Update of the Week (19/06/17 - 26/06/17)

Screenplay

I started of the week with outlining and majorly rewriting the screenplay, throwing it into a second draft. I am more confident in this version as I shifted some things around, tuned others and I feel like it provides characters with more motivation and characterisation, without skipping the world building achieved during the elongated scenes from the first draft. I am not sure about the pacing too much, and the little dialogue there is needs a rewrite.

Current version can be viewed here: Wintermute_27062017

Shotlist and Workflow

With screenplay in a much better state, I started thinking about all the shots and drawing out some thumbnails. With everything I want to try mostly established (atleast for now), I decided to craft a workflow diagram, providing at least some idea of the organisation and workflow I will be approaching. Only think that cannot be properly conveyed through the diagram is the iterative approach I want to take. As an idea, and in practice, I hope it will work as a process based on feedback, which, with non-destructive approaches to most of my work could help me iterate on all the ideas and designs all the way to the renders.

Workflow Diagram

FX Tests and picking up new programs

I wrote down a list of FX I want to include in the short, and started testing some of them out. I hope to get to a point, where I could reliably replicate my tests during production of my short without causing delays.

I wanted to start out not too big, because I need to learn the tools properly, so I tried to create sparks made with sword dragged on a floor.

Version 1:

The first idea was to create an emitter from curve, which would be the path of the sword. The results were fairly okay, but I couldn't figure out how to animate the curve, and I came up with other solution.

Version 2:

For this version, I decided to still emit the particles from curve, but make it smaller and parent the curve to the blade. Soon enough I noticed I need to parent the curve to the joint instead, so I did that instead. The result can be witnessed in the video. It looked way better, but I still wasn't entirely happy with it, as curve emitter lacked a lot of options.

Version 3:

For this one, I parented a directional emitter onto the joint and placed it into middle of the blade. I achieved better results, but It still felt lacking. The reason was that all the particles looked like they were following a single line and flow, with not enough randomness.

Version 4:

So far the the version I am most happy with, uses 5 emitters parented to the joint. One is the same one from the version 3, the second one creates more, but shorter particles closer to the blade and ground, third one boasts similar attributes to the first one, but tweaked as to add more randomness, with the last two creating short, semi random outbursts of particles to the side as to stimulate irregularities. They are currently keyed thrice with constant spacing, but that can be manually randomized.

The issue I encountered with the nParticles currently, are my inability to properly shade them in Arnold and render them out correctly. Arnold documentation hasn't provided answer yet, but I will try once again I update to Arnold core 5 from 4 that I am currently using. They are currently rendering out as huge spheres, with my only progress at it so far was to dye them red. 

Arnold Render of the particles

This issue prompted me as well to try and pick up SideFX Houdini. I want to avoid having to render particle effects apart from the whole scene, but I will see what can Houdini offer. The apprentice version is free from their website and there are plenty of tutorials online, to help me start. I will try to create another version of the sparks inside Houdini, and then compare them and see which I like more, and how familiar I am with Houdini for the rest of the project, and then decide if I want to continue with Houdini or use inbuilt Maya tools such as nucleus based systems and FLIP based Bifrost. Besides FX, I can see how I could implement some geometry based procedural tools from Houdini into my pipeline.

Side note: OpenVDB seems rather promising, so I will make sure to take look into that.

In the upcoming week, I will try to look more into houdini with more tests, and start on the shotlist and storyboard.