Vision Clip Mixing

Has anyone created a virtual vision mixer using vvvv? I saw some info on Catweasels work and I have been thinking of working on something along those ways in my continous goal to disconnect from VJ apps. The main objective would be to make a A/B/C Source with multiple Visual Samples. I like Tonfilms switching system too, just doing some thinking and brainstorming.

I’m working on a VJ mixing system for VVVV-patches. I’m creating it for my dissertation. When I’ve presented it in the university I’ll put it on the site for free. At the moment I have problems with the annoying short freezing when you change a patch. If somebody helps me to find a workaround for opening patches without freezing, it would be gorgeous!

i tend to have all patches in memory and disable what i don’t use.

so if you have a MIDIfaderbox connected to the Alphapin in this example patch, the patch is only activated when you really use it.

disable.v4p (4.6 kB)

Good idea! Thank you kalle

I am doing something similar, but I was disabling directly the renderer - If I understood correctly, just disabling a quad on which the renderer is connected as a texture is enough?

Would that work if I were to use a shader with multiple texture inputs and a selector input to choose which one of those textures would be used in the pixelshader output?

r.

you are right:
you can create a spread of textures and use GetSlice (Node) to switch between them.
that is usually quite performant.

The solutions works pretty well for 5 or 6 patches; anything more and my computer starts crying

(Got 2 fps when I loaded 31 patches at the same time and my machine crashed heroically after a few seconds of rendering :D)

If anyone wants to experiment with this, I made a little c# program that might help.

The idea of the program is that it scans your switcher Patches folder, and creates a new patch (called AutoGenerated-Switcher.v4p”) with all the patches in your Patches folder connected to a GetSlice. It also routes the Audio Analysis, Control and Screen Size inputs to each and every one of your patches, so you only have to place it in a patch and mix it with a blend (it generates both the A and B channel outputs)

To use it: download it, compile it (or download the .exe in the next post) and then put the .exe in your switcher folder (or, more cautiously, in a new folder where you put a copy of your Patches folder – the folder must be named Patches). Once you open the autogenerated patch it might take a while to load (on the extreme case, with 31 patches (10 different videos, at least 5 with medium density 3d and xmeshes, and lots of other stuff) it took almost 4 minutes to place the node using “Open in Patch”.

It is a little brute force the approach in the visual studio project, but I didn’t want to spend much time on this as it was just a proof of concept (which didn’t work, at least in my case – I’m also building my own svvvvitcher, and also was looking for a way to reduce or eliminate the pause between the patches loading)

PatchSwitcherCreator-VSProject.rar (27.2 kB)

and the runtime here

PatchSwitcherCreator-Runtime.rar (5.7 kB)