K3d object mirror

An object transformer that mirrors 3d objects in a configureable way would be cool. In fact any more k3d object modifiers would be awesome.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

cwright's picture
Re: K3d object mirror

I think you can accomplish a mirror by scaling with a negative value (this will break face culling though - you'll have to reverse it I think).

I wasn't very proud of the way object modifiers were accomplished (because they aren't configurable, and thus require more software to add new deformers, and they're all done on the CPU, which is slow for complex meshes). Too bad I didn't think of a better idea until I went to Apple...

blackburst's picture
Re: K3d object mirror

I actually meant more of a "symmetry" effect rather than flipping the whole model, with moveable mirror positions. But good to know with the inverse scaling :)

"Too bad I didn't think of a better idea until I went to Apple..." Aghh so cruel...What better idea? Stays in the loop right? Oh well. I really love these things, if not programatically brilliant they allow you to make some really snappy audio rective mesh fx. I am basing a lot of stuff on these patches, in your opinion is there something else I should be looking at/forward to for these kinds of functionality?

gtoledo3's picture
Re: K3d object mirror

If "better" is anything that depends on Apple getting GPU drivers right, the CPU path might still be the more workable option. K3D has saved my butt at times when Apple hasn't had their act together with GPU drivers.

gtoledo3's picture
Re: K3d object mirror

It would have been reasonable for you guys to have written a separate patch (and keep K3D how it is), that allows you to write new deformers in QC and live compile (ala Mirek's obj-c plugin). I've made and tested this, it works.

In retrospect, I think my biggest problem with the implementation was the concept of the custom port (and in other places where kineme does this). I think of the two options:

1-Person links a virtual or structure port to something with data it "isn't expecting", QC throws an exception.

2-Person makes an input splitter inadvertently, and bakes a non-existent patch into the composition that throws exceptions and leaks.

...that the first was the better option.