New to Pixinsight / What to do I need to do and know

Hey guys,

I couldn't resist any longer and bought myself a pix insight license.
Until now I used APP for Stacking and affinity photo for further processing.

I already processed my first image with pixinsight, and I need to say, that this program
is like nothing I used before.
The amount of processes and possibilities is big.
That big, that I have no clue what to do.

What I would like to know from you guys:

What is your pixinsight workflow?
Is there still a reason to use processes for denoising and image sharpening in pixinsight or will I be fine just using NoiseXTerminator?
What are the best places to learn about image processing with Pixinsight? Youtube, Blogs, Astrobin etc.

Thx and cs
Andi

I would recommend watching videos a to learn. I can almost guarantee if you dont youll get frustrated as hell. Its a great piece of software and once you grasped the basics everything else will fall into place. Adam Block has some great starter videos, however there are loads on you tube. Everyone has there own workflows, nothing wrong with asking others flows. Have fun learning it always amazes me when image pops, bear in mind also youll end up bining alot of images.

PixInsight has a Youtube channel, https://www.youtube.com/pixinsight. They are currently in the process of releasing new videos. This is theannouncement from the PixInsight Forum, which has the complete list of the videos they plan to upload. By the way, the PI Forum is also a great resource, with members who will happily look into your data for issues.

My advice is to start with the official PI channel (and Adam Block's) for now, and try to fully process a dataset (from raw data to a final jpeg for Astrobin

) within PI, using only the necessary steps.

While I am not an APP user, I believe that after a while you may find APP falls short in features (probably except its capability to process mosaics). Regarding Affinity, it should be a great piece of software for the money, and even if in theory you can do everything in PI, it is nice to have it handy for the final touches to your images. Anyway, it is great to have a lot of tools at your disposal, but sometimes when you are frustrated learning a new tool, it is tempting to switch to the software you are more comfortable with (like abandoning doing physical exercise for the nearest available plate of pasta

). This will result to messy folders with intermediate and temporary files used for data exchange between software tools, and a partial knowledge of the new tool which you are trying to learn.

A lot of people will probably criticise me for recommending the following, however I found using the Weighted Batch Pre-Processsing script (WBPP) the easiest way to make the transition to Pixinsight from a stacking/integration point of view.

The Light Vortex site recommended by Andrea is great. That’s where I started. Step by step, menu by menu. Once you get familiar with the software, there are any number of other sites that will go deeper into specific areas.

Read , watch videos, and try, try, try.

I started with PI almost two years ago and had a rough start.
Like Andy said, WBPP has been refined so well, I personally see no reason to use a collection of single processes to pre process data. However, you should learn all of the single processes that make up WBPP so you can get a good idea how each works and how to troubleshoot WBPP when you have issues. However, the default settings work really well.
I used (and still use) every written and video learning tool out there. I’ve read and are constantly re-reading books (my personal favorite is “Inside Pixinsight” by Warren Keller). Warren is a member here on AB.
I’m sure that I've read it 5 times and each time I make notes on things. The more you learn, the more you learn. I’ve probably learned just as much by running into trouble and troubleshooting and figuring it out.
I don’t think I’ll ever master PI. At 67 years old, I decided to purchase it as a lifetime learning experience.
After about six months in, I learned enough to blow away any processing methods I’ve used before.
P. S. Adam Block’s courses (paid) are long, but phenomenal! I’ve also learned tons from his free videos on YouTube.
good luck enjoy!

Jim

Yeah, Youtube and get your hands on the keyboard and try different workflows. Get a basic workflow down first, one you can be consistent with, and don't be afraid to try different techniques and play around.

Also, the Adam Block fundamental training series was excellent at filling in details I couldn't find anywhere else. He has a quick start and a fundamentals course, but you have to pay for them. Highly recommended, though.

And I like WBPP, too!

I love this hobby!

Hello Andi,
Congrats for moving to PI! Like you I started with APP and in all honesty still use it for my processing over WBPP. Adam Block does an outstanding job with his tutorials. I think that the fast track is a great place to start. Best part is if you like that series, Adam applys the price paid to the fundamentals (which is what I bought) if you decide to continue the learning process his way. Here is the link: https://adamblockstudios.com/
In my case I downloaded all the video’s (over 300) so that I had the library and go thru them as time allows. I should also mention that Adam also has his own YouTube channel that he posts video’s for the AP community as well. So check them out too. Some are actually included with his Fundamentals subscription.

If you prefer the “free YouTube” option there is a good YouTube video that is for Beginners I highly recommend by “Mitch” for me this was probably the best set of videos that helps you learn how to get around the environment and setup process icons save them, make settings adjustments etc… its a 12 part series. https://www.youtube.com/watch?v=HIXJJqew6rQ

Pix Insight’s forum has many tutorials as well: https://pixinsight.com/forum/index.php?forums/tutorials-and-processing-examples.10/

Also there are many other YouTubers that do other workflows etc that you can learn from as well. Entering into Space, Visible Dark, lukomatico are just a few examples. Many of these guys cover the basic or standard workflow that you would use to post process your image.

