet, lunar, solar, spectroscopy using different captures over time with different equipment. :)
Carlton
I highly recommend you start looking at capturing flats. Flats, darks and bias frames can be used often over various imaginf sessions and dont need to be taken each time usually. If you are taking apart the imaging train exposing imaging sensor and lens elements to the air they quickly pickup dust that shows up in your image. But apart from that all images have some degree of darkening in the corners as less photons reach those corners of the sensor so its advantageous to have flats to compensate. It makes a bigger improvement to end end result than darks or bias plus having a flat even brightness across the shot lets you build up better mosaics of frames to a wider image.
PixInsight is difficult to learn due to the terminology and that its basically a "desktop" with tons of tools in the menus, its unhelpful to know where to start. As Ivo mentioned its a "linear workflow", meaning you start with using one thing, then run something else, then something else. What these things are are entirely up to you, the software does not guide you along in anyway. PI can be run in demo mode and if you go to the light vortex tutotials online and follow their proposed PI workflow that will get you from your captured subs to a nice image. The are easy enough to follow, they try to explain the Why as well. I highly suggest you take a full set of lights darks flats and bias frames first. dozens of each at least then go through the LVA PI workflow, just follow it as it stands and see what you end up with and how it felt to do. PI is actually easy to use, the hardness is understanding what is going on with each tool. Others may tell you to only start withPI's BatchPreProcessing script which is basically a one click version of that workflow but it wont teach you anything and it uses technically less accurate settings to accomodate the hands off approach. But go through the LVA workflow a couple of times and see how it feels, maybe take it down as your own notes as you go so you can understand and repeat it with other data.
You just may not find PI is for you right now, thats fine, just experience it for yourself to find out. You may also have a look at AstroPixelProcessor, its both simple and powerful and I think is a good progression from DSS towards PI. Its approach is familar to DSS and PI and has good power and fits nicely between the two. But check your capture format is supported first.
I'm honestly not sure where ST fits in to the panetheon of AP Software, it seems its own thing to just clean up an integrated image. Its such a difference approach its not easy to really compare and maybe it cant. Analogy: you need to travel from Sydney to Melbourne. You could take car, train, ship or plane. They achieve the same end result but are meaningless to compare for this purpose alone. But the choice of holden or ford for the car has meaning. Here I see DSS as maybe ford, PI as holden and ST as maybe a ship... they all get you from A to B but so radically different a good comparision is not suitable.
Its too subjective. for MYSELF only I like APP and PI and find ST difficult to understand what I need to do. For others it will be different again. For yourself its likely you will differ from me but I think experiencing tthem yourself is the only way you'll make the right decision especially when understanding others advice on the products.
For myself, I regret buying AstroArt and Nebulosity. I don't regret buying StarTools. And I think AstroPixelProcessor and PixInsight were money well spent. I have one or two other commercial packages as well but they will be in my regret pile as I can't even remember them and certaintly dont have them currently installed.
irwjager
02-08-2019, 12:41 PM
A thousand times this! Best advice in this thread. They are really not optional, and you're making your life much harder than need be during post-processing.
If you want to make the most of PI's 45-day trial, at least give it (and yourself) a fair chance. Processing without flats will - likely - not be representative of what you can really achieve with PI (or ST for that matter!).
Easy. :) Use the software forever on as many PCs/Macs/Linux boxes as you like (provided you own those machines personally). You are entitled to updates 2 years from purchase (though this has not been enforced and any license ever issued will work with the latest version for now). No license server, no Internet connection required.
PI, AFAIK (and do correct me if I'm wrong), also entitles you to updates until 2.0 comes along, effectively coming down to the same thing (though a licensing server scheme is in place for that).
That thing, usually leading to the "a-ha moment", would be Tracking (quick video explainer (https://www.youtube.com/watch?v=i4wKPvjFtkU)). Perhaps the only problem with Tracking is that it requires a significant portion of the workflow be done in StarTools. Without Tracking engaged or allowed to do its thing, StarTools will perform no better than regular software and you may wonder what all the fuss is about.
That... is not ST's approach or raison d'être. I was hoping my previous post would have shed some light on what powers StarTools, but I appreciate it's all pretty technical.
ST employs the exact same algorithms you find in PixInsight with the same controls and parameters (unless obsolete due to Tracking). I was hoping the table I linked to explained the rough equivalents (as well as notable differences) in terms of PI terminology.
I would love to do a personalised workflow/tutorial with a dataset of your choice! At the very least you would do ST's development a great service, in case there are any areas of improvement.
