Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - aworonow

Pages: [1] 2 3 ... 14
General / Re: Error message using Batch Preprocessing
« on: 2019 March 06 14:01:40 »
If you are not using overscan (generally, most imagers do not), click the Bias tab at the top and uncheck the "Apply" checkbox in the Overscan box.

If you are using overscan, then the values you specified must be incorrect/inconsistent.


General / Re: MAD and Average Absolute Deviation
« on: 2019 February 26 08:46:16 »
You are correct...we statistical types (my career) use MAD most often for a deviation from the median.


I have written a new, logical, rigorous, flexible, and user friendly--Excel spreadsheet that makes generating subframe-weighting equations (for use within SubframeSelector).

Here is the link to a zipped folder that contains the spreadsheet and the instructions:

The reason to use the SubframeSelector (SS) is not just accept/reject subs, but to assign rational weights to each frame. This is an important operation, and if you do not use SS, you should learn to do so. The idea of a rational weight scheme requires a way to evaluate variables that make a subframe better or worse. Within SS the facilities for evaluating subs is improved over the previous version, but still it is rather minimal and not very interactive. Because of this, so folks have written Excel spreadsheets to help. I am among those folks. Furthermore, as far as I know, at this time mine is the only one that runs on the new version of SS.

If you've used PixInsights SubframeSelector Script before, then you probably notice there is another version now that is under Procedures. It runs much faster and uses updated algorithms for FWHM, star identification, .... , and has 3 windows. It also is now programmed (the weighting equation) using JavaScript.

Alex Woronow

M42 is a prime (and rare) example of an astronomical target whose brightness range is greater than the dynamic range covered by our cameras. The best images of M42 require a short and a long (relative, speaking) to cover all that range. Then the HDR capabilities in PI are used to blend the two separate stacks together.



It works for me. I tried it on a fully processed image from a grayscale camera (RGBHS). It added color to the biggest brightest stars (and to some quite small, but, presumably, bright stars). The effect was to dull the stars a bit, but I think that can be remediated without much difficulty.

Thanks for a nice addition to the tool set.


BTW, I often use Repair HSV Separation to good effect, but this is faster and does the job well.

Would be nice! Could do a crop on all the images in a ImageContainer and analyze those then and manually transfer learnings to the full frames. OUCH!


New Scripts and Modules / Re: Debug Environment for Scripts?
« on: 2019 February 02 05:32:34 »
I agree with Mr. OldSkyEyes. I'm trying to write a small script, and am neither familiar with the PI library, nor with JavaScript. (I' bought 2 books on the latter.) Other than copying code from other scripts and hoping they are compatible, I'm at a loss for how to efficiently get to the end game. A rich-featured debugger would be a real head start.


I have a process container that, as the first process, has a pixelmath process that creates a new image. I want subsequent processes to act on that new image. However, the subsequent processes act on the parent of the new image instead.
If I click the container processes one at at time, then the pixelmath's new image is active as the next procedure executes, and as expected, acts on that new image.
Is there some reason that the process container does not work the same way? Is there a coercion I can use to make it do what I need?

Thanks, Alex

Bug Reports / Re: SubframeSelector does not handel repaired 'nan'
« on: 2019 January 08 07:16:26 »
I guess I was not entirely clear. The 'nan' is written by the program when my equation is in error. After it is written into the Weight column, the problem arises that the program appears to try to read the field before calculating an new equation, and throws the error I reported.
The only field I change is the Expression.


Bug Reports / SubframeSelector does not handel repaired 'nan'
« on: 2019 January 08 06:50:53 »
I entered an equation that cause the all the weights to become 'nan'. Repairing the equation (setting it to 1-.5) and pressing the right arrow in the Expression window then causes the console to report "*** Error [001]: ReferenceError: nan is not defined" and no execution." Clearing the measurement table and remeasuring does not clear the nan weights, and the same error persists.
Even clearing the Measurements Table and the subframes, reload and remeasuring, the nan is still there (Cached?).
And, even exiting the procedure and coming back in, everything is restored, including the nan's.

Only by clearing all three windows, exiting the procedure and coming back in, can one remove the dreaded 'nan'.


Bug Reports / Re: SubframeSelector needs exponentiation operator
« on: 2019 January 07 09:25:07 »
Juan, Thanks for the quick reply. It's been 20 yrs since I programmed, so I am unlikely to try such a modification. I thought I'd just use the old SubframeSelector script, but I see that it, and the new one, do not agree on the values of most measurements, or, often, even the relative ranking of the measurements. Presuming that "newer is better," I guess I'll bit the bullet and modify my excel equation.


Bug Reports / SubframeSelector needs exponentiation operator
« on: 2019 January 07 07:34:15 »
In the latest version of PI, the new implementation of the SubframeWeighting process does not accept exponentiation as either ** or ^. As many of us use an external program (usually a spreadsheet) to construct a weight equation, and, in my case, one that uses '^', this is a significant issue. I have not tried using math.pow(a,b), but that would entail a lot of rewriting and testing of my spreadsheet-equation generator (and maybe of other folks too). Could '**' be implemented, maybe, please.
Until then, I'll have to use the old script.

(I use ^2 to get distances from the best value of a variable normalized to the range of the value.)

Thanks, Alex

General / Making alpha channel truly transparent???
« on: 2018 December 05 07:43:46 »
Hi, I have an image in which I generated an alpha (transparent) channel. Now I want to display that image in PI such that the transparent parts are, actually, transparent and show whatever is below them on the PI workspace. All I can seem to get is various opaque checkerboard patterns and such. What is the procedure to get a transparent transparency?

General / Re: Load Process Icons at startup?
« on: 2018 November 14 15:12:08 »
What I do is make a project template...that is  a new project that I have imported all my icons into. Then, I copy this project and rename it to being actual processing.

Pages: [1] 2 3 ... 14