sirogi

Hiya, Thwarted at step 0 of trying to use the crowd pipeline with my custom agents: Trying to use the amazing "agent from FBX feature", but so far, the rig am wrestling with does not export cleanly out of Maya. Given a simple run cycle, I (and much smarter people than I) tried exporting a "nice" FBX to no avail. The joints export fine, animation and all, but for whatever reason the mesh is frozen to a single frame (And that's just reloading the FBX to Maya to check the export). Will keep debugging the Maya side separately (This isn't a Maya forum, last I checked ) Hoping someone here could shed light on a few points: -On the Houdini side, what is the expectation from the FBX file format? how complex is the rig-skin representation? is the skin actually bound to the skeleton, or can they exist side by side? (geo cache and animated skeleton for manipulation by Houdini) -Are there ways to export separately then re-combine in Houdini? (abc for the mesh for instance.. stick to the FBX for the joints) Any help is much appreciated!!

m..mmarc you're 27?!! oh my! Happy b-day!

nice shells..... Try enabling shadows on the lights to get that extra "solid" feel and get rid of the brightness inside... you could also set the shadow intensity to something less than 1, this way you let some of the light sip in....that doens't compare to true translucency, but hey....would look nice... shells are always cool

um....yes... as in...yes I hope for one too

he he he....take another look at prman now, go ahead... I just started to play with prman, after messing around with vex, and I have to say (so far), except for the syntax difference, they're the closest 2 languages I've used. let's just say that VEX is less fussy about type casting.... differences (for the shader writer) are skin deep. for instance in VEX you only need a vector for everything whereas in prman you have "point" "vector" and "color" types.....let' s just say that VEX is easier and rman in safer (I'm NOT arguing which is "best", ok? ) but they're fundamentally the same, dealing with the same concepts... both closer to C.... Are you sure you're not talking about RIB files? (the prman equiv. to IFDs...scene descr. files...) those still don't look like C++, but they sure can get confusing ....IMHO regardless shading is possibly one of the coolest areas of CG, add lighting and serve hot! Cheers!

...and for finding a good source of shaders to disect and play with go to: http://www.renderman.org/RMR/ (ok, so they're all prman, so what? ;p) and try to start with the simplest ones....it's all about divide and conquer. another cool one to get into procedural textures: http://www.accad.ohio-state.edu/~smay/RMan.../rmannotes.html and last (should be first) but by no means least, our own good ol' Jason's tutorials on odforce itslef (remember, it's more than the forum) ;P http://odforce.net/tips/ (7th and 9th topics) you even get an i3d primer! guess THAT shoudl get the ball rolling Cheers!

err...actually, for the ballon thingy, I just exported one bgeo, and then imported one file...never neede to export/import a sequence, but I heard it's been done... good luck!

WAHA HA HA HA!!! I second that motion....dunno the guy personally, but I bet his vtm would make it on the comedy channel...easily!

absolutely, (the onlihe help in the SOP) if you look at the construct: [lc<]pred[>rc]:cond=blahblahblah... the ":cond" part is where you plug in the controlling params.... for instance premise : A(0) A(i):i&lt;5=FA(i+1) A(i)=+60FA(i+1) would execute the first rule 4 times, and switch to the second one after that. always remember the L-systems read the rules top to bottom and use the first rule that is valid, so you can have a host of rules that are all A, the L sys will go through them one by one and only apply the first valid one... and the first 2 variables can have a special meaning for the length and thickness say so A(k,l,m) the variables (k) and (l) are multipliers for the length and thickness (uh...check the docs for this maybe the PDF, haven't used L-sys in a while...) don't forget to use the "b,c,d" user vars in the "Values" tab, they allow you to control features for the whole system . For instance, in the preceding example replace i<5 by i<c set the generations to something higher than 5 and play with the c slider, the system should change immediately....it's a way to affect the whole look of a system from one variable... Most of the stuff here is a direct derivation from thew book, I just had to adapt the syntax a little bit, but the book is full of insights like these... hope it made sense...now go forth and go nuts! Cheers!!

asked myself this one a million times!!! yes, people do, and I know exaclty how you feel about what you're doing...I been trying it on and off for the past few moths, and in short, trust the metaball pruning method for simple constraints. if you want to get more sophisticated, you can increment variables inside your L-system rules, and then disable or enable rules based on it, problem is, it doesn't take into account "world" constraints (as in: crawl that wall, go round that column) you'd have to do that with good ol' SOPS (creep SOP)...hmm... hope I didnt' confuse you yet...wish I had time to tell you more or post an example...more to come when I find the time...it can be really rewarding... and cs00Bren....the best way to find examples with the rules is the "Algorithmic beauty of plans book"...or the siggraph papaers from the same author....I can't remember anywhere else where you could find such examples...but they do a good job at pointing you in the right way... cheers... and please...keep using it, that makes at least 3 of us that are interested (on this forum at least)....

gagg** ....better late than never!!!

Dang!!! You managed to steal away 15 min from me! yeah...I bet alot could be accomplished once those "proxy groups" get nito VOPS!!!! a LOT of things..... the file is my 2c. galactos.zip

Sure sounds like apprentice....

don't bother, man....make no difference whatsoever

I' bored... want new poll.... please Marc? may we have some more?