Image Generator comp not animating in Vuo 2.0

Going through some old comps with Vuo 2.0. This comp has meshes that animate in 1.2.8 but produces a still screen in Vuo 2.0 beta4 (version How can I fix it?

This setup with published ports is somewhat typical of a bunch of comps I have. Any tips for Vuo 2.0?

lissajous_whitney0.vuo (26.8 KB)


in Vuo 1.2.8, when an event triggers an input port of a subcomposition, it flow through all the input ports. In Vuo 2.0, it flow only through that input port. If you run side by side your comp in 2.0 and 1.2.8 with “show event” activated, you’ll have clear how this impacts on it. In 1.2.8, you can clearly see that all nodes are continuously triggered. In 2.0, you clearly see that only ones connected to “time” are triggered, and whole parts of the composition never run.

I tried connecting with an event cable “time” to other parts of the composition - quite randomy, without really dig into what it does as it’s not that simple. This way i got the comp to behave more similar to the original one - animates, even if perhaps at a lower fps. I bet that reasoning better about what nodes of the comp should really run every time the comp is triggered, will drive you to the result.


jersmi.lissajousWhitney0.vuo (28.5 KB)


Thank you, that’s helpful, I’ll dig in. So far with this comp, having the Vuo 2 Comp Loader open with Show Events selected is crashing the 2beta4 editor after 20-30 secs…

Ps. no love trying to use an iphone to create this response.  

So far with this comp, having the Vuo 2 Comp Loader open with Show Events selected is crashing the 2beta4 editor after 20-30 secs…

Created a bug report: Crash after running for ~30 seconds with Show Events enabled

Reworked this “test version” seeking to optimize and explore Vuo 2.0 features. Simply connecting the Image Generator “time” output to Build List started up the comp/animation, that was easy. Still wrapping my head around 2.0 event flow.

Regarding substantial performance issues, Build List (obviously) performs poorly constantly pumping out torus objects. (Why use a 3D torus to make a 2D polygon? Best way I found in Vuo to have nice corners on polygons, that’s it).

This reworked version drastically improves performance (no surprise) by converting the torus/polygon into an image shader for Make Points Object, then only using Build List to generate the list of positions for Make Points Object. Still working out how to have a couple features I enjoyed from the original (manipulating size of points, etc.)

Any other tips/new Vuo 2.0 methods to optimize this comp? Is there a way to dump or optimize Build List (the bottleneck, currently set to Drop Events)? Is there another way to generate/animate in real time a point list (i.e., list of positions)?  

lissajous_whitney02-points.vuo (29.4 KB)

I guess the best would be to calculate the list instead of build list.
And build it once, then retrieve a part of it instead of always recreating it, that way, I’m getting 60 FPS, where I got about 12 FPS with your comp.
If you know the parametric formula you could also use Make Parametric Points.

Here joined are 4 comps, all building the lists once, and retrieving them using Magneson’s List Custom Nodes (see node gallery) :

1 - Your build list system.
2 - Using Calculate List, simple version.
3 - Using Calculate List too, but with noise also.
4 - Same but with squares instead of polygons, as I liked your first upload.

You will need to add z-positions math back, as well as the image filters, removed them for debugging. Also I find the blur nodes to be quite computing heavy. Need to reduce the image size too much, at least on my 2012 iMac.

Technically one could use stock nodes only, retrieve one point and enqueue them, but I’m not being able to use Make Points Object with one position only (Bug Report Link)

Hope this helps.

Capture d’écran 2020-04-02 à 00.57.21.png
Capture d’écran 2020-04-02 à 00.59.21.png (48.7 KB)