Is PixInsight an application in terminal decline???

[rodmichael]

I am a new user and a new purchaser of PI.  I am also a subscriber to IP4AP and I have been dutifully going through Warren Keller's tutorials.  I bought into the idea of PI because of it's touted specificity for AP as opposed to the general use nature of PS and the relative economy of PI compared to PS.

However, I have become concerned that I may have bought a dying horse supported mainly be the energies of a single developer and a handful of passionate believers.  Things that stimulate my concern:
1. The place where I buy my equipment is populated entirely by 10 active imagers and PS users.  For them PI is much more difficult to use than PS so they don't.  I see a website and webpages where copyrights are several years old.
2. I see a users' forum that is not very active.
3. I see development that is entirely dependent on the single owner/developer and the sometimes contributions of a "community" of wannabe developers who develop, for example, an acquisition/preprocessing module that requires Windows users to run on a "virtual machine" and has apparently died a few years ago.
4. I see the principal developer making a statement that a package update will not be forthcoming for one or more years yet primarily because he's all alone.

Is my concern about the health of PI unwarranted?

[Duncan]

Provocative.... 

I'll share my thoughts here as a newbie to PI also. I'm sure there will be many others chiming in.

As background, I've been imaging for ~13 years. That makes me relatively inexperienced compared to many who post here, but I know my way around imaging and processing pretty well by now. I acquired PixInsight maybe a little over a year ago, but it has taken me some time to transition over from my previous tools. I have previously used many different software packages, most recently Maxim which I have been using pretty much exclusively as my front end processing for several years. That has always required some backend work with Photoshop also.

In response the question posed in the thread title, my short answer:

No.

Longer more detailed version:

PI for sure has a learning curve that is steep and arduous. I also have several experienced imager friends who have tried to make the transition over to PI but found it difficult (this is something of an active discussion among us right now). It took me quite some time to get to the point where I was comfortable using PI to the extent that I was getting better results than with my previous tools.

However, having got to that point it is very clear that there is much more capability in PI than anything else I've used previously by an order of magnitude. The results I'm getting now are far superior to anything I could achieve before; there's a huge array of advanced processing tools that you just won't find anything comparable to anywhere else. I should note, for the first time I don't have Photoshop as a back-end "fix" for stuff I couldn't quite get right otherwise.

As to the pace of development, I guess my perception is a little different. There's no reason to expect particularly rapid development of new releases of the platform itself, since it's the framework for the tools. I'm not sure what would need adding to it right now... The tools and scripts that run on it are a different matter, and I'm seeing a fair number of auto-updates whenever I boot up the system. Compared to the cadence of updates/new releases that I saw with my previous software of choice (Maxim) I'd say PI is more actively developed by a substantial margin.

Not sure what the acquisition/preprocessing module you're referring to is. I haven't seen anything that needs running in a VM on Windows (I'm 100% Windows myself). However, if you get stuck on something specific, there do seem to be people here who will respond.

Anyway, like I said, these are just my own thoughts as a pretty new PI user still climbing the learning curve. I'm sure other people will have their own thoughts...

[rodmichael]

Thank you for a very thoughtful response.  I was afraid (especially after rereading my original post) that I might have come across somewhat abrasively, which I had not intended.  I was just kind of bummed out by this thread http://pixinsight.com/forum/index.php?topic=9852.0.

I am a total newbie to everything except observing.  I'm 70+ and just trying to get a start in AP.  Figuring out the software end of things is perhaps the most intimidating component of this hobby, especially the post-processing.  When I look at the PI forums, I am intimidated by the jargon and the very technical nature of the discussions and have wondered if it is the right place to start to learn post-processing.  I have a difficult time deciding what would be more difficult, learning from scratch as I am trying to do or learning another platform, e.g., PS and then relearning post-processing anew with PI as you apparently have done, with 12 years experience on PS or whatever.

So thanks for the nice response to an angry post.

[NGC7789]

1. The place where I buy my equipment is populated entirely by 10 active imagers and PS users.  For them PI is much more difficult to use than PS so they don't.  I see a website and webpages where copyrights are several years old.

Tools you know will always appear easier than tools you don't. Unfortunately it's hard to evaluate and compare tools until you traverse the learning curve of both. However, you can compare and contrast the guiding principles of PS vs PI. PS is primarily a composition tool. It is focused on subjective alteration to suit the user's desired outcome. It serves MANY kinds of images (camera generated, paintings, etc.) PI is an data analysis tool primarily for astrophotography. It is focused on revealing information existing in the image to suit the user's purpose. If that results in a pleasing image that is a side benefit. Only you can say which one suits your interests. In my opinion it not a question of which is better. Both are arguably best in class. It's a question of which class is your class.

2. I see a users' forum that is not very active.

This is just false. There are dozens of posts on a variety of topics every day, including this one.

3. I see development that is entirely dependent on the single owner/developer and the sometimes contributions of a "community" of wannabe developers who develop, for example, an acquisition/preprocessing module that requires Windows users to run on a "virtual machine" and has apparently died a few years ago.

