Hi

I am not sure I understand the displace along normal VOP
Shouldn't the output attribute dispN be the modified normals?

When applying this VOP node on a simple grid, the dispN outputs (0, 1, 0)

Any idea?

Thanks
B

computing dispN requires shading derivatives to be defined, which is only true in displace and surface contexts. If you are displaying points in SOPs using an attribVOP, this is under the cvex context, where no such derivatives exist, so a simple fallback value is all that is returned. However, do not fret, as the attribVOP has a checkbox that will recompute normals for points with changed position for you when enabled. So you can leave the dispN output disconnected, and use the checkbox instead.

Hi Jsmak,

thanks for your answer, I did not explore that second tab with the checkbox.
Normals are updated actually.

Now, regarding VOPs and CVEX, I think I am a little more confused.

I though that in an Attribute VOP I was under the VEX SOP context (and in a Volume VOP under the VEX Volume context…)

The documentation actually says:

The VOP network runs in the CVEX context, not the SOP context. This means SOP specific VOPs and VEX operations are not available. Instead use the From File variants and the provided operator input strings.

Attribute VOP [sidefx.com]

But in an Attribute VOP I can use an Add Point VOP or an Add Primitive VOP, which I would classify as SOP specific operations. Or are they?

So… ummm, well I am a little lost here, could you shed some light?

Thanks.

To be sure, the Attribute VOP is a SOP, however the ‘SOP’ context mentioned in the documentation is a context of yore.

The SOP context was used by the old VOPSOP and VEX SOPtype builders, and has since been superseded and deprecated.

It has a different scope of commands and inputs than the CVEX context used by the attribute VOP. For instance, SOP context can only run on points, cannot create or destroy geometry or output attributes for a different class. It used global variables instead of bindings. Reading of inputs worked a bit differently too, as it used special attribute input nodes instead of using opinput file paths.

I wouldn't worry about the sop context too much as it should no longer be used to build new networks ( however some older HDA still have vopsop nodes internal to them.)

Thanks.
I think what makes it difficult for me is to have started learning VOP and VEX in the midst of the transition:
I cannot really refer to what it used to be, and I can’t tell the before from the after when reading the docs and tutorials.

Ok, so from now on I should consider all VOP and Wrangle nodes to run CVEX?

And for instance, this kind of nodes, like Displace Along Normals VOP, that have some functionalities in displace and surface contexts but not in cvex, how can I know about this?

For this one I guess that’s what this part of the doc means?

Note that in the SOP/POP contexts, the shears and object space deformation are not implemented. They exist as parameters for completeness only.

Displace Along Normal [sidefx.com]

bebe
Ok, so from now on I should consider all VOP and Wrangle nodes to run CVEX?

unfortunately no
VOP COP and VOP CHOP and corresponding wrangles are still separate dedicated contexts cop and chop
also all the shading contexts are dedicated contexts like surface, displace, light, shadow, fog

but the rest is mostly cvex

Ok, I see.
thanks everyone