Hello all, i’m kind of new to realtime animation software it seems. I’m coming from a compositing/motion grapics background (titling/logo animations/greenscreen-compositing etc), a little bit all-round video editor. Never knew about the existence of touchdesigner or other competing software. Last few years i have used alot of blackmagic software, especially fusion. it is in theory very capable of doing what i need,
Only there is the thing that, long story short, render times for very-very simple things often takes so much time, not to mention issues/bugs like leaking cache (yes thats a thing in fusion for most people it seems), that creativity often drops to a point where i can just pull my hair out. And i’m not a beginner/unexperienced animator. I also use adobe after effects, and it is often somewhat faster/more reliable, but i love the node-based way of working. blender is also interesting and fun to use, but not really that suitable for pixel-based manipulations, and the compositing part is not that fast/suitable. Then there is nuke/apple motion/natron, all with their own strenghts. anyway, i have used alot of similar software.
So i thought to myself, it’s 2020, looking at app’s on your phone, things shouldn’t be too hard to be animated in realtime right? (also looking at realtime render-engines like unity/unreal/eevee) So went looking for possible alternatives, googling for realtime motion graphics, and stumbled across touchdesigner. and i couldn’t believe my eyes, everything looked realtime, to a point where it just blew my mind. It looks at least very close to what i see as the ultimate tool for getting creative with pixels. But i just touched the surface, saw some video’s on youtube, so i can’t really say something useful about it yet.
So before i dive completely into the deep, do people also use touchdesigner for non-realtime animations? thinking about logo animations/lower third animations, compositing jobs (tracking, chroma-keying, etc) Are there people coming from a similar background as me, who can share some thoughts about this?
One Thing you have to know about TouchDesigner is that it is more a Toolbox then a Graphixs Software. There is alot of other functionality hidden inside.
Also, not everything is clear on the go or works as you are used to. For alot of things tools just do not exist or are built by the community or have to be built by yourself.
If you really want to do animation only (and no programming, realtime visuals, interface/app design) maybe take a look at Notch instead
oh yes, i’ve also see notch, and it looks amazing. but boy, it is expensive, and subscription-based. I’m not afraid to try and master software though, i could see the advantages of having to build your own tools. if i could eventually get fancy results. I have watched a very long introduction video on youtube for touchdesigner, showing some basic operations, and alot of tools that were used for things that normally a motion graphic designer wouldn’t really make use of, i was baffled by what i potentially could do. notheless, notch does look attractive just watching the showcase…
As @alphamoonbase said, Touchdesigner is more a toolset to build tools. You can do a lot of basics like compositing, animate things or deal with interfaces. But comes a time when you will need to build your own system to generate what you have in mind and eventually use some scripts or shaders if you need very specific things.
Knowing that you can build a whole software with Touchdesigner tells a lot.
It is indeed really powerful though and i am using it for a lot of purpose : from generative / interactive content, previz, realtime/rendered animations, image processing/compositing… But will sometimes need to dive deep to get what you want.
One thing for example : there is no timeline in Touchdesigner. You can manage to use some tools to get close but it is basic tool.
So yes, i couldn’t recommend Touchdesigner more as it is a really powerful and satisfying tool. If you need a real compositing software though, like a realtime Aftereffects, maybe you should also have a look at SMODE
hi, I’m coming from a totally different background, but also trying out a ton of different node-based tools at the moment - one thing I would note is that the real-time-ness depends a lot on what type of complexity you’re working with. for example I was messing with a text based logo made out of particles, and could easily make it out of around 1000 sprites, but replacing them with spheres with more than like 10 polys each quickly took it down to a few frames per second. Maybe some of the game engines you mentioned would be more optimized for things like that, but it depends on the case.
also, if you’re considering Blender, I just learned the other day that there is a very TD-esque add-on for procedural animation called animation nodes: GraphicAll — Blender Community
There are definitely a ton of options with node-based programming right now, I think the most similar ones I’ve seen are vvvv, or cables.gl if you want to work on/export to the web. Plus there’s also Jitter for MAX/MSP or GEM for Pure Data, both music environments which let you patch visuals too.
TouchDesigner does have a timeline and you can do sophisticated keyframing with the Animation COMP. In fact you can have nested timelines that would mirror the way pre-comps work in AfterEffects. But turning TouchDesigner into AfterEffects would be a lot of work. The forum has some examples headed in that direction.
Still, don’t overlook what you can do with the Animation COMP, the Timer CHOP, and some Switch CHOPs. For work I was given pre-rendered video, and I animated in detail ~6 minutes worth of ~8 custom parameters with the animation COMP, and those parameters influenced an interactive system synchronized with the pre-rendered video.
Also, if you’re at 1-10 fps, consider making a “playblast” toggle that helps you render at 60 fps but doesn’t affect the “simulation” aspect of the system. playblast - Wiktionary
good idea, have done that by hand. you’re just making some kind of switch toggle connected to all the detail levels and resolutions?
that sounds good for most cases though I could see it possibly getting tricky with physics simulations - I posted here before about collisions being kind of buggy - maybe that’s not the best thing to do in TD anyway though
Yes, I said 6 minutes of animation but it was actually more like multiple 30 second scenes sharing a switch CHOP and each scene had its own parameters.
If your goal is non-real-time, then resolution can be sacrificed during workflow too. Physics accuracy is harder to sacrifice but could be reasonably tuned down with a toggle. But speaking off the cuff like I normally do I think running a Flex particle simulation cranked for accuracy and with 2 million points isn’t as bad as rendering those points as 100+ vertex spheres in a Render TOP.