It is true that PI is primarily the work of one owner/developer and I, like probably many others, sometimes wonder what will happen on that grim day hopefully far in the future when Juan is no longer with us. But there are no guarantees in life. Adobe is not guaranteed eternal either. Regardless of what happens to the maker, the software will continue to do what it does. I recommend evaluating the software for what it is now, which is a lot. Referring to the other contributors as wannabe's is simply insulting and shows your ignorance. And the acquisition/preprocessing module was only introduced in March of 2016. How could is have died a few years ago?

4. I see the principal developer making a statement that a package update will not be forthcoming for one or more years yet primarily because he's all alone.

PI has an unusual update paradigm.What PI considers a .x release is often equivalent to a major update (or two...or three) in other applications. And between those are many individual tool updates and introductions. In fact the acquisition module you referred to is one of those. Juan's reminders that he is alone is to temper people's expectations even though he always exceeds mine.

If you want to use PS then do so but the investment in learning PI is almost certain to reward you.

[rodmichael]

3. ... for example, an acquisition/preprocessing module that requires Windows users to run on a "virtual machine" and has apparently died a few years ago.

... And the acquisition/preprocessing module was only introduced in March of 2016. How could is have died a few years ago?

If you want to use PS then do so but the investment in learning PI is almost certain to reward you.

[/quote]

....I was just kind of bummed out by this thread http://pixinsight.com/forum/index.php?topic=9852.0....

The thread above which appeared to be the rollout of an acquisition/preprocessing module written by a PI user in 2014 is that to which I was referring.

I'm not really interested in using PS over PI.  As I noted just above, I chose PI based on what I know which, admittedly, doesn't amount to a hill of beans at this point.  My question seriously is will PI be there when in 1 or 3 or 5 or in whatever number of years I have spent learning how to use it.  PS has been around for 25 or 30 years and seems not to be diminishing.  It has a whole corporation behind it.  I guess it looks more secure to me.  But it doesn't have the advantage of having been written specifically for AP by astrophotographers.  That's primarily what attracted me.  It has a budgetary advantage, but when one has spent North of $20K to get started in AP, a software package costing a few hundred $$$ more is not a big deal.

I appreciate the frankness of both responders to to my initial post.  I hope that I can come to use, appreciate, and enjoy PI and its capabilities as much as they.

[msmythers]

As just a note it might help to know that PI has been around for a long time now. This forum is the best place to find out about PI, in my opinion, anything about PI.


Mike

[Duncan]

....I was just kind of bummed out by this thread http://pixinsight.com/forum/index.php?topic=9852.0....

The thread above which appeared to be the rollout of an acquisition/preprocessing module written by a PI user in 2014 is that to which I was referring.

Understood; there's probably several things worth pointing out here, in no particular order:

PixInsight doesn't have any native capabilities for image acquisition and/or imaging system control. That works fine for me; there are plenty of other systems that do that, and trying to add that sort of capability to PI would be a huge defocusing IMHO. Unfortunately, that sort of hardware interfacing is something that *does* really require a large team of people working on it constantly just keeping up with all of the many different cameras/mounts/guiders/focusers/domes etc etc and the ever changing morass of drivers and operating systems. That sort of thing tends to rapidly turn in to a maintenance & validation nightmare. Just my 2c of course... others may reasonably differ on this.

On the other hand, as AP processing systems go, PI has a very open set of interfaces which allow third parties to write their own modules. There are other systems that have scripting interfaces (and I've tried my hand at writing plugins for Maxim in the past, for example) but they're in general nowhere near as extensive as PI in terms of what you can usefully access with them. So if somebody wants to write their own modules to do system control, they can do that. I'm *definitely* not the best person to comment on this, as a PI newbie, but I'm really not expecting that ever to be a mainstream PI feature.

Additionally, though, you mentioned preprocessing in addition to acquisition. Preprocessing is certainly within the domain of PixInsight. Best case, the data you bring in to PI should be as near as possible raw bits straight of the analog-to-digital converter in the camera. This is a general statement about any AP image processing software; you want raw data that has been "messed with" as little as possible, and messed-with-not-at-all is optimum.

[Juan Conejero]

Rumors about our death have been greatly exaggerated. The future of PixInsight is guaranteed even when I can no longer work on it. This is because PixInsight is much more than an application, it is an open development platform and a project much more ambitious and richer than a simple computer program. When I am gone, or if the project is no longer viable for me for any reason, what will happen with PixInsight is already decided and planned so that the project will last in the way that it can be most useful for the majority. So PixInsight is here to stay and evolve, there's no reason to worry about this. In the future there will be other interlocutors, but they will certainly be more communicative and easier to deal with than I am.

By focusing solely on this forum, it may appear that the development of PixInsight has stopped or been slowed down. Nothing is further from reality. If you take a look at our official open source repository you'll see that our code base is under intense and permanent development. As of writing this, the last update to our open source development framework (PCL) took place yesterday, when I updated our XML parsing and generation tools. At present I am working on the final XISF specification document, which I hope to publish very soon, as well as on the XISF reference C++ implementation, so that the new format can be adopted easily by other applications and development teams. After completing these tasks, my intention is to release the next version 1.8.5 of PixInsight in February, and later there will be new versions of DBE (which will be DB, DynamicBackground), LinearFit, PixelMath and StarAlignment, among other essential tools, as well as new tools. The planned deadlines may be delayed somewhat, but all ongoing projects will be completed. As you can see, my agenda in the short and medium term is pretty full.

