Houdini 11 Sneak Peek

[danilo2]

I'm wondering only, what exactly is multithreaded now.

RBD, fluids, fur ?

On the demo video, fracturing tool looks great, but the fluid simulation isn't "super cool" (for example in comparision to RF).

Appart from transfer sop - what is multithreaded now?

[anamous]

I don't feel that the FLIP/PIC demo scene is really representative, both rendering (amount of motion blur on the geo) and simulation wise - 270k particles isn't close to enough in terms of the implied scene scale. And the FLIP solver is made for much higher particle counts (think millions). The demo video is supposed to demo POP control over FLIP simulated particles, not so much the advantages of FLIP itself.

[mrice]

It is an actual light source which means that it contributes to direct lighting and gets sampled much better than the old emission approach. It also evaluates the shader on the source geometry to determine the light amount/tint. In other words, it's pretty neat.

Sounds great, can't wait. I hope I'm right in thinking that the improvements behind this will be applicable to much more than mesh lights

[danilo2]

I don't feel that the FLIP/PIC demo scene is really representative, both rendering (amount of motion blur on the geo) and simulation wise - 270k particles isn't close to enough in terms of the implied scene scale. And the FLIP solver is made for much higher particle counts (think millions). The demo video is supposed to demo POP control over FLIP simulated particles, not so much the advantages of FLIP itself.

Ahhh I hope so! I cannt wait to check it out

Do somebody know if rbd solvers are multithreaded now too ?

[Macha]

Not sure cos I haven´t tried them but the new Mantra is much faster

Interesting. I never felt it was slow in the first place. Compared with MR it's very fast, especially with motionblur and DOF. The only bit that drags is starting the initial render window but I always thought that was a windows problem. What kind of stuff is it faster with now?

[danilo2]

[...] The only bit that drags is starting the initial render window but I always thought that was a windows problem. [...]

starting window is windows problem (on linux it shows up immidiately).

Edited June 6, 2010 by danilo2

[Stalkerx777]

starting window is windows problem (on linux it shows up immidiately).

MPlay starting IS windows problem, but easily can be fixed by providing DNS addresses manually. If DNS obtained automatically via DHCP, it will cause long MPlay start.

About Mantra render. Yes, of course Mantra is faster then MR, mainly just because it is "reyes" renderer. MoBlur, DoF, Displacement are strong sides in micro-polygonal rendering. But don't forget about heavy refl/refr and GI, witch is strong side of "ray-tracers" like MR. So i really like mantra,but,be honest, its not so fast in some areas as it could(and should) be. So, waiting for new mantra 11, and i hope, that someday, i can finally say "Good by MR".

[lisux]

MPlay starting IS windows problem, but easily can be fixed by providing DNS addresses manually. If DNS obtained automatically via DHCP, it will cause long MPlay start.

About Mantra render. Yes, of course Mantra is faster then MR, mainly just because it is "reyes" renderer. MoBlur, DoF, Displacement are strong sides in micro-polygonal rendering. But don't forget about heavy refl/refr and GI, witch is strong side of "ray-tracers" like MR. So i really like mantra,but,be honest, its not so fast in some areas as it could(and should) be. So, waiting for new mantra 11, and i hope, that someday, i can finally say "Good by MR".

I have done jobs in Mantra using mostrly ref/rfr, lots of water with displacements, in studios that usually only use MR.

Mantra was faster and we used it instead of MR.

In our case it was justa matter of shading tweaking to get better render times than MR.

But anyway I agree that Mantra needs more speed for rautracing.

Macha, PBR is faster now.

[cabra]

I'm sure there will be more news. Anyone know or have any idea or think of why the name "uber" the new things?

I feel that when people argue they need more speed I think you are wrong. go behind a ghost! of a trend, a haircut. Practically has been transformed into new tone of voice. In some cases I think it important to improve these aspects but not in those of us who we are dealing. Houdini can not go higher than it was years ago. It is already too high!:

Edited June 7, 2010 by cabra

[Macha]

I'm sure there will be more news. Anyone know or have any idea or think of why the name "uber" the new things?

Don't know why but that word has become fashionable in English in recent years. It's the german "über", meaning over/hyper/extra. In this context it refers to an all encompassing shader I suppose.

[pclaes]

Anyone know or have any idea or think of why the name "uber" the new things?

It just means that whoever is working on the shaders/materials in houdini is in touch with some of the (houdini) shader writers in industry and/or perhaps reads the forums. "uber" only applies to the shader, not to all new things. I don't know when the term "uber" first got mentioned, but it became equivalent to "all encompassing" material - just shorter.

I feel that when people argue they need more speed I think you are wrong. go behind a ghost! of a trend, a haircut. Practically has been transformed into new tone of voice. In some cases I think it important to improve these aspects but not in those of us who we are dealing. Houdini can not go higher than it was years ago. It is already too high!:

With multiple procs in all machines these days, speed is only the consequence of parallel architecture. When people ask for "speed", the way I interpret it is they ask for an architecture that allows to use the power of all their processors. Requesting for a feature like "speed" is actually a huge deal as it will make the software future proof, but also requires major rewrites and rethinking of some algorithms. Speed is and will probably always be on the wishlist, so in that sense I do not think it is a trend. People are just more spoiled with real time physics in games or gpu accelerated calculations and are wondering why their powerful workstation and software should not be able to achieve similar speeds.

I can not speak for the rest of the world, but the amount of houdini users in London has definitely increased over the last years and in the CG survey houdini's future expectations and potential was listed very high as well. So you can be sure houdini will go higher.

[pclaes]

jinx...

[cabra]

for me they used the word "uber" as a joke. (And it's amazing)

I said that I agreed on that the thing ought to be faster but the special considerations of any users is a very complicated issue that users of Houdini as any other could not resolve sudden blow.

Consider it right that people have different interpretations and consider one thing over another, but one above all others are very stupid.

If the thing right now is not as fast as you want then it is better then silence or what can not speak better shut up. ji ji.

one of the reasons why people prefer to avoid use houdini is the difficulty and speed. and the truth that neither is true. Where does that?

It is not uncommon for people to be misled by the comments and it would be really great that experienced users keep silent on issues that can not be resolved with words.

maybe my English is not as elegant as ordinary people who usually use. sorry not my native language.

Houdini for win!

Macha yoooouu rocks!

[mightcouldb1]

Bullet solver pleaseeee

[old school]

Looks like a fun release . I'm curious about the new volume sops - if they made it in.

Yep.

Yeah, I know, all the new big ticket stuff, but the day-to-day is what counts as much and the new Volume SOPs are more fun than a barrel of monkeys.