It's a bit concerning in particular that you say the results you get from APP (which doesn't at the moment even have essential post-processing tools like deconvolution or noise reduction) are "99% as good". This leads me to believe something is going really wrong and I would love to understand what that is!
This leads me to believe you may have been using StarTools without Tracking (e.g. just for one or two modules)? Because this describes exactly what StarTools Tracking does; you can be confident that a screen full of noise coalesces into a clean image at the end when you turn Tracking off and final noise reduction is applied. Always.
Do let me know if I can be of any help - I'd love to understand what is going on!
irwjager
02-08-2019, 12:43 PM
You are in for a treat with 1.5!
Will send you a link to the latest 1.5 Release Candidate either today or tomorrow. Thank you!
Outcast
02-08-2019, 01:20 PM
Thanks to all who have contributed...
Particular thanks to Sil for some closer guidance on the importance of flats & bias shots & to Ivo regarding Startools itself.
Ivo, I have sent you a PM
Cheers
Disclaimer - I'm a Startools licence holder who has recently purchased a mono camera and filters. Processing my first few LRGB shots in ST I decided maybe it was time to move to PI as I didn't think ST was cutting it for the LRGB (was great for my OSC pics though!). My change to PI came around the same time as Ivo was working on a new version of ST and he offered me a try of it.
Initial impressions are below - in summary though I'll be shelving any further PI work as I think there have been some significant steps forward made on ST that have made an immediate difference to my images (same data processed on the new version of ST as I used for the previous version)
This review is coming from someone who thought he had squeezed all he could out of the program and was looking for something more. To be honest after a couple years using ST I had the impression it was a DSLR/OSC processing program....I'll admit now I was wrong with that impression. (well, at least looking at this new version which appears much more feature rich for LRGB and NB processing)
This is not a paid review, I don't know Ivo and I already have a licence for the program and so I have nothing to gain or lose by giving my thoughts on it. To be honest, if Ivo hadn't made any significant steps forward from my perspective I wouldn't have even bothered typing this post.
First impressions - it's a very similar setup as what I am use to however it seems he has cleaned it up a little....looks a little more 'slick'.
First up you can notice a reorder of module icons that it seems to me now follow a workflow. Previously I jumped from here to there between modules....flows a lot better now in my opinion.....making it even easier to use.
The LRGB module has now been moved to a 'compose' module and offers a significant step forward in functionality for LRGB processing from what was there previously.
I've noted a new preview button. Previously the was a before/after button that allowed you to see changes you made with a module....now there's a pre/post tweak button and well as the before/after. Great feature to maker sure what you are doing is making a positive difference. Great comparison tool.
Colour module now has some tweaks as well that are focused on narrowband processing.....I look forward to trialling the 'Hubble' preset with some data later tonight.
A number of the modules seem to have some enhancements and/or additional features. The small play I've had with these seem to indicate they are not there just for show, they are useful tools that I guess are based on feedback Ivo has received or through thinking through new ideas.
Entropy module - I'll need o have a bit more of a play with this one however initial impressions are good. Pulls out some great detail.
Noise reduction (tracking module) - as always, this is a killer and it always blows me away the amount of unwanted noise it easily sweeps from my pics
All in all it seems Ivo has put some significant work into improving Startools and from what I have seen they are all positive changes. He should be congratulated for giving the community and cost effective and very powerful tool.
Thanks Ivo....this new version has definitely put a stop to me moving over to PI. From what I have seen this new version offers the same simplicity in processing that to me just makes sense! Great work mate.
irwjager
05-08-2019, 11:50 AM
Greatly appreciate the review Jon! As stated earlier, if anyone else wishes to have a go at the 1.5 preview. Do let me know.
The Compose module is indeed a big one; besides doing synthetic luminance creation for you, it allows you to process LRGB, LLRGB, LSHO, LLSHO, etc. as if it were simple DSLR/OSC (RGB) data.
It does so by processing 2 datasets in parallel; e.g. separating luminance and color completely, yet processing them simultaneously. No need to process separately and combine later through two separate workflows (as you would in PI). It also allows the Tracking engine to cross-consult the two datasets to enhance fidelity and detail.
The new Entropy module is a direct consequence of now having both datasets available separately, but simultaneously. Color is consulted to gain clues about detail prevalence in the luminance dataset.
Cheers!
