Generalised Hyperbolic Stretch Version 2.1.0 - Forums

Topic starter

Apr 2, 2022 · Edited: Apr 2, 2022

Generalised Hyperbolic Stretch is a Pixinsight script designed to provide a comprehensive and integrated environment in which to design, appraise and apply image stretches. It takes its name from the generalised hyperbolic equations that provide greater control than traditional approaches, eg histogram transformation.

I have today released version 2.1.0 of GHS. A number of features have been included in this update, including the following:

Stretches are now recorded in a view’s history with a fully parameterised script instance, allowing this to be dragged and repeated on another view for example.
You can now transfer the selected readout point on the histogram to the stretch parameters by clicking the reset button to the right of the relevant parameter’s slider.
You can now toggle the preview between “before” and “after” by ctl-click (cmd-click on a Mac) on the preview image.
The maximum zoom available on the histogram zoom slider can now be increased in the preferences dialog (use with care on slower systems!).
A new stretch visualisation using a greyscale mapping has been added below the histogram – in common with other chart features, this can be turned off if preferred in the preferences dialog.
The histogram display has been improved for single channel (R, G or B) stretches so the histogram of the stretched channel can more easily be judged against the unstretched channels.

Plus a number of other changes/improvements/fixes.

The script can be obtained by adding the following repository to your Pixinsight>Resources>Updates>Manage Repositories list: https://www.ghsastro.co.uk/updates/. Alternatively it can be installed manually via the Feature Scripts facility from the following download: https://github.com/mikec1485/GHS/releases/tag/v2.1.0

Many of the changes introduced in this release have been motivated by the excellent and helpful feedback we have received from users. I would like to take the opportunity to thank everyone who has downloaded GHS and particularly those who have taken the time and trouble to provide their feedback. Your feedback helps to make GHS as useful as it can be for everyone, and your comments help motivate Dave and I in developing the script.

Clear skies and happy stretching

Mike

Helpful

reglogge

Apr 3, 2022

Thank you, thank you, thank you! It's amazing how quickly you add new and very useful features. GHS is by now an absolutely indispensable part of my workflow.

Well Written Respectful Supportive

Mike Cranfield

Topic starter

Apr 3, 2022

Thank you, thank you, thank you! It's amazing how quickly you add new and very useful features. GHS is by now an absolutely indispensable part of my workflow.

That's great to hear - thanks for the feedback.
CS
Mike

Well Written Respectful

Torben van Hees

Apr 3, 2022

A big thanks from me, too. I’ve always been struggling a bit with stretching and using GHS I feel I have much better control over the process. I think it brings the “PI experience” (data-driven tools - working with numbers instead of ”visual impressions”) to stretching.

Mike Cranfield

Topic starter

Apr 3, 2022

Torben van Hees:
A big thanks from me, too. I’ve always been struggling a bit with stretching and using GHS I feel I have much better control over the process. I think it brings the “PI experience” (data-driven tools - working with numbers instead of ”visual impressions”) to stretching.

Thanks Torben - I completely agree with your comments there. I think @David Payne's insight of using the GH equations in the stretching process has been something of a game changer!

CS
Mike

David Payne

Apr 3, 2022

Thank you, thank you, thank you! It's amazing how quickly you add new and very useful features. GHS is by now an absolutely indispensable part of my workflow.

There is no keeping up with @Mike Cranfield . In general I am testing something he put in four beta versions ago. Thanks and kudos from me too, Mike.

Jeff Ridder

Apr 3, 2022

This is a great and timely update! I updated yesterday, used it, and only later read the above about recording the stretches in the history. This turned out to be very fortunate since I had to go back to linear to fix a problem, but really REALLY liked what I had. NO problem. Fixed my issue then dragged the stretches from the previous onto the update. Saved me a boatload of time. https://astrob.in/eb2u7o/0/

Mike Cranfield

Topic starter

Apr 3, 2022

Jeff Ridder:
This is a great and timely update! I updated yesterday, used it, and only later read the above about recording the stretches in the history. This turned out to be very fortunate since I had to go back to linear to fix a problem, but really REALLY liked what I had. NO problem. Fixed my issue then dragged the stretches from the previous onto the update. Saved me a boatload of time. https://astrob.in/eb2u7o/0/

Wow - that is an absolute cracker Jeff - I love it - I'm so glad we got the update out in time to save that!

Klaas Pieter

Apr 9, 2022

I’m going to try it! It sounds very good!