8-10 months ago I’d have never said that I’ll give PI a try but the fact is its the best post processing software on the market and I’m glad that I made the transition to it! I think APP is a good stepping stone to PI.

Hope this helps!

Dale

My 2c.

The most efficient way to get up to speed is Adam Block's fundamentals videos. Yes, not free, but he keeps them reasonably current and there is a LOT of detail there. Personally I prefer the written word, but these have enough information to make it worth the slow pace of video in general (not saying his are slow, saying all video is slow compared to something you can skim to find stuff).

The lightvortex site helped me get started, but it is worth a bit of caution as, unless there are recent updates, they are pretty old and PI is a rapidly changing process. There is also some bad advice in there for local normalization (as well as being out of date).

I also liked Rogelio Bernal Andreo's Mastering Pixinsight, with the caveat that like all books they get out of date quickly as well. The best part about that book (at least for when I read it) was that it explained more about the paradigm of the PI user interface - what all the circles, squares, triangles etc are for. So very much of the documentation presumes you just know (by some genetic trait I guess) what all that is, and it is NOT the same as any other UI anyone will have seen.

I use WBPP, I love it, but it is vital (I think) that anyone getting started in PI go manually through that whole process many times, looking at intermediate results, exploring options (a lot are detailed by Adam in the videos). That it does a good default process is a testament to the effort that went into it recently, but you will never be really good at PI if you do not understand all the things done underneath the covers, since you may find your issues there (a lot of people getting started have poor calibration, and if you go ahead and stack without checking that, it's a garbage-in-garbage-out situation).

There are MANY outside documentation resources, of various quality. But if you are just getting started, I recommend Block's videos as the most efficient way.

Linwood Ferguson:
There is also some bad advice in there for local normalization (as well as being out of date).

I don't see how what is in there can constitute "bad advice". Am I missing something?

andrea tasselli:
Linwood Ferguson:
There is also some bad advice in there for local normalization (as well as being out of date).

I don't see how what is in there can constitute "bad advice". Am I missing something?

Calibrating the individual darks is one such bad advice. While in theory there should be no problem in doing so (setting aside the unnecessary computational and storage cost), in practice (in PI) calibrating your darks may lead in clipped master darks.

Die Launische Diva:
Calibrating the individual darks is one such bad advice. While in theory there should be no problem in doing so (setting aside the unnecessary computational and storage cost), in practice (in PI) calibrating your darks may lead in clipped master darks.

I was referring to the part about normalization, not dark calibration. But in the specific I still don't see it as "bad advice". I do it routiinely for my DSLR and CCD work (but not CMOS) and it most definitely does not lead to clipped master darks. I'll stick my neck out and add that without it you'll end up with poorly calibrated light frames, if you do remove the bias signal from said light frames as you should for CCDs (and DSLRs).

There's an astrophotographer named Mitch who posted a 12 part series of You Tube videos called PixnSight for the absolute beginner. They are about 4 years old now but still relevant and very detailed. Lots of updates since then but these should get you on your way. Adam Block's videos and tutorials are excellent and well worth the price though he gets a little more technical. Adam updates his tutorials to keep up with the changes and upgrades to PI as well. I started with Mitch and bought the Adam Block tutorials and that progression has worked well for me.

Good luck and clear skies!!

Patrick

andrea tasselli:
Linwood Ferguson:
There is also some bad advice in there for local normalization (as well as being out of date).

I don't see how what is in there can constitute "bad advice". Am I missing something?

I haven't looked in a couple years, but the section on local normalization at the time put it forward as a routine thing, and gave pretty much routine instructions for doing it. Juan himself (the pixinsight author) once said on his site "I wish he would take that part down", as at that time the recommendation was to use LN in only specific cases. LN has changed dramatically since then, so if that section is not updated, it is now out of date also. But that is what I meant by bad advice - I figure if the author of the program says don't use that advice, that's a fair attribution.

Postscript: Just to be clear, I used and valued what I learned from that site. I just think due to its lack of updates it needs to be used with caution, much as if you used a Windows XP manual to learn about Windows 11.

Linwood Ferguson:
I haven't looked in a couple years, but the section on local normalization at the time put it forward as a routine thing, and gave pretty much routine instructions for doing it. Juan himself (the pixinsight author) once said on his site "I wish he would take that part down", as at that time the recommendation was to use LN in only specific cases. LN has changed dramatically since then, so if that section is not updated, it is now out of date also. But that is what I meant by bad advice - I figure if the author of the program says don't use that advice, that's a fair attribution.

Postscript: Just to be clear, I used and valued what I learned from that site. I just think due to its lack of updates it needs to be used with caution, much as if you used a Windows XP manual to learn about Windows 11.

