All this looks really interesting and well done, IMO.
My only criticism would be that the environment of the concept sketches is a lot better than the black dark background in the renderings, but I suppose you will update that for the final renderings.
So, if I can ask, how did you mapped the animated mesh, in the end? I mean, after the boolean and remesh modifiers (not UV map, I suppose)?
This is composite render. Later I will render in passes and produce final images in postproduction, with dust, smoke, light effects…
I made a single mesh with UV map, which I separate on body, arms and legs. These bodyparts share the same material and UV map.
I gave up from using Remesh, because I cannot achieve realistic materials without UV mapping. I do use Boolean, but only for animation of cutting arms and legs in particular movement phases. But, because Boolean cannot be limited to vertex groups, its needed to cut body into parts.
Latest test for first two scenes:
I liked the idea. But I do not understand what’s going on with the left arm from 1,35 to 1,50 seconds . Toes touch the heel when walking. It looks very strange. P.S. sorry my English, please. I hope you understood me
I don’t mind about Your English, mine is probably even worse.
Everything about this project looks very strange. I want to have action poses like sculptural stylisation, with cutoff arms or legs, with reduction of body parts which aren’t so much important. It’s a challenge to find good balance between expectation of a viewer and experiment in animation, between communication and artistic model. Missing arm isn’t some kind of bug, it’s an opportunity to show additional body contour in the movement… or maybe… I’m just out of my mind and making complete nonsense… :spin:
I like it too.
Some thoughts:
Maybe there are other ways to make a distinction between actual sculpture forms and animated representations so that it would be more clear to the viewer. Perhaps less adding and removing body parts, replaced with one clear effect that makes sense. For example, the sculpture that walks needs to have arms at the cut for the transition, so maybe it could start without the arms, especially if the actual sculpture doesn’t have them. Could make them appear before the walking part is over.
It would simplify the animation but maybe that could be replaced with additional shots that may have a switch between removed and visible body parts, showing some aspect of the model and still having the effect make sense to the viewer, or at least make it clear it’s not a glitch.
After the cut the arms need to be there to make the transition and to show the function, but maybe the group could consist of ones with arms who wash themselves and ones without.
The transition at the cut could be clearer. Cut when the hands are halfway and have the movement continue after the cut. It’d be like: stading up -> going down -> halfway -> cut -> going down -> there.
Thank You so much for your critics and suggestions. I have waited for this kind of conversation. In next couple of days I’ll set whole animation in preview state and then come back to fix problems.
First experience with arms and legs disappearing is a feel of some kind of a glitch. This is for sure biggest issue I have here. Thanks for ideas how to fix it
I finally manage to came to the last frame with initial setup:
Hard to give useful critique in this thread 
That is meant as a positive comment, what you’re doing is unique, complex and good enough to make it hard.
Some thoughts again,
I like how the key points in the audio matches things that happen in the video. If you’re stopping the animation just to wait those though, maybe there is a way around that by mixing the audio. Could use the less busy bits in it as margins perhaps.
The title might not be prominent enough. It’s the title and what the clip is about so maybe that would need to be more clear to the viewer at the beginning. The strings in the audio start from frames where there’s no interest point in the video. Could maybe experiment with varying camera movement speed to get the interest points in front of viewers’ eyes quicker. Constant camera movement is a bit boring, at least for less interesting things.
I don’t mean making the whole clip hectic, it’s not a action movie with Scarlett Johansson in it, just maybe increasing and decreasing speed to make the interest points in the video and audio match better, to have the viewer wait less for the interesting parts, or to have it move in a better position for a key event.
The bird sequence looks quite strange. The character movements might work but obstructing the bird from the viewer feels odd. The camera moves at a constant speed again and the viewer has to stay with it. When the bird comes down between the two characters, that place is where the viewer has been led to look at, wants to look at, but can’t see it because the camera moves behind the character. It’s a bit disappointing ride in that sequence.
Thanks, JA12. These are really useful critics. There is nothing You said that I can disagree with.
Music is painful topic at the moment. This is only a sketch made from Audiomachine tracks to markup the mood. Because it’s a temporally solution, I didn’t pay enough attention to matching sound and video.
Title really should be more noticeable.
Camera really could be done better, as You propose.
Thanks a lot, I appreciate it.
Here is latest animation test for Venus of Jale, in stereoscopy, with final music. Final music is a song “In the Water”, by Isidora Zebeljan.
I would like to make in next iteration better DOF, some motion-blur and atmospheric effects and still need to fix some issues with stereoscopy.
NOTICE: Youtube Player has an option for playing a 3D video in various stereoscopic formats, or in 2D version. These settings are available on the gear icon.
2D version:
Big improvement. Congratulations on being so close finalizing a mountain of tasks with this. I think you made it work, especially those transitions which went from weird glitch - look to self-explanatory to the viewer, and they support storytelling between sequences in the animation.
This is looking fantastic!
This is one of the most interesting projects on Blender Artists at the moment - can we get it in the gallery when it’s finished?
Can we have a non-stereoscopic version?
- 1, really one of the best projects related to Blender. Is it BI or Cycles?
NOTICE: Youtube Player has an option for playing a 3D video in various stereoscopic formats, or in 2D version. These settings are available on the gear icon.
Thank You so much for these positive reactions. It’s such encouragement for me, to endurance with this project.
It’s rendered with Cycles.
I must admit it’s quite exhausting working on this kind of a production. It’s really push the computers and nerves to the limit.
Maybe You guys have some advice for issues I have. This test animation is done in passes. All effects, DoF, Motion Blur, etc. were planned for post-production with these passes. But I hit the wall with DoF and Motion Blur. I believe issues I have are consequence of using scenes with linked objects and Cycles.
DoF is crappy when added with Defocus Node, especiallly on overlapping contours and point of focus in animated sequence. Only flawless results I had with rendered DoF. For final render I’ll for sure go with rendered DoF.
Motion Blur is crappy when added with Vector Blur Node, with blur limited only on some random squares over image. Motion Blur is crappy even when rendered, with some long blured lines all over the image. My plan B is to ask someone with After Effects to run Force Motion Blur on final images.
I’m would like to have some smokes in the scene, but smokes which goes through geometry would look stupid, because linked objects in simulations are out of the question.
It’s looks like 3D option for Youtube is missing in some browsers. I’ll publish 2D version tomorrow.
2D version: