PixInsight Forum (historical)
PixInsight => Wish List => Topic started by: georg.viehoever on 2013 December 05 01:26:04
-
I am kind go tired seeing all those questions about the basics PI and image processing in this forum, things like
- correct setting for RAW format
- Color casts (->color calibration)
- STF vs. histogram transform
- applying the same process to multiple images
- tuning StarAlignment parameters
-....
How about writing a book about the basic of astronomical image processing, with tutorials on doing it in PI (and maybe some other suitable software)?
Georg
-
That sounds interesting.
-
well i have long thought about doing this but i am a terrible writer, don't necessarily have the authority to do it and to do it right would be an incredible amount of work. just one process could fill an entire book!
rob
-
Georg,
You assume that somebody would actually read the book before asking a question here.
An optimistic point of view IMHO.
-- bitli
-
I think a FAQ section in the forums would be very helpful. Sometime ago there was a wiki for PixInsight where the FAQ could have been hosted, but it disappeared.
-
Hi
A book would be very nice , but would be a huge commitment for someone :o
But A " newbies start here" section might be a good idea with pointers of where to look for info( vids written tuts etc) and or basic information
Not quite a Wiki but with user input ( keep things concise as some thread here are very long)
Regards
Harry
-
And that newbie section in many cases would only need to be pinned threads from the forum.
How does one get a thread pinned anyway? I've seen several that I've felt ought to be pinned.
-
Harry:
Your YouTube videos are the best documents for "newbies" like myself that I've come across. Excellently done. If you take them in order, they even sort of produce a coherent workflow for future use. It is that lack of a suggested workflow that I think is what baffles most newcomers to PI. That and a lack of explanation of some of the parameters and settings that should be applied rather than defaults. If the default setting doesn't work, you are relegated to trial and error to figure out what to do. Reading the flyover texts for each button helps, but only if you understand the technobabble that soem of them are written in. Some are easy to understand, some are written in an alien exo-language. :-[
-
I may be kind of old fashioned, but I still like written documentation. I believe there are people who prefer "learning by doing", others like videos, and others like reading documents.
When following the forum, I am so tired of answering the same questions again and again, such as
- correct RAW settings
- STF vs. HTF
- creating/saving/loading process icons
- setting up StarAlignment so that it actually aligns
- linear vs. non-linear
- how do process containers work
- ....
I could probably continue for quite a while. I know that those questions are from entirely different categories, some are very basic, some are advanced. But even after following this forum for years, I have the feeling that I am missing some of the powers of PI because I just dont know some of its tricks. As an example: how is anyone supposed to discover without hints that it is possible to transfer the zoom level of one image to another by dragging/dropping the tabs onto each other. Certainly not by accident, and not by coming across the one or two messages in the forum that mention this.
Would it be possible to find a dedicated group of people writing such a book? Does it need to be a commercial endever, or would be an open sourced book/wiki/PDF be the way to go? Would the PTeam contribute? Has someone else other useful ideas?
Georg
-
Perhaps a wiki is the way to go. With good peer review and lots of authors it wouldn't take so long to accumulate a valuable resource. While I'm no expert I could see contributing content. For me the main obstacle is that my computer is so slow that generating meaningful screen shots can take longer than writing the text.
-
Yes, I think that in the first instance a wiki is the quickest way to get things moving. A book would be a big effort that would (optimistically) take at least two years to appear. Additionally, a wiki could provide the basis for a future book. I suspect also that various members of the PI team would be unable to resist contributing to a wiki.
Geoff
-
Hi All
There was a few years ago a attempt at a wiki here , which was not overlay successful :-[ and it was taken down
IMO I do not think it will work well and we need better more concise information , I need to think about this hard
and offer alternatives as I do not wish to be seen as " Ney doer "
Kind regards
Harry
-
...As an example: how is anyone supposed to discover without hints that it is possible to transfer the zoom level of one image to another by dragging/dropping the tabs onto each other. Certainly not by accident, and not by coming across the one or two messages in the forum that mention this....
I belong to this group of users. Is there another hidden feature?
-
...IMO I do not think it will work well and we need better more concise information , I need to think about this hard and offer alternatives as I do not wish to be seen as " Ney doer "
Hi Harry,
your videos did more for popularizing PI than any other single effort. This was and is a fantastic approach to teaching the basics, and I can only hope that Juan one day will invite you to his private island (you have to see his private relativistic space cruiser first, though) ;) . But while videos are good for teaching certain procedures, they are essentially useless for teaching people about accelerator keys or the fine points of setting up RAW imports.
