+- Gimp-Forum.net (https://www.gimp-forum.net)
+-- Forum: GIMP (https://www.gimp-forum.net/Forum-GIMP)
+--- Forum: Extending the GIMP (https://www.gimp-forum.net/Forum-Extending-the-GIMP)
+---- Forum: Scripting questions (https://www.gimp-forum.net/Forum-Scripting-questions)
+---- Thread: Proof of concept two (/Thread-Proof-of-concept-two)
Proof of concept two - Tas_mania - 01-14-2024
I improved the animation I posted on the other forum
![[Image: MM-FlagVirtualSphereRotate.webp]](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEjH_TBxWkbTuuHQocwywAszy4wofUG-HbMG-kwDH057NBnW4UMtH8DoygEfG6y_Dy53Wb_mwxxM-kEt3bJmNxY7usT2woJIE5TFybfK0WKkirnRaQLCxbcGJbj_4RXQHyly5MjJvkJpO0nuAddbGM7AVv6KOK1I0O6HHTg05_P4-Da6KKVkp38zFfsacu8/s720/MM-FlagVirtualSphereRotate.webp)
Started with a rectangle select of a fractal. I used mathmap flag chained to animated sphere. This makes a continuous animation. The sphere has an alpha background so it's a virtual sphere.
That makes the front animation. The back animation is the front reversed and made to look like it's on the inside of a sphere with G'MIC Distort lens. (not ideal because the front is alredy distorted)
Gimp has a tool called Script-fu Spinning globe. I checked it out because I'm looking for something that maps an image to the inside of an object. I was surprised to see that it was written in 1998. It makes the back and front similar or identical but I don't understand how it does the back. It feels like it was made in the 1990's
I packed the 90 layers of the back and scaled them to meet-up with the 90 packed layers on the front. I think the back is more of an optical illusion than a properly mapped object.
If anyone knows of a better way of doing something like this I would like to know?
RE: Proof of concept two - Ofnuts - 01-14-2024
spinning-globe is just a bunch of calls to map-object(sphere). This one doesn't map to the inside of the sphere, but if the sphere is transparent you see the back of what is on the opposite side (this is also true of map-object(cylinder)):
[attachment=11017]
If you want to make it look like it is inside, flip it and rotate the sphere by 180° along Y:
[attachment=11018]
RE: Proof of concept two - Tas_mania - 01-17-2024
Thanks ofnuts. You may have been referring to 'ofn-path-to-cylinder'. I'm going to check that out. For now I have resolved this riddle.
![[Image: ProofOfConceptResolved.webp]](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEicEpF2ILnW5SKFnxWe_PjZIHFWX9-Eor9APkSk4rn2tJNHpm8SRFuJ-xSvWB_SGYrVy50u4DGrpECvS_QtlgFXdcioPxcE9wkT5PIkK4iTz8_fgclIOlgkLPRUA_G6645E5QgkgDempu_yIdIQaTYwT1cAwIYuhK1z-gRzhp223DdNuS1zOUSG0J0YSug/w640-h360/ProofOfConceptResolved.webp)
I was overthinking this. The back rotation is just the front reversed because only the sphere is transparent not the animation on it
I scaled the back smaller by 20px but only to the vertical - leave the horizontal the same so it joins-up with the front series. Also applied a GMIC colour blindness filter to the back, to give the back some spatial difference to the front.
While doing this I found the back wasn't a continuation of the front because it was reversed. The fact there is an animation on the moving waves makes this hard to see but it can be solved by this method.
pack the back series of layers > flip horizontally.
unpack and reverse layers > pack and flip horizontally again and unpack > rename layers.
I don't understand why this is but it seems to work
Finally I used ofnuts interleave layers to merge front and back series, they need to have the same number of layers and be the same size.