|
|||
Plato's Shadows--"Z'ha'dum"--3D Fractal Animation Music Video | |||
Download File | |||
Previous Image | Next Image | |||
Description: My most ambitious fractal video ever! This took 2.5 weeks on 3 video cards (2 980s, 1 1080) to render. 6800 frames at 4K resolution. The Shadows of Z'ha'dum* are thought given form in the reflections of the minds' eye, cast from the fires of Plato's Caves. Music: Electronic industrial metal music with crunchy guitars, a fat analog synth lead, and arpeggio'ed bass. By Stephen Sink. 4K Video: Made with Mandelbulber v 1.21 and mclarekin-ported formula "SpinV_ABox-Kali." All video and music by Stephen Sink and Paigan Productions. *Z'ha'dum is a Babylon 5 reference. Stats: Views: 521 Total Favorities: 0 View Who Favorited Filesize: 15.54kB Height: 128 Width: 128 Discussion Topic: View Topic Keywords: SpinV_AboxKali Posted by: Paigan0 ![]() Rating: ![]() ![]() ![]() ![]() ![]() Total Likes: 1 |
|||
0 Members and 1 Guest are viewing this picture. |
Tas_mania | October 11, 2017, 11:40:07 PM Very cool! I enjoyed the graphics and music. Also, thanks for the detailed description. Looks like OpenCl will be a part of 3D fractal rendering for a long time to come. I'm amazed you use the 'Creative Commons' license on your videos. I hadn't noticed that before. That's an innovative and generous road to take. I'm wondering if you have seen the 'Zoomquilt' videos? https://www.youtube.com/watch?v=INqKqk-ifpc They are not fractals but have an interesting infinite zoom effect. They were made collaboratively about 6 years ago. |
Paigan0 | October 12, 2017, 01:47:05 AM That Zoomquilt movie was one of the most awesome things I've seen in a bit! I so want to do that! Any idea how it's done? And thanks for the comments! |
mclarekin | October 12, 2017, 06:07:35 AM cool steve, maybe your best so far! zoomquilt! You can do this now in v2.11_1, but it requires an amount of human input. I have suggested automating this with Buddhi about this a while ago . Now with the speed of openCL, this feature has has become more practical. In my simple mind it does not look complicated tin automating rendering an animation inside an animation inside an animation ... etc Like, we can insert an imageB into an imageA : http://www.fractalforums.com/index.php?action=gallery;sa=view;id=20529 so what we want is, that when rendering frame imageA.001 then we are inserting imageB.0001 from the folder of imageB rendered frames. If zooming in from imageA into imageB, then as soon as none of imageA is in view then we have finished with rendering imageA frames. It needs a backwards approach so if having three levels you must render the deepest animation first. Image C frames are inserted into image B frames, then Images B frames are then inserted into image A frames. If you get what I mean. ![]() |
Tas_mania | October 12, 2017, 06:13:53 AM Hi Paigan0, somebody called 'iKnitYogurt' has an explanation here that seems plausible. https://www.reddit.com/r/explainlikeimfive/comments/35v7os/eli5how_does_an_image_like_zoomquiltorg_actually/ So it could be images with an alpha channel (semi-transparent) zoomed in to gradually reveal a new image also with an alpha channel. Maybe it was done backwards and then the clip was reversed? In my limited experience 'frames' and 'layers' are interchangeable in graphics and video editing. ![]() |
claude | October 12, 2017, 08:34:58 PM I guess it works like the Kalles Fraktaler Movie Maker (and my own zoom assembler): fading between adjacent key frames while zooming in does lead to slightly "strobing level of detail", in that you can sort of tell where the next keyframe is, because the central rectangle is more detailed perhaps that could be ameliorated by using exponential map representation of the zoom video? http://www.mrob.com/pub/muency/exponentialmap.html |
Paigan0 | October 12, 2017, 11:48:10 PM http://www.videocopilot.net/tutorials/earth_zoom/ This tutorial for an "Earth Zoom" might just be the way to do it. https://www.youtube.com/watch?v=5aWwoPxF1zg (although this is not for completely different images, which I'd like to do. But this gets me closer. I have After Effects!) Last modified by: Paigan0 October 12, 2017, 11:49:40 PM |
Tas_mania | October 14, 2017, 12:58:05 AM That Earth Zoom tutorial was interesting. The gag about hot air balloons and fishing line must have fooled some people. ![]() I use mostly open source software so it's good to see how after effects does it. In gimp there are no 'parent layers' but layers can be 'linked' and manipulated as blocks. Here is an example of perspective applied to linked layers. https://west-tamar-talk.blogspot.com.au/2017/07/a-visit-to-animation-gallery.html I think in the zoomquilt video there are often visual elements close to or around where the next image is being introduced. I think this is to get people to look into the zoom but not notice the transition to a new layer. (zoomquilt in slow mo to see how it's done) It's a mix of editing techniques that includes the psychological element in my opinion. It's impressive your 5 minute video took 2 1/2 weeks to render. Your viewers are getting a massive amount of CPU cycles and server-side optimization with music thrown in as well. |
Paigan0 | October 16, 2017, 03:33:48 PM At 2K views this morning over the weekend, this is my most popular video in just a few days. Rock on! |
Sabine62 | September 03, 2018, 12:33:22 PM Thanks a lot, Stephen, very much enjoyed the ride, both fractal and musical ![]() |