I think the wiki failed because we did not have a central editor (wikis tend towards chaos if nobody "manages" them), and because the markup language was a major obstacle for many who wanted to contribute. Not sure how to get around this.
Maybe others with more experience on web based documentation projects can comment.
Georg
-
OK, Here's the thing.
When I started imaging, I had no idea what I was doing. In a period of 3 months I learned I need to take many images, flat, darks, bias. Started to understand this, and then start hearing about "flat darks, and other names so same things.
My point is with me I dove right into this, realizing I had be doing it all wrong. So I joined a Yahoo group, where people taught me all this about taking many different types of subs. That was daunting enough.( especially flats, still is). After I had a few images, maybe only 4 or 5 subjects, I used DSS to do all the ( everything it does to get the final stacked image). I read the documentation, but really did not get all of what was really going on. Just knew if I setup with these parameters, It turned out pretty good. I fight a lot of LP. After using DSS ( which is a great product by the way ) for a couple months, I started to hear about other products, mainly "Images Plus" and "Pixinsight". After corresponding with several users of both, I decided to go with PI. (Dang glad I did too).
For me, either one would be a learning curve. But I was particularly interested in the DBE and other noise reduction features. Did not know how they worked just heard they do.
OK, All this being said, I had no experience, in what integrating, calibrating, stacking, aligning, etc. really was all about. Remember DSS just did it. No real understanding of what an image consisted of.
But I was up for a challenge. So Before I took the trial, I went to Harry's sight ,and you tube, and watched a lot of videos. I was very impressed. So I started the trial.
The problem was nowhere in the videos, did it ever really explain, if you use a DLSR, this is how the image science behind that works. ( meaning bayering-debayering, RAW format, etc). What integrating, calibrating, aligning, stacking, etc. really does. For PI this is needed info.
So once I had it, I started reading all the forums, but I have to say, I was lost most of the time, because I just could not follow the conversations.
But, over the last few months, I have received invaluable info, from those like ( pfile rob, and others). I am sure he has figured out, I did not know much about imaging at all when I started this. But his patience has really helped to try to grasp this.
This is new stuff, a science in itself, and I have to say PI does not pull any punches.
As far a the basic questions, this is to be expected, because we just don't all start out as intermediate imagers and processers.
But I can certainly say, that using PI, will educate you pretty fast. But there is so much to learn.
A good example is Pixel Math (what I see as one very powerful tool), but if you are not an engineer( and I don't see why you need to be to be an astrophotographer), then you may be taken aback by the scripting language needed to use it.
I am barely a scripter, so I at least understand a formula when I see it, But I still cannot, look at my screen and say I want to do this , using this and this, and know what formula to put in the command line to get it.
I think for me, the lack of immediate local documentation is an issue. But I have learned to find it on other sites by now.
Everyone is different, in the knowledge base, and everyone will have different questions that need to be answered.
I cannot even begin to comprehend some of this yet.
It would have helped me to of had much more imaging experience before I began this, but I am not one to back away from a challenge.
But Georg, there is just no way for many of us to know what you and others already know and take for granted about this science.
As far as your above issues, I have read the documentation on a lot of it, but if I don't understand the science behind much of it ie: correct RAW settings, how am I going to know what works for me? Another is linear vs non-linear, I knew the difference, but did not really know why a process would or would not work with it. That's just the way it is.
This is why the questions are asked.
One thing I will mention, is many times I try to use the search to see if I can find a previous thread on a subject, but I have to say for me the search has not been helpful on many instances. So I find myself just asking a question.
I do not have an advanced setup to work with, but that is how I want it for now.
This software has allowed me to create images I never thought I could ever achieve.
And I know there is a whole lot more I can do with them, I just need time to learn them and experiment with them.
I have only been doing this since this last April. And using PI since last July.
Most sites that encourage PI, mention , a steep learning curve. No kidding!!!!.
Scott
-
Hi Scott,
so how can we make this learning curve less steep. The technology behind PI is amazing, and I am always wondering how Juan pulls this off. But I feel that few of us really have the ability to leverage all the power that PI gives us. IMHO, providing (more) decent documentation would certainly help, and explaining a fuller view of the concepts of astrophotography and PI is another. I am still puzzled about the best way to make PI a better place.... How can we making this learning curve less steep?
