hi,
i have a vellum cloth setup, nothing special a simple dress (made with marvelous and remeshed for houdini), about 10K prims, so fairly low-poly.
vellum solver settings are mostly default, time step 4, 250 constraint iter...
when i simulate, it takes about 1.2 sec per frame, which leads on a 200frames simu in about 4-5min simulation time and the simu i consider not final, since it flickers when it comes to rest, so settings should be raised up...
nevertheless, when i simulate in marvelous, the dress then contains about 50K prims and it simulates (in cpu mode) in about 30secs, no flicker and it looks awesome...simulation settings fairly high (for animation)...
i am wondering why vellum solver is sooo slow (opencl is enabled in the solver)...and how can be it improved? somethin in the solver settings that has to be enabled to make it faster?
i would like to simulate in houdini, since it has more flexibility on settings things up or art direct the cloth...but the simulation speed is sooo slow, comparing to marvelous designer...
Found 45 posts.
Search results Show results as topic list.
Houdini Indie and Apprentice » Vellum Cloth - Solver Speed
- soulcage_dpt
- 45 posts
- Offline
Houdini Indie and Apprentice » vellum post process "apply weld" uv issue
- soulcage_dpt
- 45 posts
- Offline
hi,
i have a cloth that has UDIMs...
when applying the welds in a vellum post process and exporting the cloth as alembic (for using in maya) the uv UDIMs are gone and no uvs appear in maya...
but when NOT applying the welds, then the alembic has uv UDIMs...the same applies when having only one uv (range 0-1)...
does anybody know whats going on and what "apply welds" exactly does under the hood? it seems to change something on how uvs are then handeled and maya doesnt like it...
i have a cloth that has UDIMs...
when applying the welds in a vellum post process and exporting the cloth as alembic (for using in maya) the uv UDIMs are gone and no uvs appear in maya...
but when NOT applying the welds, then the alembic has uv UDIMs...the same applies when having only one uv (range 0-1)...
does anybody know whats going on and what "apply welds" exactly does under the hood? it seems to change something on how uvs are then handeled and maya doesnt like it...
Houdini Indie and Apprentice » vellum solver sop - external collision friction
- soulcage_dpt
- 45 posts
- Offline
hi,
i want to change the friction of the animated collision object and i dont know where to adjust these property directly on the vellum sop solver...
do i have to dive inside the dopnet (static object?)? or is there another way of doing this?
and is the friction of collision objects also paintable (to have only some areas having different friction)
i want to change the friction of the animated collision object and i dont know where to adjust these property directly on the vellum sop solver...
do i have to dive inside the dopnet (static object?)? or is there another way of doing this?
and is the friction of collision objects also paintable (to have only some areas having different friction)
Edited by soulcage_dpt - Nov. 25, 2023 05:29:07
Houdini Indie and Apprentice » Aces color problem
- soulcage_dpt
- 45 posts
- Offline
...now i am a bit confused...
the osl i provided works not as expected...especially with the "inverted ACES"...dont know why and have to investigate (or ask redshift forum)
with a map generated from the hexcode and mapped to the grid with "inverted aces" in the rsTexture, works as expected...
as a "attribute from map" and the Cd as a rsUserColorAttribute feed into the osl, doesnt work...very strange...
the osl i provided works not as expected...especially with the "inverted ACES"...dont know why and have to investigate (or ask redshift forum)
with a map generated from the hexcode and mapped to the grid with "inverted aces" in the rsTexture, works as expected...
as a "attribute from map" and the Cd as a rsUserColorAttribute feed into the osl, doesnt work...very strange...
Houdini Indie and Apprentice » Aces color problem
- soulcage_dpt
- 45 posts
- Offline
1. houdinis 19.xx.xx color picker has a bug...it always has linear-srgb as input...
so if you feed in the hexcode above, it thinks it is in linear...so with the little "apply color correction" activated in the right top corner of the color-picker, you get a double-color-transform (see also in viewport)...
so you have to first linearize the hexcode before feeding in the values...(from #B030DE-srgb to #6E07BA-linear)
2. to then get this "linearized" color to match exactly in ACEScg (which will be impossible with a normal color-transform from linear-srgb to acescg), you will need a so-called "inverse-aces" transform...
redshifts ocio-config has one built in called "invert ACES 1.0 SDR-video" but this is a bit tricky because this config only assumes image-inputs.
on a color sop this conversion can only happen with some fancy color-matrix-maths (dont know if there is online color conversion out there, that converts from inverse-aces to acescg...)...
