How to save an individual Process Icon

[Niall Saunders]

Quote from: Niall Saunders on June 24, 2010, 15:08:21
especially now that I can save individually selected ProcessIcons.

How do you do this?  Copy/paste the code in the History explorer?

Can you save Process Containers?

I tried looking for answers to this last night and couldn't...  I think I ran into a thread from a year ago that said this wasn't available, but I am not sure.

Thanks,

Jim

Jim,

First you need to use the little blue "NewInstance" triangle associated with your process - but, instead of 'dragging it to an image', you drag'n'drop it onto the Workspace background - thus creating a 'Process Icon'.

You can now give the process a 'meaningful name', by clicking the tiny little 'N' at the right hand side of the icon. Just be aware that you cannot use spaces in the icon name, nor can you start with a numeric character, nor can you use a host of other 'illegal characters'. So, that often means that you will end up resorting to CamelCaseToBeAbleToMakeYourNameUnderstood

Now, make sure that you only have ONE of the ProcessIcons selected. Right-Click, and then select "Save Selected Icons..." - this will allow you to save ONLY those icons that you had selected, i.e. ONE ICON in this case.

As a further suggestion, whenever I 'rename' the ProcessIcon (as described above), I also Ctrl-A and Ctrl-C the name I am about to use. That leaves the name in the Clipboard, and I then Ctrl-V the name into the filename box in the Save Icon user dialogue.

I also keep a close eye on the Processing Console dialogue to ensure that I didn't mistakenly select - and therefore save - more than one icon. (This is easy for me, because I keep the Processing Console permanently open on a third monitor - so I DO get so see all the PI feedback, all of the time - lots of useful stuff that I used to never be aware of, becuase the PixInsight Console was just always 'in the way')

HTH

Cheers,

[varmint]

First you need to use the little blue "NewInstance" triangle associated with your process

Thanks Niall.  I was hoping I could somehow save the details from the History Explorer for the work done on a particular image/step.  I'll have to be more careful about using Process Icon's first.

And since it's been a month since I last gave anything a go, I figure I'll be redoing some processing anyway.

Jim

[oldwexi]

<I was hoping I could somehow save the details from the History Explorer for the work done on a particular image/step. 
I'll have to be more careful about using Process Icon's first.>

Hi Jim!
But thats exactly what you can do.

I move the blue instance from the history explorer to the workspace, rename it, save it and so
i have a process container saved with all the processes and parameters and steps i did.
Thats the good part.
The bad part is i have to save the mask files, work files and their History Explorers too.
This is much work at the end of a long session at 2 o'clock in the morning.
I would appreciate if a save all files with their history explorers would be there - have not found it so far.
Gerald

[Carlos Milovic]

Hi Gerard

That will be the main focus of the "project system". It has not been implemented yet, but has a high priority at this stage. It may be included during the 1.6 or 1.7 cycle.

[varmint]

I move the blue instance from the history explorer to the workspace, rename it, save it and so
i have a process container saved with all the processes and parameters and steps i did.

Thanks Gerald...hmmm, I will have to look for this.  I can see the blue triangle on each Process Icon, but I didn't notice it on the History Explorer.  Also thanks for the Tip on saving the Mask Files and such.

One thing I noticed last night with saving Process Icons is that you have to enter your parameter values before you drag the blue triangle to the Workspace if you want them to be saved with non-default values.  Not a big deal, but glad I caught that.

Also, I'd be curious to know why PI only lets you load one Process Icon set at a time?  I was saving individual Process Icon instances for my Image Integration steps but could only open them all if I selected "Merge Process Icons"...maybe that's just a semantic thing?

I know someone's going to bring up Process Containers...I'll look at those next, but last time I tried I couldn't see where to Save a Process Container.

Thanks for helping this noob out.

[Niall Saunders]

Hi Jim, Gerald,

I certainly know that Juan's intention is to provide for a 'single-click' solution to the 'I need to go to bed NOW' problem. That way, ALL your open images, each with their full processing history, can be saved in one file. Now, although I have downloaded the pre-release Beta version of v1.6.1, I just simply have NOT had time to install it and see if this option has made it into this release.

In the meantime, it is actually still worthwhile getting a 'feel' for how you can already manipulate the information contained in the Processing History. See how an individual step in the History Explorer can be extracted, either as a Process Icon, or as an 'apply-able' process in it's own right - capable of being 'dropped' on an active image. Have a look at the difference between extracting a 'single step' in the History compared to extracting the 'full history', as a Process Container.

Note how PI can only have one 'active' Process Container, yet you can have multiple sets of 'process container contents' saved on your workspace. Each time you double-click on a saved set of Process Container steps, you will see that a warning message will appear to tell you that you are about to over-write the 'active' Process Container.

In History Explorer, note that you can move back a step at a time, or can move back to an earlier stage altogether.

Note that any individual step within a Process Container list, just like in the History Explorer, is just a single Process in its own right, and that it can therefore be dropped on the workspace to create a new Process Icon, and that this new icon can be renamed (if desired) and can also then be saved to disk (as needed).

