I won’t lie – 11 to 17 December 2016 was rife with distractions. Rife. Like the English actor. In order of distractiness they were: Oculus Touch controllers; a not-very-good banjo ukulele; 14-year-old video footage that loads into Blender without extra conversion; and lovely relaxing warm weather. But I still got a few things done!
I’ve got a rough draft of 3D visuals on a stinger shot which takes place at a majestic bus stop. I’m already considering ways of re-doing it to make it funnier, including a poster instructing Gronky not to bury the bus stop.
In my head, Gronky’s compulsion to bury things is building up to a backstory where he’s been sent to the desert to bury things and mark where they are… possibly even created to bury things and mark where they are. Not that it needs to be part of the story – it’s vague sub-text and only makes Gronky compulsive about burying things and putting little flags next to them. Maybe I’ll explore Gronky’s situation more in a later story.
I’ve also built the layout version of the robot. It’s a bit crap and low poly at the moment, but it should be fine for layout with a couple more tweaks (like a Floor constraint on the mono-leg). Here’s a sneak peek of the robot flying around.
It is the 12 to 18 June 2016 and off we go. Pointy is very talkative today.
At the last blog I had 55 shots covered in layout. As of right now I’ve got 64 shots covered and 14 more shots to go. Five of those are titles of one sort or another, and nine are narrative shots. Five of the narrative shots use a camera framing that I’ve already established, and after that there’s four shots with framings that I’ve yet to set up.
Let’s talk production management for a bit.
How do I choose which shot to work on next, given that not all of my days have an equal amount of time?
Here’s the shot list for scene 1 as of today. Completed shots are buttery yellow. Shots which already have a complete camera setup in another shot (e.g. all the 01_03 shots) are white. Shots with no camera setup are coloured anything but yellow or white.
Note the one row second from the right with single letters in it. y means “shot’s done”, e means “easy”, t means “tricky”, h means “hard”. (That should really be two rows.)
For weekday evenings after the day job (Mon-Fri), I tend to tackle e shots
where the camera and assets are already in place. On Sundays and public holidays, I tend to do new camera setups or tackle tricky/hard shots. The idea is to create as many easy shots as possible for the coming week so that I can maintain some kind of momentum even when on days when my time and energy are limited.
For example, dialogue shots and simple reaction shots are best left for weekdays. Camera moves, complicated animation, character interaction, sims and anything else with the potential to be a runaway timesink is better left for days when I have the time and energy to stick with it.
For instance, last Sunday I was using particles to roughly simulate flying dirt. You can key a texture that controls Density with a cyclic F-Curve which makes particles emit in discrete puffs instead of flowing at a constant rate.
But in the actual animation file where Gronky’s digging, I could only manage a stream of dirt. What worked in the test file didn’t work where I needed it to. Since it’s only layout, I postponed any final solution to when it’s time to do it for real. It was definitely a timesink. And hence, something better attempted on a weekend where I have time to sink.
So why not stick with the particles and fix them? Again, it’s a question of setting priorities and maintaining momentum.
During this first pass, it’s OK if shots are less than perfect – if sims malfunction, if the action is underpolished, etc. It’s more important right now to get coverage in layout rather than slowing down to nail any given shot on the first try. That means embracing the reality that anything could be redone – even a really nice shot might not work in the edit. Nothing has to be perfect yet. It just needs to be there at all.
Even though I’m only creating one tree at a time, the forest has to look nice too.
Thanks for reading and I’ll see you again next week. 🙂
This week I took Sunday off to catch up on a backlog of movies but I made up for it during the week. I’ve now got more shots covered in layout than not. The end of June target for full layout coverage looks achievable after all. Woo!
This layout feels nicely different to anything I did in 2015. I’m treating it much more like animation
– using stepped/constant keys instead of interpolated keys. This gives it a much rougher yet snappier feel. Hopefully steppy layout will carry through to snappy animation!
It’s a long weekend here in Western Australia. I’ve taken an extra day off from the day job to stretch it into a four-dayer. There’s at least another twelve hours of work to get full layout coverage and I’ve got three free days to play with.