From where I stand it is pretty routine processing to me but I also see it is embedded into WBPP (not that I'd use it, btw) so they must have considered worth having. While the process might have been optimized to a degree I don't see why the specific section recommandations can constitute bad advice. It is up to you to decide whether you want to follow its advice or not and then look at the results. The data usually does the talk pretty strightforwardly.

andrea tasselli:
Linwood Ferguson:
I haven't looked in a couple years, but the section on local normalization at the time put it forward as a routine thing, and gave pretty much routine instructions for doing it. Juan himself (the pixinsight author) once said on his site "I wish he would take that part down", as at that time the recommendation was to use LN in only specific cases. LN has changed dramatically since then, so if that section is not updated, it is now out of date also. But that is what I meant by bad advice - I figure if the author of the program says don't use that advice, that's a fair attribution.

Postscript: Just to be clear, I used and valued what I learned from that site. I just think due to its lack of updates it needs to be used with caution, much as if you used a Windows XP manual to learn about Windows 11.

From where I stand it is pretty routine processing to me but I also see it is embedded into WBPP (not that I'd use it, btw) so they must have considered worth having. While the process might have been optimized to a degree I don't see why the specific section recommandations can constitute bad advice. It is up to you to decide whether you want to follow its advice or not and then look at the results. The data usually does the talk pretty strightforwardly.

What's in WBPP is a completely different implementation that what was there at the time lightvortex wrote that.

But that's fine -- bad vs good vs mediocre is a matter of opinion, maybe that was not the best wording to use to raise the issue.

andrea tasselli:
Die Launische Diva:
Calibrating the individual darks is one such bad advice. While in theory there should be no problem in doing so (setting aside the unnecessary computational and storage cost), in practice (in PI) calibrating your darks may lead in clipped master darks.

I was referring to the part about normalization, not dark calibration. But in the specific I still don't see it as "bad advice". I do it routiinely for my DSLR and CCD work (but not CMOS) and it most definitely does not lead to clipped master darks. I'll stick my neck out and add that without it you'll end up with poorly calibrated light frames, if you do remove the bias signal from said light frames as you should for CCDs (and DSLRs).

I have very limited experience with local normalization but I have found in practice that calibrating the dark frames from my DSLRs resulted in calibrated darks with zero values. For reference, see Section 6.2 of Bernd Landmann's Guide to Preprocessing of Raw Data with PixInsight, which I forgot to mention in my initial post and it is a recommended read for all PI users.

Die Launische Diva:
I have very limited experience on local normalization but I have found in practice that calibrating the dark frames from my DSLRs resulted in calibrated darks with zero values. For reference, see Section 6.2 of Bernd's Guide to Preprocessing of Raw Data with PixInsight, which I forgot to mention in my initial post and it is a recommended read for all PI users.

I have several DSLRs. Nikon D610, D7000 & D5100 as well as Canon 40D and 1300D. Also Fujifilm XT-1. None of them gave me the results you suggests. The weirdest it ever got was with the XT-1 with a minimum level of 10^-11 and the D5100 with a minium of ~10^-12. Yet none has been zero. The Canon 1300D master dark after super-bias subtraction looks like the following:

Hardly close to the bias mean value the guide you suggested say it should be. So I suppose there is variation in the results to be expected depending what you're shooting with.

I agree with Adam Block fundamentals it’s worth the money. In addition you can’t go wrong with noise and star exterminator. All good investments and will save you time.

Another thought: If you are just starting in PI and using WBPP, I would turn local normalisation off initially … it adds huge amounts of processing time and doesn't always improve the result. You can always try it later on if your gradients look iffy, but with good quality source images it doesn't add a lot in my experience.

There's a very good youtube video by galactic explorer where he reprocesses an old M31 using PI. I'm a beginner, using the trial version and I found it really helpful.

Trevor Bray:
I agree with Adam Block fundamentals it’s worth the money. In addition you can’t go wrong with noise and star exterminator. All good investments and will save you time.

+1. I am mid way through Fundamentals and it's excellent. PI is a bit different from other things (such as Photoshop) and so it is actually worth putting the time to learn it from the ground up. You can just watch a video on this or that workflow, but IMHO it's far better to develop and understanding of what's going on first. PI has its own internal logic but it's a very good one, I think.

Andy Wray:
Another thought: If you are just starting in PI and using WBPP, I would turn local normalisation off initially ... it adds huge amounts of processing time and doesn't always improve the result. You can always try it later on if your gradients look iffy, but with good quality source images it doesn't add a lot in my experience.

Totally agree with this.

Andy Wray:
A lot of people will probably criticise me for recommending the following, however I found using the Weighted Batch Pre-Processsing script (WBPP) the easiest way to make the transition to Pixinsight from a stacking/integration point of view.

I would use this script too, if I didn't have to watch out for bad subs where the stars trail, or are completely unusable. Subframe Selector and Blink help me to make sure no bad subs go into my final stacks.

One of the fastest ways to learn Pixinsight is via content written by renowned experts: Adam Block's tutorials, RBA's "Mastering Pixinsight" etc. That way common mistakes/pitfalls are avoided and you can learn in the most straightforward/efficient way. There are some things that can only be refined through experience but at least you will have the know-how/basics to improve that experience quickly.

Related discussions