Be aware that any process icons saved to disk, and then re-loaded, will appear in the same Workspace as they were initially located, prior to having been saved. They will also occupy the same physical 'location' on that Workspace - so be prepared to go 'looking' for them when you re-load them. They will NOT end up where you 'drop' them, although they CAN be 'dragged in' from the likes of Explorer, and can also be 'double-clicked' to open them from outside of PixInsight.

As I think I said earlier, be aware that you 'may' have multiple Process Icons selected across several Workspaces. If you only intend saving ONE actual process, then just make sure that was ALL that was saved - by paying attention to the Console Output (very easy, as I said, if you have the luxury of multiple monitors, and can 'float off' the Console to an auxiliary display) - the console output will tell you what was actually saved.

Have a play around and get a feel for the system - if you are still struggling, come back here and I may then have time to throw some quick videos together.

Cheers,

[oldwexi]

Hi Carlos!
Thanks for the information.
I am sure i will  like this new "project system" .

I like the use of PI very much, as for me PI has the most functions in one program.
Before, i had to use 5 programs, one for this function and one for that function with converting from fits to tif etc....
The functions in PI are more detailed, precise and handle most of the deep sky image processing needs
IMO. (But i am still and for a long time in the learning phase.)

Gerald

[Niall Saunders]

Also, I'd be curious to know why PI only lets you load one Process Icon set at a time?

You can have multiple Process Containers on the workspaces, but only one 'active' Process Container

This is the same analogy as having multiple 'Histogram' Process Icons ('Histo' chosen by way of an example) - each process having a different set of parameters - but you can only have ONE 'active' Histogram GUI, and this will be 'loaded' with the parameters saved as part of the workspace Process Icon.

You can save ANY selected 'group' of process icons (which can also include any number of Process Container 'icons', if you wish) and can save these anywhere on your storage network that you choose.

However, when you 'reload' an individual icon, or a selection of icons, you need to be aware that these will either TOTALLY OVER-WRITE the existing 'live' set of icons, or they will be MERGED WITH the existing set. It is your choice.

Further, all active icons (processes and images) must be UNIQUELY NAMED (this is an 'absolute' requirement for the way that PI works internally), so you should take the time to experiment and see what ACTUALLY HAPPENS if, for example, you start with a new, clean, instance of PI (i.e. with NO process icons on any of the workspaces) and then try to - for example - load the same saved icon TWICE.

Does this help?

Cheers,

[Niall Saunders]

I can see the blue triangle on each Process Icon, but I didn't notice it on the History Explorer

When you open the History Explorer, you should find the 'blue triangle' in the normal location of 'bottom left' on the History Explorer GUI.

If you drag this to the workspace, it will create a 'Process Container' icon - which has the ubiquitous 'blue cube' icon - as used here on the Forum to denote a section with new activity.

However, try just clicking once on a single line of the History Explorer window. First, note that the actual 'code' of the process can be viewed in the right-hand pane, and that this code can now be viewed as either PJSR-type code, or (a new feature) as XPSM code. The (PJSR) code can even be cut'n'pasted into your own PJSR script, if that is what you need to be able to do. (If you cannot see the right-hand 'code pane', then you will need to click on the little 'double right arrow' icon at the bottom of the History Explorer GUI).

Alternatively, just click on a line, and drag it to the Workspace - and you will create a Process Icon associated with the actual process that was applied at that particular stage in your processing history.

Also note that, if a process was applied 'through a mask', then this will be indicated in the History Explorer GUI. The name of the actual mask that was associated with the process will be shown. So, if you hope to be able to recreate ALL of your processing steps, you will need to save THAT image as well - so that you can re-use it as a mask at the appropriate time.

Obviously, to be able to re-use an image as a mask, you may also need to carefully save all the key processes used to CREATE that actual mask as well.

This is where you just HAVE to be 'properly organised'. It is all very well just hacking and clicking like a pixel-crazed demon, but I think it is nice to know 'how' you created that APOD, just so that you can 'do it again' and learn from whatever method you used. For those who know me, I am 'anal' in how I go anout recording my processes. I save individual images and their processing histories at several stages throughout the overall process (nowadays this also includes all the key stages in the Calibration and Integration processes as well). It is only a little bit of hard-disk space, and a few extra seconds after all - but, personally, I find that it really reinforces the 'how' and 'why' of my processing. What it does NOT do for me is improve my images - they ALWAYS end up looking like cr@p

Have fun.

Cheers,

[varmint]

Thanks Niall for the detailed and thorough descriptions, they really helped.  I just attempted everything you mentioned so I have a better feel for it now.  The only thing I didn't test was saving a Process Container.

Your comment about organization, and saving as you go, is a very important one.  I had saved a set of process icons, but not many of the intermediate files (and Masks)...eventually I'll share my progress, but right now I can concur with you.  All my images are looking like cr@p, but I can see potential in them...

[Niall Saunders]

Happy to help - sometimes actually 'doing' an exercise serves you better than just watching someone else 'doing it', which is why I decided not to throw in a video here (no offence to anybody else producing videos).

It can be useful just to 'play' with PI, especially whilst getting the hang of the user interface itself. Once you have THAT under control, then you are more ready to start playing with the controls within the processes. And then it just gets scary all over again

Cheers,