Beauty Approximations

Dusting off the files. Reconnecting all the wires. After a year of silence and rendering has recommenced. 

Starting with this shot , there will be a lot of these fluids so I had to make the setup completely procedural. I now basically just build the shell of the target and automatically get the animation.

Moving on. Pretty excited about this one as the liquid will reapear. 

350 shot complete . Added background image from a photo of mine, a Chicago thunderstorm at sunset I took from a high building. Think I got enough materials from that evening to patch together what I might need. Also spiced up the smoke element as I felt the shot needed something more. Latest cut.

Virtually done with 330 (see top image). Starting the following shot which is basically the same thing but with camera lower and further out. From this I also attached a  before/after of the manual resculpting I do to every background model in this segment. Most of my geometry is done procedurally but this is an important step to add a more organic feeling to the scenes (it's done in 'zbrush'). Got a fairly standard approach to these shots now so most of my time is spent on waiting for renders.

Base for next shot, done for now. Big chunk in the middle will start twitch, an edge crumble, let out steam and some goo will start sipping through. So just like the previous shout but a bit further in progression and clearer. 

Added smoke element. Maybe I need some more dust, but might move on. Anyway,  it's a short one. Also updated the edit . Starting to see I've got some grading issues, might be an area where I should accept my shortcomings and consult an expert. 

As the film got to the outside of the crystals there's been a lot of hesitation on the next step. Think I'm getting closer though, got this simpler solution with this shader and chaotic mesh (in this case, boxes on a sphere). The idea is to create drip shapes from which boxes starting to poke out, gradually forming the crystals. 

Working on a lsystem element that will work as a closeup detail to the bigger crystal growths. I've used wheat references and it's pretty close to those but I think I want the growth to be planar so I can place it on arbitrary surfaces. Starting to work in shorter intervalls and then casually plan and meditate over the next step. I think this works well with my stressed scheduele and the testing phase the project is in.

As the goo will sipp out, I want its movements to be driven by a irregular, kinda creepy force. This is a test for something that will be a underlying skeleton in the goo. I think I'll put some twitching movements in it and then start working on it's surface. This is made by an Lsystem with an animated angle function and gravity force applied in the rules. 

I'm not too happy with the look of the "blob" in the shots I've done so far, it lacks interesting details. So to get more complex movements I started to play around with liquids with variable viscosity . You can hint more rigid chunks in there.

Now where was I... A lot of distractions these days- both on and off screen. But I squeezed out another shot and updated the cut .    In the thinking department I'm working on a stop motion look on the crystal formation. Might do a concept test of using swimming displacements textures and non-consistently solved dynamics to get a non linear looking animation. 

Now got a first version for all liquid shots above ground. I'm quite happy with the look in a few of them, like the one in the still to the left . Here's the latest(200MB) . After this the liquid will trickle through the ground and start gaining some color.

Per request I've uploaded the latest version of the shed video (200MB). www.bhenriksson.se/projects/shed/shed_seq001_v017.mov

My friend and former college Simon Gustafsson has made this lovely still of the tree structure. I needed a more trained eye to nail design and composition as starting point for my 3d model. It's a lesson from previous project- I need to outsource vital decisions to specialists. You might say I need to understand my weeknesses rather then to try to become the complete generalist. Would love to get back to work on this project right away but that short film got a deadline in February.

Liquid expansion overview (missing 30 frames). Feel I want to change a few things here, a clearer light coming from the energy blob outside top screen, the speed may be too fast and the liquid comes out from nowhere (which is fine but that 'nowhere' is in frame). I may also need to establish that the colour of the energy inside changes during these shots but I think I'll get them in decent shape and look at them together before making that call. So I'll probably just go to the next one.

Previz. Goo grows outwards and sips down underground, last frame is a cut through of the underground. I reckon I can throw out the overground shots fairly quickly but should move carefully to reach a nice underground look. I vision some kind of tunnel system where you see the light spread and bounce around as the goo moves down. At the same time the liquid becomes less frantic and the overexposed colours cools down to more saturated colours. Finally it comes out on the other side in and starts to build the crystal. I might need to start move the camera at some point as I think that from the camera and forward I got a free camera to emphasize all the space we got as we're now outside.

Two shots with the new goo element.

After a standstill on the liquid to solid development I decided to go back a step and work on the liquid. Maby if I get a clearer view of how this should look I will get some hints to how the crystal formation should happen. The sample is the liquid with the floor and some sparkling particles visible in the refractions. Next up is to mix it into the blob.

Dynamically creating wires . Next step is to make them join to each other as they're getting close. Later on their trail will create meshes that is the inside of the crystal. Update: did a quick mesh test, a lot of intersections but I feel that there is some niceness in there. Might wait implementing the join function until I see this animated and with a more continuos surface.