Georg
-
Hi Georg,
Unfortunately, that is a tough question to answer. And there are most likely many answers. I can only speak for my self by saying, being a Test Engineer myself, I am a fanatic when it comes to documentation. In my line the better the doc the less the customer issues I have to deal with. But I am aware you cannot make someone read. And you can get yourself in over your head trying to keep up with it.
So I say documentation is essential, and every feature does need it. The LIVE feature that many tools (but not enough) have is really nice. I can look at the doc, as I am working the image. This is very handy.
The first tool I learned to use, (well almost the first)was the DBE tool. This helps out with my LP issues. But in reading the doc, (which is fine by the way), was confusing, because being a complete lamen at this, when I read the mouse over explanations, I had no idea what it was saying. I knew nothing, and still am learning about sigma units, mean background, 2D-surface spline, etc,etc,etc.
So in many cases more elaborate documentation isn't that helpful. But for some I am sure it was great info.
One thing that would be nice is some kind of doc "explaining a fuller view of the concepts of astrophotography and PI" as you said. . But I think it should include some simple items like DSLR formats explanations like the ones in the "Format Explorer", but more verbose. At first they meant nothing to me. Now I have somewhat of an understanding, the the 3 bottom explanations in the top section make sense. But I have no idea what the top one (Create super-pixels mean). It's things like this, that I can't expect you to document, I just appreciate the answers that I get.
Another thing that goes with the format is the "hints" in some tools. Had no idea what they were for. Even after reading the description. Do now of course.
I just don't have an answer for you, except docs for each tool. Obviously a book like those from Jerry Lodriguss on PI would be great. But that is a heck of a chore.
One feature of PI, that I am just learning to use, are all the scripts. Lot of talk about the mschusters scripts in the Scriptbox.
But once again, if you don't know what the science is behind it and understand what the parameters are to be and what to expect, the tool doesn't makes sense. Also even how to download a script and execute it properly is not explained. I have run into dependency issues when trying to execute some scripts.
My biggest gripe (and this a gripe to myself), is I do not know what to expect for numbers on my images. I am trying to learn about FHWM, noise values, sub exposure estimates, I could go on and on.
I would appreciate someplace where I could get some amatuer and non-professional data images to open with some tools and get ideas of different values. But I don't know if just a sub or just the stacked image would do, or possibly the final processed image. But I would like is that I could open and compare to learn more about what expectations or goals I could look forward too. I know my data is not the best, but I have no idea what good data numbers look like for some of the tools and scripts described above and what I should shoot for.
So you see, I think just knowing that there is someone , and usually several users that can answer questions as they arrive, is the most important attribute and most invaluable feature of this forum, which as far as I am concerned is part of PI.
I think you said it all when you said "The technology behind PI is amazing", however that means it is an advanced piece of software, and will require a lot of educating and learning.
Scott
-
Just to pick up on Scott's point, I have previously shared my cribsheet with the forum, but it is now up to Rev17 and still I keep adding new stuff. The problem is twofold, first, I build it to match my preferred workflow so it has lots of personal elements and, second, if I didn't constantly tweak it the thing would be out of date quickly or would miss out on my recent learnings (PixelMath, with a huge thanks to Gerald's tutorials). Just imagine if we had a book a year ago. It would need to be substantially rewritten just based on the changes we have seen in the last few months culminating in the Ripley improvements. And it would in some ways provide false security that you are getting the most out of the program when that would never be the case.
I guess I am firmly on the side of you don't get full value for the vast investment Juan and his development team have made unless you dig in and get your fingers dirty. We all come from some processing background so newbies rarely need that kind of primer. Its more about learning new PI ways of doing things (and IMHO they are so much more logical than anything I used to do with Layers in you know where). And if you are looking for theory, we already have Berry's Handbook of Astronomical Image Processing which remains the bible (I still thumb through it regularly).
And for people like me, and I am guessing Scott as well from his note, we are not numbers guys so just a "open this, - set these parameters - hit apply" type of written instructions will not teach us how to get a true understanding of what is really going on. That's why this forum is so valuable and really one of the best I have seen as far as a teaching tool for newbies, wannabe non-newbies but still drinking from a firehose (and I include myself firmly in that group), and old hands alike.
For what its worth,
Jim
-
I found this about wikis http://therealkatie.net/blog/2013/dec/11/your-wiki-dump/ . Certainly something to consider should we resuscitate the wiki approach.
Georg