So. Will I smash through the remaining 39 shots like a demon or will I go and look at quolls on my day off? Hopefully both, but find out next week for sure! 🙂
So when I woke up last Sunday, I had 12 shots in layout and no scenes complete. As I write this today, I’ve got 28 shots in layout and I’m one shot short of covering two entire scenes.
I’ve been using the Non-Linear Animation editor a lot. Shot 03_02_A contains dialogue and gesturing as well as a walk cycle where the character is moving across the screen. Even though the NLA can be a pain to work with, managing all of these separate sub-actions in a single action would be even more unpleasant. 🙂
It’s really nice to see Pointy and Gronky moving around again, even if said motion is very low def and provisional.
What’s not so nice is discovering bugs in the rigs. Gronky has a weird one right now where turning his wrist changes the curl of his fingers. Pointy’s face also contorts unexpectedly when the rig is scaled -1. The plan is to press on until I’ve got complete coverage in layout – no point stopping and losing precious momentum to fix minor problems. I’m carefully logging issues to one side and they’ll all get seen to before animation proper begins.
I’ve been playing with my shiny new Oculus Rift a bit too. Did you know that the latest builds of Blender have a “spherical stereo” rendering mode? Did you know this mode lets you render 360 3D videos that are fully compatible with the Oculus 360 Video software (as long as you render out in H.624 and name it correctly)?
I’m not sure how much traction a 360 degree promo video would get, but a short test could be fun to try. 🙂
Thanks for reading and I hope to see you again next week. 🙂
Right. 15 to 21 May 2016. Let’s do this. Sorry for today’s delay but I’ve been waiting for a storm to die down. 🙂
Last week was quiet. On Sunday, I tore through my to-do list for the 2D animatic like a thing possessed and knocked it on the head. This gave me a draft soundtrack and shot list which I was happy with, so it was time to kick off layout again. This means I get to play with the 3D versions of Gronky and.Pointy again too. Yaay!
The layout stage is for figuring out where to put the camera and where
the characters, props and scenery go in front of the camera. Basically I’m composing individual shots and making sure they flow nicely from one to the next while telling the story effectively.
Before I started working on shots, I created a template file to serve as a starting point for the majority of the shots. The screen is set up how I want it, there’s a render of the 2D animatic set as a background image with the soundtrack set up in the VSE, and all the assets are rigged and constrained and ready to animate.
In the template file, I’ve split up the assets like so:
Cameras and light on their own layer
Terrain on its own layer
Pointy and Gronky on their own respective layers with empties following Pointy’s head and Gronky’s hand
Pointy’s hat on its own layer parented to an empty which tracks Pointy’s head using location and rotation constraints
Gronky’s walkie talkie on its own layers parented to an empty which tracks Gronky’s hand through constraints
The plastic kangaroo on its own layer parented to an empty which doesn’t currently track anything
any given layout shot, I can just turn off any layers I don’t need and
get to work. Setting up the template took me a couple of hours but it’ll
save me a lot of repetition.
To start work on a shot, I set the start and end frames to the start and end of the shot in the 2D animatic – the animatic is visible as an overlay through the camera. I place the camera and assets to match the sketch, then I remake the 2D frames in 3D with a few extra bits like you can see above. The 2D version is of course just a starting point.
Once the shot is finished in rough, I boomsmash (playblast) it using OpenGL render, then drop it into the master edit over the 2D version.
I’ve got eighty shots to draft up and eleven of those are roughed and
in. The beginning has been slow going –
mainly fixing the assets and working around the usual health problems.
The current objective right now though is to have a first draft for
every single shot, then fix the problems in a refining pass.
So the movie’s pottering along. Not racing and not dead still or going backwards, but definitely moving forward! The more shots I do, the more I get back into the flow of it. I’m hoping to have layout all done by the end of June in time for a short winter break before I push ahead into animation. We’ll see!
Thanks for reading and I hope I’ll see you again next week!