Thanks for the comments...something going very wrong would likely be me. Deconv and noise reduction are parts of my workflow but not highly essential (or at least I'm not using them properly). If you are familiar with the PI workflow suggestion tutorials on Light Vortex Astronomy thats a little simpler than what I use. I incoprporate steps to remove stars temporarily to stretch nebulosity. No I may have not used your tracking feature or possibly its not in the version I have (sorry, it could be ancient)
Just to put into perspective, I mostly do DSLR capture with a lens on a regular tripod. So I have for a session/target about 500-600 Lights, Maybe 50+ Darks, Flats, Bias (in any combination, depends how my body is feeling at the end of my session I may take some, all or none knowing I have the Lights and can work well from those alone if needed). I also periodically take fulls sets of Darks, Flats and Bias frames as my camera is permanently set up because I cant use it as a camera anymore so the lens stays on always so sensor dust is rare for me. Using so many subs (and I rarely have ones to throw away) noise is lowered well upon integration.
My workflow uses registration (aligning) of frames, calibration/debayering of frames etc, processing with bias/flats/darks, and then integration of everything into a single 64bit FITS file. This is my starting point for processing (what people call post-processing?) to take this file which is still in linear state and stretch the signal to get to my pretty picture to show etc. This gives me the option at any time to reprocess from my starting point Integration frame. Ah one last thing I try to do first to this file is plate solve and save it with metadata and render an annotated jpeg to refer to easily when I want to work with combining multiple sessions integration frames to enhance a target or create a mosaic.
This is all one shot colour untracked and a huge amount of data for software to chug through. Something I like about APP is I can throw a ton of data at it with any permutation of darks, flats, bias and it will make use where it can of all that data. So for me its flexible in that I dont have to make sure i take certain calibration frames. Plus it can mosaic all the lights without, so on nights where I chase a comet for example, its rare that I am ever able to see the comet by any means and pointing my camera at the correct bit of sky is hit and miss, so a mosaic straight off of subs throught my capture session gives me a result I can plate solve and find out if my target is even in the data or in every sub or maybe just at an edge so I can decide how to proceed with processing just the frames I want.
Longer term goal for me is to try to do a whole sky image with all these tiny patches.
Lunar and solar are sometimes with same dslr setup. But planets and moon I do with a 6in sct with dodgy tracking and a zwo cam. So that data is captured as SER video, debayered/centered/cropped in PIPP and exported to BMPs to run through AS!3. Again this is a lot of frames and noise is ironed out and autostakerts wavelets sharpen the images well enough, anything further in other programs really introduces sharpening artifacts which I dont like. I also prefer the colours as they are: natural and muted.
I dont recall if ST I can just throw all my data (such as I have) at it and go from there. APP does take it all in and saves a linear FITS file plus it does and automatic stretch/post process for screen display which you can save. Its this stretched version thats very close to what I can do with PI just without the time and effort. Plus it has things to tweak I havent tried yet. I never got into DSS which could take my data to a master integration frame easily but stretching was difficult to control for me. Having good knowledge of PI by the time APP came out it was clear to me it would sit well for DSS users wanting the next step up but not able to afford PI etc. Though APP only supports some manufacturers RAW formats it doen't yet cover all common ones, its slowly being improved as the developer has time etc.
I would be happy to share a set of subs with you Ivo, message me if it'll help you out, I'll just have to put together an external hard drive to copy files onto. I think I still have my comparison of PI and APP somewhere but I had drive crashes in the past year and had to delete a ton of astro data I was working with or finished (my originals are safe but only organised by date) I dont recall if I posted to this forum (I may have) a part comparison. I did it purely for my own interest after buying APP. Its ease made it seem quicker (which i think it is but not vastly slow, never timed it) and now Its a simple way for me to get a result as I tend to just shoot whatever when I have a chance, rarely after a target and m31 the only target i want more of is never visible from my home. As I said its easy to mosaic my subs which move since im on tripod i rarely recompose the frame so i can see if i even got a particular target and then put the time in PI to process carefully. So I don't know if ST has a place for me currently. Also to add perspective I suffered a stroke 5yr ago which paralysed the left half of my body and i've faught to regain some leg use to be able to stand and walk again my arm is unresposive and body twisted so how I used to do astronomy 5 yrs ago has evolved differently with chances in my interests and physical/mental capabilities. I think I bought ST early on and really liked it as a package and your philosophy for it and I wanted to support it. And yes it has yet to grab me in a big way, some of its capabilities I have really liked but they couldn't be used in conjection with PI to get what I needed done at a step. So I couldn't take my working fits into ST for one single tool to make a particular adjustment and save back to same fits format and load back into PI. As its not an "everyday" tool for me I tend to forget to retry it, update it and find current tutorials etc. My life has so little free time to experiment and learn but it also changes with more restrictions as my body declines.