Miguel G.

Nov 3, 2022

Hi,

I don't know where to submit a bug report.

Just downloaded v2.2.3, where the script is also available as a module. But the module fails when loading:

<* failed *>
*** PixInsight API Error: Module: /[…]/PixInsight/bin/GeneralizedHyperbolicStretch-pxm.so
Cannot load library /[…]PixInsight/bin/GeneralizedHyperbolicStretch-pxm.so: (/lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.33' not found (required by /[…]/PixInsight/bin/GeneralizedHyperbolicStretch-pxm.so)): Module load error

Running linux mint 20, kernel 5.4.0. System completely updated but glib reaches only 2.30 version

m

Mike Cranfield

Topic starter

Nov 3, 2022

Miguel G.:
Hi,

I don't know where to submit a bug report.

Just downloaded v2.2.3, where the script is also available as a module. But the module fails when loading:

<* failed *>
*** PixInsight API Error: Module: /[...]/PixInsight/bin/GeneralizedHyperbolicStretch-pxm.so
Cannot load library /[...]PixInsight/bin/GeneralizedHyperbolicStretch-pxm.so: (/lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.33' not found (required by /[...]/PixInsight/bin/GeneralizedHyperbolicStretch-pxm.so)): Module load error

Running linux mint 20, kernel 5.4.0. System completely updated but glib reaches only 2.30 version

m

Hi Miguel

Unfortunately there are several different versions of Linux available. The GHS module has been compiled based on Kubuntu 22.04. Unfortunately it seems your installation has a version of glibc earlier than 2.33 so will not run the new module. Sorry about this.

Helpful

Miguel G.

Nov 3, 2022

Hi Mike. Thanks for the reply. Yes, just as suspected.
My mint version is based on ubuntu 20, which is an LTS version supported until 2025. Kubuntu 22.04 seems pretty new. Just wondering if other users with LTS versions but lower than yours would face a similar issue. It's just a guess and of course I can deal with this since I can still happily use the script version

silversilk

Nov 3, 2022

I know I am late to the game on this topic but feel I need to add some gratitude for the tool, as well as the online resources via youtube and your website on how to properly use it.

I have seen a dramatic improvement in post processing as a result of being able to put a proper stretch to the data. I am learning that you can really make or break the data in this crucial step. All the hard work, time, and money spent on data acquisition (equipment purchased, finding good backfocus, eliminating tilt, collimation, good guiding, acquiring a good amount of total exposure time, getting to a dark site, etc, etc, etc…) can all be undone or wasted with a poor stretch.

Thank you again for providing such a controllable, and repeatable method to help us get t he most out of our data.

Dave

Respectful Supportive

David Payne

Nov 4, 2022

Thank you Dave for the kind words, and you are very welcome. It has been a lot of fun. I think @Mike Cranfield's work on porting the functionality over to a module/process is a giant leap for GHS. It will be interesting, now, to see if other scripts make use of the transforms.

I agree that stretching is the single biggest transformation that we make to the data, yet in some ways it is one of the simplest because an individual pixel output brightness is only dependent on the input pixel brightness and the transform. Most other processes are more sophisticated in that they not only look at an individual pixel brightness, but also the brightness of pixels surrounding it (eg. noise reduction, sharpening, deconvolution, wavelet transforms).

Clear Skies!
Dave

Well Written Insightful Respectful Engaging Supportive

Hermann Pais

Sep 6, 2024

@Mike Cranfield - THANK YOU! Thank you!

GHS is amazing. It salvaged an entire night of observing because I undersampled. 90 subs of 180 minutes each, taken through a Celestron EdgeHD8, Optec Lepus 0.62 Telecompressor, ASI533MC-pro, Optolong L-pro (bortle 7 skies).

GHS, using David Payne's 3 steps (hyperbolic, linear, hyperbolic) helped me produce a result that at least to my beginner's eyes is beyond what I expected. I'm truly thankful for your work and can't sing the praises of GHS enough.

This weekend is supposed to be the last good stretch of clear skies in the Pacific Northwest, and I'll try to image once again. This time, of course, extending my subs to 500 or 600 seconds.

Helpful Respectful Engaging Supportive

Mike Cranfield

Topic starter

Sep 6, 2024

Hi Hermann

So glad GHS has helped you produce this very nice image. Credit must also go to David Payne my collaborator in GHS development!

CS, Mike