3. but there is a confortable way to achive this...redshift OSL.
here is a OSL-file that takes a color as input (dont forget to linearize it first because of the color-picker bug) and you specify a "input color-space" and a "output color-space" (in your case here "invert ACES" to "ACEScg")
but be careful with this color-transform, very bright colors will start to emit light, because i.e. 1.0-srgb will become 16.0 in acescg...
so if you feed in the hexcode above, it thinks it is in linear...so with the little "apply color correction" activated in the right top corner of the color-picker, you get a double-color-transform (see also in viewport)...
so you have to first linearize the hexcode before feeding in the values...(from #B030DE-srgb to #6E07BA-linear)
2. to then get this "linearized" color to match exactly in ACEScg (which will be impossible with a normal color-transform from linear-srgb to acescg), you will need a so-called "inverse-aces" transform...
redshifts ocio-config has one built in called "invert ACES 1.0 SDR-video" but this is a bit tricky because this config only assumes image-inputs.
on a color sop this conversion can only happen with some fancy color-matrix-maths (dont know if there is online color conversion out there, that converts from inverse-aces to acescg...)...
3. but there is a confortable way to achive this...redshift OSL.
here is a OSL-file that takes a color as input (dont forget to linearize it first because of the color-picker bug) and you specify a "input color-space" and a "output color-space" (in your case here "invert ACES" to "ACEScg")
but be careful with this color-transform, very bright colors will start to emit light, because i.e. 1.0-srgb will become 16.0 in acescg...
Edited by soulcage_dpt - Sept. 27, 2023 06:29:02
Houdini Indie and Apprentice » group-name from name attribute
- soulcage_dpt
- 45 posts
- Offline
thx a lot for the hints...
@Enivob, this works perfect, thx for the snippet (had to rename the name-attribute since it contained a namespace-character ":" and it seems group-names dont like ":")
@sepu, the "groups from name" sop doesnt exist in 19.0.xx, seems only be available in 19.5.xx, but this also works perfect in 19.5.xx, the renaming of the namespace-character is then automatically...thx
@Enivob, this works perfect, thx for the snippet (had to rename the name-attribute since it contained a namespace-character ":" and it seems group-names dont like ":")
@sepu, the "groups from name" sop doesnt exist in 19.0.xx, seems only be available in 19.5.xx, but this also works perfect in 19.5.xx, the renaming of the namespace-character is then automatically...thx
Edited by soulcage_dpt - Sept. 26, 2023 03:54:10
Houdini Indie and Apprentice » group-name from name attribute
- soulcage_dpt
- 45 posts
- Offline
hi,
i have some points with a name attribute.
i want to create groups with the group-name from this name-attribute, but the group node doesnt accept @name as the group-name.
does somebody know how to achieve this?
(i need these group-names for better selection points on the rivet obj node...)
i have some points with a name attribute.
i want to create groups with the group-name from this name-attribute, but the group node doesnt accept @name as the group-name.
does somebody know how to achieve this?
(i need these group-names for better selection points on the rivet obj node...)
Edited by soulcage_dpt - Sept. 25, 2023 15:17:50
Houdini Indie and Apprentice » Replacing Assets in Layout LOP after painting
- soulcage_dpt
- 45 posts
- Offline
hi,
i have scattered some assets with the layout lop and now i need to replace some assets with other assets...but i cant figuere out how to replace them with others...
does sombody know how to do that so my scattering stays the same but with other assets?
i have scattered some assets with the layout lop and now i need to replace some assets with other assets...but i cant figuere out how to replace them with others...
does sombody know how to do that so my scattering stays the same but with other assets?
Houdini Indie and Apprentice » Extracting KineFX transform matrix to OBJ-Level
- soulcage_dpt
- 45 posts
- Offline
mestelathx a lot.
Rivet?
yes rivet seems to be an option, but i also need scaling, which seems not possible with rivet. or is there another trick to extract also scaling?
CYTEi tried packed, but no really performance gain. the copy to points acts in my case more like a contrainer and not an instancer...
