dyei nightmare

found the answer, was connected in wrong output, and i blame the magnetic wires think in network, view, houdini trying to guess where to conect everything is just wrong.

can you tell why this vellum set up doesnt work with dopnets but it works with a solver in sops??? vellum_for_od.hip

Im developing a pin contraints set up, i have like 19 geos configured as solid, but its taking so long, do you know a way to optimize vellum sims? Also i was thinking on asigning an specific pos with a falloff via sop solver, could it be faster? but a dunno if that could work

but why

whats up... what do you eat to be a vfxtd? I was programming this morning and felt that I lacked concentration because I did not have a good breakfast, so I asked myself what could be more convenient to eat when you are a programmer, looking for it I found that it is advisable to eat nuts, oatmeal and chocolate. because even though we are mostly sitting down, this implies a good mental effort when you spend a lot of time on this. so, in your experience what food have you noticed that benefits you when programming?

i dont think a lot of coffee is good for concentration

yes, i wonder what it means for sesi this investment

i have the same problem

jellyfish.mp4

path deform's rigidity not very clear, this feature is not clear, where should be the atribute per point to make a variable stiffness? in the curve? on the model? the tool tip says: "foreach point inte model, this attribute sets what point on the curve the model point should map to." to me, makes more sense to have this attrib on the model, but this doesnt work, neither on the curve, so where it should be?

@Noobini whoaa! no, no, dude, maybe im not explaining well my self, what mean, is that i had this misconception, of what i needed, i didnt uploaded any file because this issue has a good explanation, you see, what i want to do is to make a dinamic offset on v, where the major curvature is smoothed, but this isnt a path, but just trajectories of v, so im thinking in another solution based on calculate curvature of v and smoothing its trajectory so it doesnt bend a lot,

thanks just found that this feature isnt helping with i want, i though it was a feature that helps you to not deform in an unaceptable way with lines with high curvature, what i need is a way to give to the model a stiffness so it doesnt bend a lot, but this feature what is doing is just reposition points acording a float value what ever it comes from, it is a remapper of the position, what need is some feature that prevends extreme folding or extreme bending in a very curved line, lets say like a strech stiffness

no matter how much control you think you have over simulations on vellum, jittering is a threat that will overcome you sonner or latter, but here is a tip. most of the times jittering is coming from a high tickness value in constraints, is like the algorithm is not confortable with itself, it reminds me when you have a very tight cloth on your secret parts and you dont stop moving so, to me is like the setup tryes to get confortable with itself, like searching second rest position, and most of the cases is because this high tickness param on the constraint.

dude! you saved me hours of this problems! yess this is

ive done this softbody set up, to show what i cant find, that is, how to in herit all the velocity from geo, and position al the same time.?????? blend velocity in the constraint doesnt move anything. and is strange because the strut constraint by itself is able to move acording incoming velocity, but not position, and pin to targets interprets position, but not velocity!!! again, yes, blend velocity is ticked, this is strange, what im missinggg here fellow brotherhood????? file vellum_for_od.hip

im in the confirmation that @torque and @w doesnt work in vellum soft bodies, at least not by default vellum_no_torque.hip

how can you orient vellum instances on the solver driving every instance with force?, i wanna turn the z axis of each instance in the direction of the force but maintaining the y axis pointing upwards, im defining a matrix with a cross function for x axis. i suspect that i have to apply a driedral() but i dont know where exacty i get the current matrix transform cause vellum doesnt show intrinsics, so at the moment my matrix data just rotate randomly due the misterious @w force so how, can i determine a rotation for each instance on vex? i know i have to use probably a diedral() but how? because i dont know where to retrieve the current matrix transform

i suspect there is no torque attribute on vellum, otherwise should it be listed here? https://www.sidefx.com/docs/houdini/vellum/vellumattributes.html

Im still trying to do this with no luck, seems like i cant make work @torque attribute, the initial vector is coming from vel, passing through diedral() but no luck atm..

well i know you have to set up a torque or w angular vel, but it doesnt work this way!! file - > vellum_matrix.hip

pcfind() not finding patchname in vellum!!!! how is it posible! that... i have a pcfind inside a solver in vellum, it doesnt find it!! patchname is $FF and it gives an index per instance, but the pcfind in wrangle in the solver cant find this attribute!!! how is this even posible, the name is not string, as far as i know, or is it? it goes as an integer, and even so it cant read it! also i would like to search attribs in the other instances with diferent patchname, i tryied this in a foreach, but it also doesnt work, starting from this fact. file ---> vellum_patchname_for_od.hip

just a little progres on something im working. playblast.mp4

done!!! it worked now