As for the eternal PS/PI question... it's a boring question, I must say. The answer is very simple: From the end-user perspective, PS is basically a tool to paint and retouch images, while PI is basically a tool to design and implement image processing procedures. PS does not require or facilitate your understanding, neither about image processing nor about the specific problems that must be solved to process a particular image, while PI does. The decision is up to the user and also simple: if you want to paint, then use PS; if you want to process, then use PI. Of course, there are suboptimal ways to process with PS and suboptimal ways to paint with PI, but for a serious user of any of these applications, the intersection of both sets is, essentially, empty. That's it.

[chris.bailey]

Ouch!

As one of your "handful of passionate believers" I feel duty bound to respond.

I think you will find that all "astro" software development teams are pretty small. Software Bisques SkyX core development team is in the low single digits and I suspect the other major players are similar. With many others it will be part time development at that. Its the nature of the beast and the market it sits in.

As users we want new "goodies" and we want them now. That is inherent to human nature. If you compare Pi now to Pi 8 years ago, the development is actually pretty immense. As an aside in that 8 years I have not paid a single upgrade fee for PI, despite being very willing to do so. Compare that to the current subscription model for PS. PI is incredibly good value.

As a moderator on a UK imaging forum I have "enlightened" several people over to PI. Others seem to cling on to using PS, Gimp and the like for grim death (these are often the sort of people who like to add artificial star spikes). I also hear from people who complain that PI is expensive, despite having just bought a full set of 2" Astrodon narrowband filters. If PS works for them, fine by me, but I often see very convoluted processes for doing things like narrowband combinations that make my head hurt, yet PI will do it with a single process or two, repeatedly. My other experience of being a forum moderator is that "lurkers" outweigh "posters" by a very large margin. If you judge forum activity purely by post number then it may seem quite sedentary but I would say the post "quality" to be high and an essential part of getting to grips with it.

Just my $0.02 (which is a lot more compared to £ than it was a while ago)

Chris

[IanL]

As one of the "wannabe developers" I thought I'd chip in here (degree in computer science, 30 years in IT including as a professional software developer and currently running a team of 40 developers notwithstanding).

PI offers a generous free trial period, so it is slightly puzzling that you are only having these doubts after you've bought it

Yes PI is harder to learn than Photoshop on the face of it, but my experience is that accomplishing the steps needed to create a good astro image are not intuitive in Photoshop and by and large one has to learn a lot of hacky techniques from tutorials. PI's tools do the actual things needed rather than a half-baked approximation of them.

In any event you have to use something to calibrate and stack your images in the first place, and every other tool available is inferior to PI in that respect. So if you've bought the best for that task, why pay the annual Adobe tax on top instead of learning the remaining tools?

To denigrate the community (not "community" as you so insultingly put it) is to miss the point. People contribute new scripts and processes because PI offers them the capability to solve their own problems, and having done so they are happy to share the results with others. Not every development makes it to fruition but many of the contributions are used daily by large numbers of people.

If that seems a bit harsh, then I suggest you revise your style of criticism to be somewhat more constructive.

[jkmorse]

For those of us who happily abandoned PS and its clunky interface that had and has no real relation to astroimage processing (which is why you need several other programs to do preprocessing before you can even use PS on an astroimage), PI has been, and I expect will continue to be, a happy home with dedicated users and a fabulous development team that extends beyond Juan himself, as stated in the his own post.  I actually find the forum amazingly active and helpful, though it is not a generic bull session which is why you see posts centered on PI and not the hobby generally.  There are tons of those if that is what you are looking for (and why I avoid those like the plague).  What I want and get here is an active group of extremely knowledgeable and helpful users who talk shop about how to get the best out of PI.   

As far as easy or hard, anything worthwhile has a learning curve.  But having spent so much on equipment to collect your data, don't you want to use the best tool by far to process it? And to help with the learning. Waren's book is a great place to start.  Harry's astroshed online has great material as well and I offer my own workbook for free that I share with around 200 users around the world.  If interested in a copy, just drop me a line at jkmorse57@gmail.com and I will be happy to send it along to you.

Best,

Jim   

[akulapanam]

If anything PixInsight is now the standard. How many astroart, maximdl, photoshop, or imagesplus processed images do you see on Astrobin, cloudynights, or stargazers lounge anymore? Not many. I do wish that PI development was more active but I suspect big things are happening on the backend.

[mmirot]

If anything PixInsight is now the standard. How many astroart, maximdl, photoshop, or imagesplus processed images do you see on Astrobin, cloudynights, or stargazers lounge anymore? Not many. I do wish that PI development was more active but I suspect big things are happening on the backend.

I guess the newbee poster has not been around enough to know PI is number 1.  We just had major update to several modules two week ago.

Btw, most people are using PI for image processing only. The acquisition stuff is more of a beta. Not a concern for most people.
SGP is excellent choice at 99 dollas US IMO.