Jump to content
LaidlawFX

Houdini 13 Wishlist

Recommended Posts

- Group/Partition Nodes to inherit attributes.

If a object is in a group and the group has the material applied the object uses this material, same goes for other properties.

"categories" are probably the way forward on this. However, I'm not up on the lighting side of things. My personal beef on this is that we've already invented several methods to deal with these workflows and I'd rather like to see some consolidation rather than invent yet another system.

  • Like 3

Share this post


Link to post
Share on other sites

Point Instancing is supported in the GL3 viewport using GL instancing for the Instance Object. Does this cover it?

As far as I understood, no. I meant different instances per point. when I feed the instance attribute with different objectpathes they do not show up in the viewport . only the object assigned at objectlevel shows up in the viewport

"categories" are probably the way forward on this. However, I'm not up on the lighting side of things. My personal beef on this is that we've already invented several methods to deal with these workflows and I'd rather like to see some consolidation rather than invent yet another system.

maybe it could also extended. categories or bundles can not be saved with assets, as far as I know. and they can not propagate attributes over their members. if I have tons of screws on a robot and want to render passes on them I end with tons of entries in the take list. if I could share attributes by groupes/partitions i could live with one entry. I still think it would make life easier with takes. All what is needed is a way to share attributes in a non hierarchical manner. probalbly no easy thing to implement but in combination with bundles and categories it would be very powerful

Share this post


Link to post
Share on other sites

"categories" are probably the way forward on this. However, I'm not up on the lighting side of things. My personal beef on this is that we've already invented several methods to deal with these workflows and I'd rather like to see some consolidation rather than invent yet another system.

I certainly agree. There are good things about the light mask, light linker, forced objects, etc within Houdini, even stuff from other packages that is simple to do, that should be reviewed, organized, and consolidated, into a good production style workflow.

This would certainly be a pretty cheap bang for the buck, not much time invested into technology development since it already exist, but with some serious production workflow style planning on how to pakcage and wrap the stuff up. I know SideFX has heard the numerous issues with this before, it would be great time to wrap all these ideas in with some good feed back.

I know two things along this vane are that categories need to have their own pane like the light linker, being able to add and remove mask and category parameters to subnets... etc.

Share this post


Link to post
Share on other sites

As far as I understood, no. I meant different instances per point. when I feed the instance attribute with different objectpathes they do not show up in the viewport . only the object assigned at objectlevel shows up in the viewport

It's not impossible to do with the current architecture, but it would be a bit challenging. I'll submit an RFE.

Share this post


Link to post
Share on other sites

Crowds ladies and gentlemen, crowds! Most people agree that Massive is closer to a nightmare than a dream, no other software package does proper crowds, there is an increasing need for crowds, Houdini's procedural system will suit it perfectly, bring on those NOP's (Neuro Operators).

H13 == Crowds

So it turns out that 'no other software package does proper crowds', needs to be adjusted in to 'only Softimage seems to do an attempt'.

Based on this demo I would say that a Houdini worthy crowd system need to be a little bit more than a arc/viz game engine crowd. In my opinion they do get a point for trying though.

Edited by Nerox

Share this post


Link to post
Share on other sites

Snapping can use some love as well. I haven't tried it in H12 yet, but in H11 its far from Maya's snapping tools for instance. Primarily in terms of speed and it often doesn't do what I expect. Like being very jittery on empty areas's or suddenly very slow.

  • Like 1

Share this post


Link to post
Share on other sites
My personal beef on this is that we've already invented several methods to deal with these workflows and I'd rather like to see some consolidation rather than invent yet another system.

Agree with that.

Share this post


Link to post
Share on other sites
Guest mantragora

Show code generated by selected nodes in VOPS. Right now we can view all code generated by network, check it one by one or just play with reconnecting parts of network to see it. It would be helpful with more complicated shaders when you decide to rewrite some parts etc.

