Maya - Houdini - Fur

[haggi]

Hi,

We have written our own hair and fur rendering tool for our last project. It has no GUI yet because we simply didnt have the time do do it. Now we will start soon the next project and it is time to think about extending our tool.

I had a look into houdinis hair tool and there is a masterclass video on the houdini website about the hairtool. The backbone is very similiar to our approach and if we would rewrite our inhouse tools, they would look very much like the ones in Houdini.

Now it is never an excellent idea to invent the same thing twice. So the question is:

Is is more useful to invest some development in a stable maya/houdini pipeline and rely on two products, or is it better to continue our work and stay with only one software and concentrate on new features and a GUI? What do you think?

[Allegro]

Hi,

We have written our own hair and fur rendering tool for our last project. It has no GUI yet because we simply didnt have the time do do it. Now we will start soon the next project and it is time to think about extending our tool.

I had a look into houdinis hair tool and there is a masterclass video on the houdini website about the hairtool. The backbone is very similiar to our approach and if we would rewrite our inhouse tools, they would look very much like the ones in Houdini.

Now it is never an excellent idea to invent the same thing twice. So the question is:

Is is more useful to invest some development in a stable maya/houdini pipeline and rely on two products, or is it better to continue our work and stay with only one software and concentrate on new features and a GUI? What do you think?

I'm not particularly fluid with hair simulations, so I can't fairly comment on that particular aspect... but frankly my blood pressure rises whenever I'm using Maya. Little things like how my particle simulations are different every time I simulate even if no parameters are changed... and not being able to change things like per-particle scale or opacity after you've simulated... or being able to remove a single particle if everything looks nice except for one distinct stray particle.

I find Maya to be so incredibly restrictive that I would highly encourage you to invest some time to further explore all of the advantages that you could get out of Houdini, not *just* for the fur/hair part of your work.

[dayvbrown]

Hi,

We have written our own hair and fur rendering tool for our last project. It has no GUI yet because we simply didnt have the time do do it. Now we will start soon the next project and it is time to think about extending our tool.

I had a look into houdinis hair tool and there is a masterclass video on the houdini website about the hairtool. The backbone is very similiar to our approach and if we would rewrite our inhouse tools, they would look very much like the ones in Houdini.

Now it is never an excellent idea to invent the same thing twice. So the question is:

Is is more useful to invest some development in a stable maya/houdini pipeline and rely on two products, or is it better to continue our work and stay with only one software and concentrate on new features and a GUI? What do you think?

Hi there,

we use a dual (sometimes triple) software setup, we recently did a job with fully fured mammoths which were modeled in max, animated in maya and simmed and rendered from houdini. Having done a bunch of fur jobs in the past i've got to say houdini's fur was the most painless experience i've ever had... and this was for a fairly painful client

Rendering was through mantra and was predictably stable and wonderful, simming through dops was also pretty reliable, some minor shenanigans but nothing to get too upset about. Generally memory or other not-really-houdini's-fault issues.

We used fbx to transfer cached meshes (so we aren't dealing with rigs in houdini) and have a reasonably painless time with it. As long as you've got a nice rest pose to add all your attributes and groups for copying to your animated mesh you'll be set.

Hope his helps,

dave.

[haggi]

This helps indeed. Thanks.

[Mzigaib]

How about speed? I did some minor tests, of course I know very little about fur on houdini, but on my tests a simple ball with fur and GI was rendered very slowly.

I do know there must be a way to optimize this, any tips about that would also be welcome.

Thanks.

[symek]

How about speed? I did some minor tests, of course I know very little about fur on houdini, but on my tests a simple ball with fur and GI was rendered very slowly.

I do know there must be a way to optimize this, any tips about that would also be welcome.

Thanks.

the challenge is to adopt basic fur shading pipeline to your needs. Fur renders fast in a basic three lights / deep shadow maps scenario. It might get complicated once you want to add some sort of global illumination to that (by baking it to maps/attrib etc) or by developing more advance shading models. I don't have much experience with other fur toolkits, but rendering speed was always the least concern in Houdini's case. More likely styling or rare cases you reach the limit of ram. Basicaly you can render a fury rabbit with ~60K fur density per unit, in close up, in HD, full quality (sampling 32x32) in less then 15 minutes per frame. Fair enough.