You mean like this ?
I don't get the same image unfortunately.

Thanks for you answer jsmack.

I was indeed using this method, but I have no idea of what to put in the "From space" and "To space" values.

Hello everyone,

I am currently rendering frames with Mantra in OCIO, and then I am using COPs to modify my final image. But my goal is then to export the final frames to jpg. But obviously when I write my exr rendered with OCIO into jpg, I have a different image from the one I had in the render view and in the composite view. Does anyone knows which parameter I need to use before writing out my frames into jpg ? Do I need to do some sort of conversion in a vop ?

Thanks a lot,

Tom

Here, it break where there is a lot of branches, which makes sense. But I would like deformation to be uniform.

Wait I'll send you another scene to explain you why, I didn't want to share a ‘complicated’ scene (the scene won't be complicated at all but I thought sending an example with just a curve would be enough), maybe you'll see what the problem is.

Thank you ! Pretty useful. Not sure if it's what my setup should use but it's a great start.

Oups sorry I just didn't think of uploading a file. Here it is

Hi everyone,

Quick vellum question here, is there any way to get the same dynamic I have on the curve with the points uniformly spaced but on the curve with the ‘dirty’ spaced points. It makes sense that it will bend more where there is more points, but I really want to avoid it and yet keep the same point spacing as the ‘dirty’ curve. Not sure if it's clear.

Also, I don't want to use a point deform or any post sop operation, I need to sim this ‘dirty’ curve.

Thanks,

Tom

Oh wait, surprise, it seems to work now…

Hi everyone,

I am currently using the instancing to do some pyro stuff, so I have points and each points are instancing a container. But I have many points, so the instancing is pretty efficient.

So I am using the Instance node, in the object level, and inside I'm loading the points that will later be containers when rendered.

But I am also loading points as a point cloud in the shader of the volume, each point has a color and that color is transfered on the diffuse's volume. So it's working fine with classing volume, but it's not working when it's instanced. So I tried with a copy stamp, with the volumes being packed before copying, it works but it's very heavy and the render is not efficient because there is so many container that are on top of each other that it's taking ages to render. So I used the volumestamp sop, and it's also working, but again it's very heaving to make each frame, because the final container is pretty big, even with vdb.

So my question is, can I avoid these two methods, the copy stamping and volume stamping, and just keep using the instancing but still be able to load my point cloud in my shader ? Or is it just not possible ?

Thank you,

Tom

Hi everyone,

Is there a way to move the color palette (the one on the network view, to change nodes color) on the left instead of having in on the right ? It might just be personal, but it's quite annoying since my parameters tab is also on the right, and the color palette used to be on the left on the versions below h16. Sorry if the questions has already been asked.

Thank you,

Tom

But what if I want to keep my twist constraint ? Multiple impacts can happen over time

Hi everyone,

Is there a simple way to freeze twist constraints after they've been deformed ?
For example in that video : https://vimeo.com/89584231, [vimeo.com] the object almost reacts as a stiff cloth, or a sheet of metal, but what if I want that to act as sheet of metal that keeps its deformation after an impact ? A bit like a car for example, which keeps his bending after impact. But I want to stay with twist constraints.

Thank you,

Tom

Thanks !

But I don't even know how to use all these nodes…
I have some idea on how it works, but I would not be able to do it myself for now.

“it is actually not as bad as it sounds once you get used to it. For rotations, I generally use a lot the Align VOP in between two vectors to generate my quaternians (set it to axis instead of matrix), then after that there are a lot of nice quaternian vops for doing those kinds of operations (qrotate, slerp, etc). if you are using quaternians, you also have the advantage that your rotations will never gimble. There is also a quaternion subfolder in the tab menu for convenience ”

Do you have an example for this kind of situation ?

I'm waiting for your tutorial now :evil:

(You have to consider the fact that I'm coming from a literary training…)

Thanks !

Hi,

Is there a way to learn about matrices in vop sop, rotation in vop sop, quaternion, transform in vop sop, etc… I just don't get it, and EVERYTIME I see a tutorial with these kind of things, it's not explained, like if it was so simple… I am not a beast in maths, so it's not always easy for me with this kind of stuff.

Thanks,

Tom

You can check tutorials about this effect in 3dsmax and Blender, and try to recreate it in Houdini.

“- All the SSS AOVs (except for sss_clr, the raw unprocessed color) are coming out blank.”

Since H15 I also have this problem.

Well it seems that the “scope:self” is only compatible with the Ray Tracing engine, but not with PBR… Is it correct ?
With PBR, and with the “scope:self”, the dirt mask just disappears.

Working great !

Thanks Tamte.