I agree that the chargeable Ready-to-Run agents in Massive have a better starting point than Golaem’s free to use Entities.
However, Massive’s fuzzy-logic state machine, is similarly limited by it’s range of mocap, which is specific to the agents that you buy.
Having dealt with new mocap, for EVERY film production that I’ve worked on, I’ve found Golaem’s ingestion tools far more user friendly. User’s can mirror actions; processing of loops and footdowns is intuitive; and the best feature of all is Motion Mapping: http://golaem.com/content/doc/golaem-crowd-documentation/motion-mapping-gmm
No more retargeting of whole libraries every time the rig changes, or we want to reuse the same actions on multiple character skeletons!
I’ve also found the rigging part of Golaem less constrained by the software: rigs for Massive are always quite different from the ones used for a typical animation production.
Good to know that Massive is taking training a little more seriously. To prompt that further, I’d point towards Golaem’s approach: offer free learning licenses, to download, and have comprehensive video and searchable documentation online: http://golaem.com/content/doc/video-tutorials
I would love to know, in more detail, how you got into the situation of baking out geometry from Golaem for farm-rendering, instead of using the license free render procedural? Was it for an unsupported renderer?
Also, Golaem agents take whatever shaders are assigned in Maya. You can choose to use the optional supplied switch shading network nodes, a they do provide decent functionality. I’m surprised that you would expect similar shading networks to work with different renderers — have you tried to do any of this with Massive?
In my experience of using Massive for 3 years — we could never fully replicate the lookdev we were trying to assign in Maya Vray, using the supplied procedural. We wrote from scratch, a whole new rendering pipeline to circumvent the impossible way that an artist’s lookdev is translated into Massive script. This is not an isolated case, it happens in most studios, if you want to try creating lookdev in another package and assigning to your Massive caches, then be prepared for a lot of undocumented fiddling around!
One way that I think Massive suceeds, is in having ascii file formats that are easily scripted. This is it’s saving grace, when bringing through to a more adaptable rendering pipeline.
With regards to Golaem’s Layout Tool, I should’ve added this link: https://www.youtube.com/watch?v=6exmgU4rl5I
Obviously the other video had some marketing guff, which you rightly called out. The Layout Tool should never be regarded as a way to produce proper simulation of crowd agents. This is, of course, just as easy with the regular simulation tools.
It is however, the BEST artist-friendly PREVIS and SHOT-FINALING crowd tool I’ve ever used. Having been hacking around with Massive for years, I’m sure you’ll agree that when the Supervisors start art-directing crowds, you will have to come up with your own tools and scripts to get the desired effect. Going back to sim, especially for really huge crowds, just to update some terrain or a single agent’s costume or pose, is a huge pain. Not so in Golaem.
So in summary: “The lure of AI”
Massive is a very cool simulation tool, has decent ready-to-run agents and formats that are easily scriptable.
It loads and runs very fast, which is necessary as it crashes frequently. There are difficulties using custom assets, especially shaders. Also every update needs to be re-simmed: simple cache edits, character/terrain updates — it creates a lot more work.
Golaem have nailed the artist friendly approach, and fits perfectly in a Maya-centric pipeline. Yes it has bugs, but the team behind it have been excellent at fixing them and responding to any queries.
If you absolutely must have AI, however painful that may be to create for a VFX production, then I’d advise waiting a few more weeks for Golaem v6! Having had the pleasure to beta test it (car traffic especially), we are confident that this software will see us through all future productions.