To put it short, what is one supposed to be able to do with P?
I guess something like that would be used internally for projections, but there must be a reason for exposing and load/saving it?

eetu.

Or one step faster, use $NVTX!=4 as filter expression in a delete SOP

eetu.

Hmm how about doing the same group-by-parented-box -thing but instead of deleting, assign another material with low or 0 opacity? or scale down maybe?

eetu.

“Phantom” toggle, found in the render tab of the geometry container.

eetu.

Ah. Coming from lw + fprime and still doing much of my work in it, I can feel you
I'm working on a lw/modo -style ubermaterial and this was a good test, heh. (I'll share it after ironing out the kinks)

My first houdini vidcap: http://undo.fi/houdini/modo_mat/modo_ubermat.html [undo.fi]


(seems like I saved the image after changing hdr from the end of the vid)

Rendertime was 1 minute on an 8core, and aa is still a bit crappy :\
(gotta learn proper GI, too)

It has to be said, though, that this is exactly the kind of scene one should not time mantra with. I just rendered a landscape with ~200k displacement mapped trees, dispmapped ground, environment area light, motionblur and 9 x 9 pixel samples at 36s per frame - that's more in mantra's alley *grin*

eetu.

Looking at billowy_smoke :

In the first post the parameter node in question is “density” which, as i understand it, is used to get the “density”-named attribute from the actual volume. That parameter is also toggled as “invisible” and doesn't show up in the material interface.

When you replaced it with your own constant parameter, you got a volume of constant density.

Steve C has probably been modifying the “Smoke Density” parameter in the material interface, which is different from plain “density”.

eetu.

(hope i got it right..

Dunno about your problem, but I really like the image!

(I do understand that the client might disagree..

eetu.

This has been most helpful for me regarding alignment issues: http://www.sidefx.com/components/com_forum/files/instancing_273.jpg [sidefx.com]

eetu.

9.5.147

Ah. I didn't read your original post properly *blush*

It seems to me like the black defects might be mantra running out of reflection bounces.

Here's a couple of tests

And your original view, with an opacity-'fixed' white ball and 20 reflection bounces


eetu.

I added an isShadowRay test in the constant shader for the constant material used for the environment sphere. If tracing a shadow ray the shader now returns opacity of 0

eetu.

Another thing you could try is calculating point normals before fracture (select Point SOP -> Add Normals), and untick any “recalculate point normals” or “consolidate” -sounding checkboxes in operations after that.

eetu.

There's a related thread going at odforce, http://forums.odforce.net/index.php?showtopic=7208&pid=48387 [forums.odforce.net]

eetu.

stevecullum
..exporting obj sequences from Houdini, but I got no idea how you might go about this.

Drop a File SOP, put it in write mode and enter something like “myobject$F4.obj” as the name.

eetu.

stevecullum
..what if I want to create something new?
Lightwave's surfacing nodes are very intuitive and quite artist friendly, but Houdini's seem a bit tricky to understand. Maybe I've missed something?

Sorry to say, but with Houdini you need to roll you own, “from atoms”.

I too am often really missing Lightwave's node texture editor for the userfriendliness and speed of use. The high level (abstractionwise) nodes and workflow fit the overall fastpaced way of working I love about Lw.

Houdini on the other hand opts for total explicit control and configurability, which does fit the Houdini Way The power is quite obvious, pretty much anything you could code by hand, you can build with VOPs - unlike LW nodes.

I think one needs quite a chunk of r&d -time with Houdini to build a base material that fits one's work and which can be “easily” modified to suit the need at hand. (a.k.a. the dreaded/loved ubershader

(I'm still learning, so take this with grain of sand)

Is there any indepth tutorials out there that deal with the creation of surfaces?

Not really, as far as I've seen :\
Your best bet is to dissect the library materials. And read a couple of Renderman books

eetu.

Thank you for the prompt investigation!
Much appreciated, not something that many other vendors would do.

Seems like it's just one more reason to switch to linux..

eetu.

First I made 256^3 and 512^3 volumes and they worked fine. When i tried to step up to a 1024 volume, i3dgen errored out after a while (“can't create attribute: density” or something like that). It left behind a 2gb .i3d

After that, i tried stepping down to 800, 720 and then 640 - always the same error and a 2gb file. Then 512 again worked fine and resulted in a working 1.9gb file. I'm writing r g and b as well as density.

My source data is rather large, but I'll whip up a procedural example hip.

eetu.

Hullo,

I'm creating volumes from image data, and from the looks of it i3dgen can't create volumes larger that 2gb. (size on disk)

Is this a known limitation?

9.1.218 xp64

Maybe >2gig volumes are a bad idea anyway - but ram is cheap these days and inquiring minds want to play..

Btw, is there a real benefit in splitting something up into several volumes, or does the tiled block format handle things efficiently enough anyway?

eetu.

I'm pretty sure that's because expressions expect the argument to be in degrees, but vops expect them in radians..

eetu.

Hullo,

I was trying some learning exercises and ran into snags when trying to play with attributes. Simple stuff and a newbie point of view, but:

1) Primitive attributes in SOPs. Simply put, how do I set the value of a primitive attribute? You can modify values of arbitrary point attributes in a Point SOP, but not in Primitive SOP?
In my case I was trying to multiply “area” with clamped N. It was easy to circumvent by setting the alpha in Prim SOP to what I needed, or probably by promoting to point attrib and back, but it feels like there should be a way to just set the value of a primitive attribute directly. A “Custom” tab for Primitive SOP? Am I missing something obvious here?

2) Not that big, but accessing point attributes from a vopnet. I was trying to access a point attribute in a COP vopnet, and there is a primitive attribute node but no point attribute node. Perhaps it could be done via a point cloud, but for completeness' sake a point attribute node would seemslike a good idea. Unless I'm missing something again

eetu.