So my methods in astrophotography are not ideal for anybody only to my own limitations, time for learning is itself a hinderance in my life and a personal failing thats hard to overcome. I will try to give it another chance soon,I wrote it on my hand, APP is at home still chugging on 500 frames i took on friday. its the only working computer i currently have so i wont muck about until it finishes or crashes.
Steve
irwjager
05-08-2019, 05:41 PM
Thank you Steve. It indeed sounds like pre-processing is the killer app for you (which ST obviously doesn't do). ST only takes pre-processed stacks. ST will be a lot less relevant if you have tons of signal, do very wide fields (let alone mosaics!), and are happy with a simple stretch for the post-processing part of things.
ST is more for those who can't wait for the calibration, aligning and stacking to be over and "get to work" on a faint object (e.g. me! :lol:). Give me the challenges; light pollution, detail recovery, low surface brightness. Going beyond diffraction, atmospheric or personal circumstantial limits (with regards to the latter, let me just say I have nothing but respect for how you are successfully dealing with yours!).
Personally I get my kicks chasing the stuff the Hubble Space Telescope has showed us over the years to the best of our "mere mortal" abilities and gear, accurately documenting and bringing out the features that show the past, present and future of these fascinating, far away objects.
At any rate, if there is a use case for StarTools in your workflow, do let me know, now or in the future.
Clear skies!
xelasnave
07-08-2019, 04:02 PM
Any chance of getting in on that link?
Alex
rcheshire
10-08-2019, 11:32 AM
1.5... Oh! Goody!
DuaneDibbley
10-08-2019, 02:34 PM
I'm a beginner who went straight to PI when I bought my telescope.
It's a supremely capable pig of an application that's horrifically poorly documented for how much it costs. Nearly $400AU and maybe 6 processes are documented? Extremely disappointing.
No doubting the output though once you've managed to trawl the internet and torn out all your hair looking for tutorials all based on different versions of processes. But like others have said in here after a year (or more) of using it you'll still be a beginner.
That said - I simply couldn't bring myself to use DSS and Photoshop and MUCH prefer PI. I just hate it while I'm using it.
I'd never heard of StarTools before. I'll definitely be having a look now though!
All that said though - I adore firing up PI and seeing is strangle my Ryzen 2700x to within an inch of its life. No other app I own happily eats up all 8 cores and 16 threads on offer :)
Startrek
10-08-2019, 03:16 PM
I initially started with Startools in AP and can’t see myself using anything else in the near future
After 12months on version 1.4 I’m still discovering different processing methods to improve my final images
I always end up with a reasonable result irrespective of the quality and quantity of the data , and I use a 10 year old DSLR with my 6” f6 and 8” f5 newts
Ivo Jager thanks again !!!
mynameiscd
11-08-2019, 09:05 AM
Hi Ivo,
A registered user of ST here but still learning.
I'm progressing slowly but still restricted by my XP machine I do processing on. I've yet to use PI so can't comment on that but have seen dramatic improvement from changing from PS to ST for processing.
Definitely need a Win 10 machine as my 20 year old laptop doesnt cope with the high demands of processing but ST works without too many crashes after binning my images early on.
I had a play with the android vs of ST and although not finished and problems with screen size i can see a future in this direction but for a simpler processing chain for the phone photographers.
You can never duplicate the the raw power of a pc compared to a phone or tablet but us humans are going down this path whether we like it or not and esp the younger generation just want something to work but dont do any work to make it happen!
I know this is a whole new debate but clip a phone onto a dob and press one button to start a series of lights, darks, and flats in one then a second button to process and the result is a reasonable image that would have taken 10 hours of hard work to do 20 years ago.
Looking at a lot of the names here who have contributed to this post i can see many who have helped me get over the hurdles of processing and without encouragement i may have given up with mediocre images.
The main thing i have learnt is that the better data you get, the less processing you need. I used to do a lot of recording bands in my home studio and a few other studios and it all goes down to recording the best sounds and performance you can get rather than the "fix it up with editing" type of recording with shifting individual drum hits and autotune on everything.
We are all so lucky with the imaging tools we have got now and what is coming in the future but they are still tools and the old saying " you cant' polish a turd" is still valid, (yet saying that I've polished a few with processing and they're not too bad).
Thanks again Ivo for developing this sothware and the time involved with coding, testing, and bug fixing.
Also thanks again to all here who have helped so much.