Have you packed your geo before copy to points or while? That would improve performance significantly.
the objects being transformed to the points are all unique...
(also i need them unpacked at obj-level because using redshift and need its obj tags for shading...)
Edited by soulcage_dpt - Oct. 10, 2022 06:48:47
Houdini Indie and Apprentice » Extracting KineFX transform matrix to OBJ-Level
- soulcage_dpt
- 45 posts
- Offline
hi,
i have an fbx animation coming from maya and using kinefx fbx-anim-import to load it to houdini.
the imported points have also a "transform" (translate, rotate, scale) 3x3 matrix attribute.
i have to constraint some heavy assets to this points, so my first intension was to use copy to points in conjunction with the name attribute from the points.
works well, all transforms are coming through, but its very very slow when scrubbing the timeline.
constraining the assets on obj-level would be more effective, but how to extract the tranform matrix from the fbx-import-node at sop-level to the TRS of a geometry-node?
is there an expression or something else to use the matrix directly?
any hints are welcome
i have an fbx animation coming from maya and using kinefx fbx-anim-import to load it to houdini.
the imported points have also a "transform" (translate, rotate, scale) 3x3 matrix attribute.
i have to constraint some heavy assets to this points, so my first intension was to use copy to points in conjunction with the name attribute from the points.
works well, all transforms are coming through, but its very very slow when scrubbing the timeline.
constraining the assets on obj-level would be more effective, but how to extract the tranform matrix from the fbx-import-node at sop-level to the TRS of a geometry-node?
is there an expression or something else to use the matrix directly?
any hints are welcome
Houdini Indie and Apprentice » Material builder - create copy reference cut/paste
- soulcage_dpt
- 45 posts
- Offline
hi,
i have a bunch of material builders.
i want to share a material by "create reference copy" from one of the builders and copy-paste them to other builders, so i can adjust them with only one "master".
but it doesnt seem to work, as soon they are pasted to other builders, the referenced channels doesnt update to the pasted builder paths...
is there something i miss with creating referenced copies on materials (inside a builder)? or is there a build-in command i can use to change the channel paths, so they point to the right location?
i have a bunch of material builders.
i want to share a material by "create reference copy" from one of the builders and copy-paste them to other builders, so i can adjust them with only one "master".
but it doesnt seem to work, as soon they are pasted to other builders, the referenced channels doesnt update to the pasted builder paths...
is there something i miss with creating referenced copies on materials (inside a builder)? or is there a build-in command i can use to change the channel paths, so they point to the right location?
Edited by soulcage_dpt - Sept. 25, 2022 06:03:01
Houdini Indie and Apprentice » HDA Asset Workflow
- soulcage_dpt
- 45 posts
- Offline
thx for the hint...i will try.
normally it shouldnt update automatically, because lets say you are middle in rendering with v1.1 and modify to a new v1.2...then when it automaticalky updates then you mess up rendering...
(but there are situations where you want to automate the update process)
normally it shouldnt update automatically, because lets say you are middle in rendering with v1.1 and modify to a new v1.2...then when it automaticalky updates then you mess up rendering...
(but there are situations where you want to automate the update process)
Edited by soulcage_dpt - Sept. 24, 2022 06:12:57
Houdini Indie and Apprentice » HDA Asset Workflow
- soulcage_dpt
- 45 posts
- Offline
hi,
i have an asset geo as an HDA...
when modifying i typically create a new version and it will be saved with the new version number, which is fine, so i see a bunch of hda files in the hda folder...
but when an older version of the hda was placed in a hip, how can it be replaced with the new version? in maya i would "replace reference" in the reference editor.
what is the equivalent to this in houdini? and how can this be automated?
i have an asset geo as an HDA...
when modifying i typically create a new version and it will be saved with the new version number, which is fine, so i see a bunch of hda files in the hda folder...
but when an older version of the hda was placed in a hip, how can it be replaced with the new version? in maya i would "replace reference" in the reference editor.
what is the equivalent to this in houdini? and how can this be automated?
