Hi everyone,

I'm rolling out a series of "Houdini Shorts" not just on YouTube but also on TikTok, Twitter, and other platforms like LinkedIn. These videos will provide additional Houdini tips, tricks, and sometimes a sneak peek into my in-depth advanced content like Pragmatic VEX course(s).

Feel free to subscribe/follow for more updates.

Hope you find these useful, and as always thanks for watching!




Exclusive content:

https://www.pragmatic-vfx.com [www.pragmatic-vfx.com]
https://www.patreon.com/animatrix [www.patreon.com]
https://animatrix2k7.gumroad.com [animatrix2k7.gumroad.com]

Follow us on:

https://www.youtube.com/@pragmaticvfx
https://www.tiktok.com/@pragmaticvfx
https://vimeo.com/animatrix2k7 [vimeo.com]
https://twitter.com/pragmaticvfx [twitter.com]
https://instagram.com/pragmaticvfx [instagram.com]
https://www.linkedin.com/in/rdx2k7 [www.linkedin.com]

New short video just dropped today:

Finally, Houdini's training began to be published in a one-minute TikTok format!
Learning Houdini has never been so fast!
🔥

ZeitG
Finally, Houdini's training began to be published in a one-minute TikTok format!
Learning Houdini has never been so fast!
🔥

Thanks man!

New short just dropped today

animatrix_
New short just dropped today

video]https://www.youtube.com/shorts/d79jCZ8aN0Y [www.youtube.com]

this one is precious

Andr

animatrix_
New short just dropped today

video]https://www.youtube.com/shorts/d79jCZ8aN0Y [www.youtube.com]

this one is precious

Thanks a lot mate! More to come 😎

Brand new short video for Houdini 20

call me a boomer, but how on earth is this tiktock format relevant for anyone using houdini?

animatrix_

I'm not quite clear on this. Are you recommending splitting complex code into multiple wrangles for performance or readability?


Regarding performance: Splitting a block of code into multiple wrangles forces you to save data to attributes, which are then read by the following wrangle. Doesn't this approach cause overhead compared to keeping all data in memory in a single wrangle?

rickapex
call me a boomer, but how on earth is this tiktock format relevant for anyone using houdini?

It's short format videos for quick tips. There are quite a few people making these as they are orders of magnitude faster to produce than full length videos while still being useful.

Andr

animatrix_

I'm not quite clear on this. Are you recommending splitting complex code into multiple wrangles for performance or readability?


Regarding performance: Splitting a block of code into multiple wrangles forces you to save data to attributes, which are then read by the following wrangle. Doesn't this approach cause overhead compared to keeping all data in memory in a single wrangle?

It's on a case by case basis. For example when implementing some algorithm like point cloud fitting, doing it all in 1 wrangle is extremely inefficient, but also forces you to use detail run over mode.

For saving attributes it also depends. If you save the attributes on the current elements using the @ syntax, the overhead should be minimal. But sometimes even using setpointattrib, setprimattrib, etc functions is worth it if you can parallelize the next steps. Ultimately you have to profile and see if you gain performance doing this vs not and how much more complicated the setup becomes vs the speed gains.

Very useful new VEX function added to H20 thanks to SESI

Now with Sky voice (Scarlett Johansson soundalike)

Sky voice again

Sky Voice is still alive despite OpenAI's removal of it from ChatGPT