|
Interactive Mesh Handles (Composition by gtoledo3)
This is a composition I thought I posted here a long time ago, but maybe I just posted it to the QC-dev list, or my blog (?). In any event, it shows a way of using the interaction patch and spooky to manipulate the vertices/uv's of a mesh. I thought it might be a handy, non-obvious kind of setup to check out. Since this, I've developed some higher-rez approaches by totally avoiding the iterator/spooky, but this has a certain immediacy and some aspects that I like. |
NIce One! Never new that Interaction patch functionality, learning something every day
I don't know if it is intended functionality at all.... I almost think of it as iterator func!
It kind of came out of a back and forth with cwright where somehow topic drifted to iterators and "wouldn't it be interesting if iterators that can't publish outs, still are doing the same queue functionality inside, and somehow, that could be sent out". As he usually does (for good!), there was probably a "try it, don't talk about it", after which, this comp came about.
Caveat that I just thought of - make sure to check for memory leaks, like with all spooky stuff. I ran this for a couple hours last night, b/c I couldn't remember if it had a memory leak from the spooky out of iterator shenanigans... it was ok for me.
Use note: once one has moved around all of the mesh, sample and hold, and still maintain position. That's how I usually use this kind of thing... (well, I throw in this sample and hold type patch, but it's equivalent ). It's always better to not have the iterator crunching away needlessly, once one has the positions established. The interactive handles are the bog down in this comp.
Also, hopefully it is noted that one can send a big structure to the grid (not even using this method), and get very precise warp (and also, control vertex differently from UV as well... not touched on in this comp really). The whole QCMesh/integrated geometry pipeline avails us that automatically.
hmmmm. It crashes both QC (and the crash reporter notification too, it seems, which is just beachballing at the moment) upon attempted opening. Nothing opened. Intel Core 2 Duo X7900, ATI RadeonHD2600, OS X 10.6.5. QC in 32 bit mode. Reproduceable.
Works in 64 bit, however. Checking it out now. Aha! very cool! Thanks! errr ... and happy holidays! :-)
As it happens no crash here with nVidia 9400 iMac, Intel Core 2 Duo, 10.6.5 both 32 and 64 bit, do you reckon this is another ATI related problem ? [not looked closely enough at the comp or kernel[s] inside to say ].
When it crashes, it locks up so quickly that it appears that it doesn't even have a chance to send a crash report to Kineme. I checked to make sure that function was enabled in KinemeCore Preferences, but I see no evidence of it having actually sent anything over the course of two attempts at opening the composition in 32 bit mode. Normally there is a notification of an option to send a crash report to Kineme (or at least there used to be). So, if anybody wants that report just let me know; I'll crash it again and copy out the default report that gets sent to Apple.