Houdini Indie and Apprentice » VDB compute velocity
- soulcage_dpt
- 45 posts
- Offline
Houdini Indie and Apprentice » VDB compute velocity
- soulcage_dpt
- 45 posts
- Offline
hi,
i have a vdb sequence with only density...
i want to compute velocity (like the trail sop) and merge it into it, so i have in the end a volume with density and vel...
is that posdible? and if yes how (maybe by subtracting current frame P from previous frame P?)
i have a vdb sequence with only density...
i want to compute velocity (like the trail sop) and merge it into it, so i have in the end a volume with density and vel...
is that posdible? and if yes how (maybe by subtracting current frame P from previous frame P?)
Edited by soulcage_dpt - Sept. 14, 2022 02:31:52
Houdini Indie and Apprentice » Houdini 18.5.696 - H19.0.383 - GLSL shader failed to compile
- soulcage_dpt
- 45 posts
- Offline
seems that i could fix this problem by installing the latest studio driver...
hope this helps anybody who has this experience...
hope this helps anybody who has this experience...
Houdini Indie and Apprentice » Houdini 18.5.696 - H19.0.383 - GLSL shader failed to compile
- soulcage_dpt
- 45 posts
- Offline
Hi,
after updating to newest nvidia driver, 496.49, i get a viewport message that a glsl shader cannot be compiled (beauty_wire_lit.prog) and this leads in not displaying alembics as unpacked geo in viewport...
in H19.0.383 this even worst, here the alembic is not loading correct (missing groups/meshes).
i also tried older versions of the nvidia driver, the only versio i got working was 466.xx.
but i need newer nvidia drivers for other apps i use...
any hints how to fix that?
after updating to newest nvidia driver, 496.49, i get a viewport message that a glsl shader cannot be compiled (beauty_wire_lit.prog) and this leads in not displaying alembics as unpacked geo in viewport...
in H19.0.383 this even worst, here the alembic is not loading correct (missing groups/meshes).
i also tried older versions of the nvidia driver, the only versio i got working was 466.xx.
but i need newer nvidia drivers for other apps i use...
any hints how to fix that?
Houdini Indie and Apprentice » Houdini Indie Network Rendering
- soulcage_dpt
- 45 posts
- Offline
Houdini Indie and Apprentice » Houdini Indie Network Rendering
- soulcage_dpt
- 45 posts
- Offline
hi,
i have 1x houdini indie.
i wrote a little python script that can submit a houdini file to backburner with hpython2.7.exe, hrender.py and redshift as the renderer for network rendering...
for testing i installed houdini indie on 2 PCs.
but only the one PC that has the license server installed can render.
is houdini indie capable of network rendering (like mayabatch) with only one license?
i have 1x houdini indie.
i wrote a little python script that can submit a houdini file to backburner with hpython2.7.exe, hrender.py and redshift as the renderer for network rendering...
for testing i installed houdini indie on 2 PCs.
but only the one PC that has the license server installed can render.
is houdini indie capable of network rendering (like mayabatch) with only one license?
Edited by soulcage_dpt - Sept. 8, 2021 06:11:47
Houdini Indie and Apprentice » Scattering Cards without Intersection for Vellum Simu
- soulcage_dpt
- 45 posts
- Offline
hi,
i am scattering cards on a character and running in the problem that this cards are intersecting each other (depending on the size, i need big cards...), so a vellum cloth simulation will not have a proper start geo...
with intersecting cards i get a bad vellum simulation, because this "intersecting cards" just sticks together...
i tried to use the measure node in conjunction with the gradient to setup a better distribution of the scattered cards, so they dont intersect each other...but with no success...
also the detangle node doesnt give me the espected result...
has anybody a hint how to avoid such "cards" intersections, deintersection??, so a vellum sim has a proper starting mesh?
(would maybe the hair-to-cards give non-intersecting cards?)
i am scattering cards on a character and running in the problem that this cards are intersecting each other (depending on the size, i need big cards...), so a vellum cloth simulation will not have a proper start geo...
with intersecting cards i get a bad vellum simulation, because this "intersecting cards" just sticks together...
i tried to use the measure node in conjunction with the gradient to setup a better distribution of the scattered cards, so they dont intersect each other...but with no success...
also the detangle node doesnt give me the espected result...
has anybody a hint how to avoid such "cards" intersections, deintersection??, so a vellum sim has a proper starting mesh?
(would maybe the hair-to-cards give non-intersecting cards?)
Edited by soulcage_dpt - June 18, 2021 05:44:58
-
- Quick Links