Hi. I’m a veteran modder with minor Unreal experience (mostly materials, lighting and particles). I’m looking to push a very large-scale, private machinima-esque project into Unreal 4. However, my initial research raises a lot of difficult questions I don’t know how to answer at this point in time, so I’m looking for some input, hopefully from someone with experience in the subject.
First and foremost, I’m not an artist. I’m working mostly with third party assets. I can mangle stuff somewhat and know how to use plugins like FumeFX to make textures for particles, but I can’t animate or anything myself. I know how to make inorganic static meshes, simple vehicles, and other kinds of props, but that’s it. I have a lot of experience porting graphics and audio and remastering them, but actually editing them like rigging or animation are entirely beyond me. This is the major crutch of the project and will probably lead to a lot of pain, as it appears Matinee was intended to be a front-end for compositing pre-recorded/mocapped scenes and not for stringing existing animations together to make a scene. I do audio and writing primarily, so I’ll have to figure out how to make ends meet. To that end…
1.) Matinee doesn’t support animation blueprints it seems. I cannot send animation commands to an actor using an animation blueprint. I rely on animation blueprints to weight certain bones to each other between character meshes with components that have different skeletons and also for syncing the body parts (though I have to manually make the same animation blueprint information between every skeleton, greatly increasing workload…). I’ve read that there’s a Matinee update in the works. Will this address the lack of animation blueprint support? Or will I have to somehow modify the assets at the source level to fuse their skeletons? (I have no idea how I’d do this, but it seems it may be mandatory to proceed.)
2.) How would you manage a very large scene with a few hundred thousand turrets on starships? I asked this question back in UDK era and was told I’d need uscript. I’m not a programmer, so I can’t figure that one out. Is there no way to manually control bones or pieces of a mesh in Matinee? If there is, I can’t find it or any information about it.
3.) What would be the ideal way to manage a large scene in general? Matinee gets super messy in a hurry with only a few animation tracks. I can’t imagine multitasking a massive fleet battle with hundreds of thousands of particles to manually keyframe on top of it, or a scene with dozens of characters fighting. I have experience with Sony Vegas, whose interface is quite a few steps ahead of Matinee, but even that got really messy making a 3 hour audiobook in with 9 tracks.
4.) How exactly does Root Motion interact with Matinee? A lot of animations I’d be pulling are built for root motion. Hypothetically, I could do basic starship maneuvering as an animation in 3ds max and use root motion with it for maximum control in certain, specific scenes. But does it play nice with preview et all?
5.) Without the skills to animate faces whatsoever, what would you recommend as a visual means to convey character speech and emotion?
6.) I’ve looked a bit at Nvidia hair in terms of video tutorials, but it seems rather complex. How difficult would it be to pick this up, and do I require any commercial software aside from 3ds max to make good use of it? I’m looking to use it to spruce up third party assets; adding fur for beasts, alien terrain objects, etc.
7.) Will supersampling apply to rendered frames from Matinee? I’m not a fan of the huge quality loss from Temporal AA, so I’m hoping to supersample frames to get rid of aliasing. On the same note, is there a way to make it dump frames somewhere other than the project’s directory? I have a few terabytes worth of recording hard drives but the project itself is on a data drive.
8.) Disabling Temporal AA removes a chunk of reflections. This has been posted about a few times, but I don’t think Epic said they would fix it, or least I haven’t seen it. Will you fix it?
My full pipeline is intended to be to build the visuals in Unreal doing whatever necessary to produce the scenes I need. I would then convert the frames of those scenes to a storage format using Vegas. Once I begin actual compositing, Vegas would be used to composite all audio alongside the final video pieces. All post processing, of what little I would use, would be done via Unreal itself, as it seems more than powerful enough to do the few effects I would desire out of a video. I then would encode the videos to my standard x264 settings via frameserver to megui. I am accustomed to projects of large scale, so the workload itself is not necessarily the problem, mostly my learning disabilities and lack of graphics skills. This would be nearly exclusively a solo project other than the few external voice actors I would be outsourcing to. Seems rather hopeless, but I’m accustomed to that, too.