  • Like 1

Share this post


Link to post
Share on other sites

Modeling workflow that is simplified and doesn't create an entire network. I don't always want to use a non-destructive workflow! It would be cool to do multiple extrusions, fusing points, etc, within one node.

Share this post


Link to post
Share on other sites

Give CHOPs some love and proper integration into the animation workflow.

Animation layers.

Integrate/seamlessly interface COPs with the render view, and increase the 3D-post-processing features of COPs. Make COPs useful as what they are (a compositing module integrated in a 3D package).

Get rid of POPs and integrate them into DOPs.

VOPs as OGL shaders authoring tool, or add an OGL context to VOPs.

Cluster deformers.

I guess it's not very different from any list I had for the last 4 versions. SESI seems to be very determined to ignore CHOPs.

Share this post


Link to post
Share on other sites
And btw... Houdini is a professional software, and After FX is not really..

This is funny.

Try to Use [...] even Houdini Halo.

I guess this is trying to be funny as well, but failing.

Share this post


Link to post
Share on other sites

Ok so this is a far off dream, but there are VEX writers and Vop wirers.

It be cool if one could translate between both, making alterations both ways and it updating the other.

One pane is the vex code and the other is the vop network side by side. I could go vop happy one sec and code crazy the next second.

I know this feels like crazy talk, and there are a lot of issues, but if I'm wishing that be cool to see in v13 v14...

This would force some structure on the freedom we have, but streamlining the styles would be cool.

Share this post


Link to post
Share on other sites

13 could be a thousand-cuts build...

or fix COPs.

Share this post


Link to post
Share on other sites

>> Houdini 12.5.37: Fixed a crash in the triangulate2D SOP that may occur when the option "Remove Edges Outside Constraints" is enabled.

So I guess it'll be 12.5 at first :)

Share this post


Link to post
Share on other sites

Generally once a major release is released they will bump the development version up a notch so they can continue developing and bug fixing for a future release. Whether this ends up as 12.5, 13, 12.1 or some other number entirely will remain to be seen.

Share this post


Link to post
Share on other sites

Please please please a cancel button that works quickly, and in any situation.

Share this post


Link to post
Share on other sites

Please please please a cancel button that works quickly, and in any situation.

Esc - they even got the keyboard manufacturers to accept it.

or for they good old fashion terminal loving

kill -SEGV pid

Share this post


Link to post
Share on other sites

I would add something to the wish list:

Automatic conversion of used pictures in the scene to .rat format (as a background process or before you make a test rendering).

And if the picture have some information of their used color space .... automatic gamma compensation if I work in linear space!

Great if some parts of mantra could be accelerated by GPU!

More character tools: a kind of AutoMuscleRig for Biped and Quadruped or even Multi-Ped.

Something like Facerobot in Softimage.

Share this post


Link to post
Share on other sites

node with having multi output.

import attrib(vopsop) - handle prim/vertex/detail attrib with filtering option(as the attribpromote).

If I have parameter for pass the float attributes name. I want to pass them "someVector[0]" form.

I think viewport's "override color with attrib" display option is quite cool. so move it out to viewport.

Share this post


Link to post
Share on other sites

Wireframe Shader for mantra!

The fact that I had to read through forums to find out about the wireframe vop, I'm guessing introduced sometime in h11, is a notch in the annoying direction. The wireframe vop needs to be turned into a proper shader, with decent presets, so the novice or the experienced can find and apply it from the material palette and get a good demo reel quality render out of it.

Wire frame renders should not be a multi step setup. Or need a special render like wren(I hate maya vector too for this use). Also using any renderer with the primary control in sops, is very unintuitive, linewidth and width are not a concept for the novice.

Slap on a wireframe shader, and get a production quality wireframe, that is all I ask. This is along the same concept as this thread jumped off from of an out of the box car render that can compete with vray, and the dozen others like it. If I can't easily do it in mantra as easily in vray or mental ray with the simplicity and control, then why use mantra, or wren(Studio production level controls aside). Out of the box quality!

Edited by LaidlawFX

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×