Archived post by PaqWak

So I was looking for a way to find the longest path in a network of edges, like the opposite from what the ‘find shortest path’ does. I remember looking for a solution previously without succes. So I asked Skynet (Chat-GPT), and here’s his solution :

The doc don’t mention that, actually the shortest path OP complain about the negative weight … but it indeed output the longest path as far as I can tell 🥳

Attachments in this post:
http://fx-td.com/houdiniandchill/wp-content/uploads/discord/20234501/11/23/image.png

Archived post by toadstorm

ok i think gemming is working again? managed to get a really high quality post or three up on the archive

i just need to migrate the bot instance to my big boy webserver instead of running off my NAS and then we should be good to go

hmmm maybe I’ll try to get image links to actually display as images for this next build so everyone can see these incredibly important posts as they’re meant to be fx-td.com/houdiniandchill/wp-content/uploads/discord/20222312/19/22\fart.gif

Archived post by SniperJake945

you can get some neat effects by adjusting the above code too, like a weighted max, with just a constant y value for the gradient can give you this like pillars rising up effect: “`c vector g = volumegradient(1, 0, @P); //taking the gradient from the second input,
float intensity = ch(“intensity”); float dx = length(v@dPdx); //voxel size; g *= dx; //scale our vector by voxel size float samples = chi(“samples”); float out_val = @density; float weights = 1.0; for(int i = 0; i <= samples; i++){ float u = i / float(samples); //0 -> 1 vector sample_p = v@P + g * u * intensity; float value = volumesample(0, 0, sample_p); float w = smooth(-.1, 1, 1 – u); //in theory this should be a gaussian, because its a better falloff, but yolo out_val = max(out_val,value * w); //weighted max
} @density = out_val;// / weights; “`

Attachments in this post:
http://fx-td.com/houdiniandchill/wp-content/uploads/discord/20222908/22/22/unknown.png

Archived post by will

Here is another go a squash and stretch, I calculated the distance between two points and then scaled a joint in-between.

I have added some shape controls on the rig vop, something to play with

Attachments in this post:
http://fx-td.com/houdiniandchill/wp-content/uploads/discord/20220908/11/22/2022-08-11_09-30-05.mp4
http://fx-td.com/houdiniandchill/wp-content/uploads/discord/20220908/11/22/squash_and_stretch.hiplc

Archived post by Peter Sanitra

Re, the difference in volume look on Karma vs Karma XPU. XPU renders at 0.25 step rate by default, and there is no control for it on the ROP, unlike regular Karma CPU. So if you have Regular Karma, have switched volume step rate to 1 on ROP, then your XPU volumes gonna looks “shitty”, compare to that. Way to get nice crisps volumes on XPU is to drop Render geometry settings node and set step rate manually to 1. I`ve seen lot of post about how XPU volumes are ugly, compare to “their engine”, how they lack detail. Yep, they do , when you render at 1/4 res…