Cheers
Andy
JimsShed
11-08-2019, 11:14 AM
My opinion of PI is that you can only get some sort of proficiency if you also buy an aftermarket text book on the product. The usability of the program itself does not always conform to the norms of what we expect a desktop program to do, and it has enough counter intuitive behaviours to be annoying to use. Eg; window resizing and magic drag/drop points. You eventually realise the overall product is an eclectic mix of dozens of individual processes, most of which you will never use or you simply avoid. Also for the most part, the help system is non-existent and you find yourself referring back to a text book.
Outcast
11-08-2019, 12:34 PM
Gentlemen,
As the OP I wish to say thankyou for the mix of very frank & honest relation of experiences with each of these products.
Ivo has kindly offered to assist me with gaining proficiency with Startools, an offer I am defintely going to accept....
The information contained in this thread has helped me enormously to understand the differences in the two programs & consider the pro's & cons of both...
I'm hopeful it has helped a few others out too..
Cheers
Ivo, I had a play with ST on the weekend briefly. Yes the reason ST never became a part of my AP is it doesn't take my raw captures and gets me through to a finished image. To me its essential like one of the hundreds of "photo filter" programs to beautify an image so you get likes on facebook or wherever. I lost interest in that superfical garbage in the '90s with photoshop filters etc which are now back it seems with mobile apps and social media... none of which interest me. I have deleted where possible the few social media accounts I ever had.
My goal with my AP is to push my capabilities and that of my equipment as best I can to get clear "true" images. If there's noise then there's noise. I dont give a damn about Image of the Day recognition or stuff like that. I already know my images are not the best around, nor can they be with my limitations, and I dont need trolls to point out the bleeding obvious to me. I'm satisfied my images are the best I was able to do at the time. It also gives me benchmarks to beat with newer processing tools and knowledge using the same data sets.
PI and APP don't do everything I would want so for me there is no perfect program available. ST does some very nice things and maybe it has a place for some images where I want to really try to make them pop but for now I'm satisfied with what I'm getting, though I would like to astromod my camera if i get a tax return this year :)
I disagree, if you READ what the product is about and its documentation it clearly tells you what it is and what its not. You should never "eventually realise" because its always there in black and white, just you didnt bother to read. Besides there is there "unofficial" missing PI manual website to help with expanded documentation (yes the supplied docs are poor to non-existent for various tools). The forums though thoroughly explain new tools and scripts in superb detail, if you read them. And the Light Vortex Astronomy Tutorials site is itself very detailed and thorough.
PixInsights steep learning curve is because its NOT a dumbed down program to hand hold people unwilling to learn. Astrophotography is all about photon energy and mathematics, a LOT of complex mathematics so you need some skill to use it. Its the program for people willing and able to build a car from a box of bits instead of just drive one already built. Its not a program for everyone.
Guess we all have opinions Steve.
Seems you are pretty insistent your view is correct with all the SHOUTING you do with those capital letters.
I like ST, you like PI.....others like other programs......but there's no need to SHOUT! :rofl::rofl::rofl:
RussellH
26-08-2019, 02:25 PM
Throwing my 2c worth in here. I’ve just purchased StarTools after spending the day with it yesterday processing my first image, so take my opinion for what it is.
First off, StarTools is from an Australian developer. Being a former software developer myself, I know how hard that is, so deserves support just for that alone.
Secondly, when PixInsight is 6x the price of StarTools, I have to ask the question if it’ll give me 6x better results. StarTools while being different to other image processing software seems to be fairly competent at being able to process images with a fair degree of flexibility and power. I can’t see why it won’t be able to meet my needs for a long time, and ai doubt very much that PixInsight will give me a 6x improvement in result any time soon.
So for now, StarTools is it for me, even with the option of PixInsight on the table. Even if PI was closer in price, I think I’d still choose StarTools for now, as it seems to offer a unique approach to image processing which brings something new to the table than just a supercharged Photoshop workflow. Time will tell if I’ve made the right choice.
Russell.
Startrek
26-08-2019, 03:29 PM
Russell
As a fellow Newt / DSLR astrophotographer and avid user of Startools, you have made the right choice to kick start your Astro processing
Learn and enjoy it’s features
multiweb
26-08-2019, 03:36 PM
I use them both equally. PI is great for drizzle integration and OSC processing. ST is very useful for the long FL stuff, sharpening, deconvolution and noise management in oversampled data.
xelasnave
26-08-2019, 04:38 PM
Imagine a fully automatic astrophoto imaging set up...all you do is say "computer photos of M16" and it opens the dome moves the scope to the object, determines the exposure, number of subs, filters etc, downloads stacks and processes the photo and posts them for you... Nah.
Alex
vBulletin® v3.8.7, Copyright ©2000-2025, vBulletin Solutions, Inc.