Projection Mapping from Rhino / GH

Replies to This Discussion

Permalink Reply by Armin Seltz on November 2, 2015 at 10:52am

I have a bit of experience in this and would probably not go with GH even though it might be tempting. The performance is not geared towards real-time performance and visual output is pretty restricted.

Have a look at vvvv. Its an amazing toolset and works very similar to Grasshopper. The same basic principles apply there and if you are comfortable in one, you will get the hang of the other quickly.

VVVV is totally geared towards realtime performance and can be made to do pretty much anything you can think of. Works well with multiple outputs, for example using Matrox Triple Head 2 Go. Obviously Syphon/Modul8/Madmapper only work on the Mac and GH only on Windows, so getting live-video from one to the other will be a bit of a challange. I know Resolume works on Windows and you can run vvvv Patches directly inside it using a framework called Sprout (similar to Syphon).

If for some reason you want to control GH from vvvv (and send data back), that is possible using OSC, which even works over the network if necessary.

Over at http://vvvv.org/forums you'll find the vvvv forums, which are very active and if you pose your question there or even look for stuff on there, you will find things that will be very close to what you are trying to achieve, because its one of the classic use-cases for vvvv.

Don't be put off by vvvv's minimalistic user-interface. It will all make sense with time and in some ways is superior to GH. If you want to get into programming real-time video stuff, vvvv is the way to go as it can connect with/handle pretty much anything you throw at it, whereas GH is really for 2D/3D-based stuff and Firefly is a bit of an exception in terms of what it enables GH to do.

Permalink Reply by Armin Seltz on November 2, 2015 at 10:58am

http://resolume.com/blog/10341/vvvv-to-resolume-bridge-is-ffffantastic

Here is the bit about using vvvv with Resolume.

Permalink Reply by Matt Heinzler on November 7, 2015 at 1:28pm

Thank you for introducing me to vvvv. I recall seeing this interface in an interesting VR project posted on CAN (http://www.creativeapplications.net/maxmsp/in-the-eyes-of-the-anima...).

It seems like an impressive toolset with much more compatibility than GH for this type of goal. Would I be able to run this reasonably well through parallels?

Really appreciate the thorough reference.

Permalink Reply by Armin Seltz on November 10, 2015 at 3:29am

No problem, always glad to introduce someone to these things. I have used vvvv lot a while ago and made this nice realtime geometry synthesizer, an early version of which was shown in a TEDx Talk (https://www.youtube.com/watch?v=uxYHlZQSADQ) - thats me doing the demo at around the 8-minute-mark :) 

That was a truly gigantuan patch in the end, but its nice because its still in real-time. In the newer version there is much better rendering and a lot more options.

So if you have any questions or need some starting points you can always send me a message.

As for running in Parallels - unfortunately I don't think that will work too well. I have actually never tried it, but for anything realtime you want all the power available, so you have to run it in native Windows. I recommend Windows 7. I use Boot Camp on different Macs and it works really well that way. Mind you it doesn't like retina screens too much for the User Interface.

Be prepared for a bit of a learning curve, especially in the field of visual output, because unlike in GH, you have to actually build the whole render process. But there is many good examples and stuff in the forum on vvvv.org.

Maybe you know this book called "Generative Design" (http://www.amazon.de/Generative-Design-Visualize-Processing-Bohnack...), which is amazing, but designed for processing. Get this book, because its amazing and there are vvvv versions of most of the things in there!

If you speak german or probably even if not, then there is a great book called "Prototyping Interfaces", which is the only book about vvvv and shows a lot of great examples, which you can download.

In vvvv itsself in the addons (called girlpower), there is a ton of examples and you can press F1 on any component and it will open a help patch that shows you what it does and how it works.

Lastly I would recommend you print out the keyboard shortcuts cheat sheet and have it handy for reference. There is a ton of shortcuts and it will take a while till you know them because they are pretty obscure, but the more you know, the more fun it is to work with.

Permalink Reply by Matt Heinzler on November 12, 2015 at 12:31am

I really like what you guys are doing. Compelling stuff.. I'd love to see any progress made, especially with improved rendering- though I think the shapes you're generating are still very interesting.

Again thank you sincerely for turning me onto this software and additional resources. I see this type of work as the direction my design is going so I'll be referring to this discussion. And a huge thanks for being wiling to answer a few questions / give me starting points!

Permalink Reply by Armin Seltz on November 12, 2015 at 3:16am

Thank you. Well there isn't any published records of the continuation of that work, because for me its the other way round. I started with vvvv and now more and more use Grasshopper, because we are more and more integrating our work into architecture. Since I am doing all this for the company I work for, its all a bit secret until ready to be shown unfortunately.

There was a version where its running on 2 computers - 1 does all the rendering and 1 is just to control it using a 27" touchscreen.

Then there was also a way you could export the animation you are creating as a Maya file for rendering, which was fun. I actually had to reverse engineer a maya animation file, but it is in plain text so it was just a bit tedious, but it worked and you could then render it properly in the incredible Octane Render Engine (http://www.seltzdesign.com/blog/octane-render-lives-again). Time constraints were a factor, so there is a lot of work that lies dormant right now, but hopefully gets resurrected at some point.

I am not so active in the vvvv forum any more, because I'm not using it so much any more, but if you are stuck or need some pointers you can always send me a message here.

Permalink Reply by Tomaas on July 24, 2017 at 2:52am

 I also wonder if I can create similar video mapping with this program?