Renderman is dead!! Or at least on its way to die!!


#41

Saying rasterization or raytracing is slower or faster than the other is inherently flawed.
Some things raytracing will be way faster at, and viceversa.

When you need indirect contributions, or raytracing eyes, both things that happen in nearly every-single-shot out there in a creature show, bruteforce is already faster than bolting a raytracer on top of a reyes engine or spending a day computing point-clouds and storing 20GB worth of data per frame before you can even stand a chance to render the final look for that frame.

There are other things too, such as very dense meshes (sometimes unavoidable) falling into a single bucket. A raytracer, once the acceleration structure is built, will just run over it like bad roadkill, REYES will still grind to a halt on that bucket before it even starts thinking. Conversely, sparse cages with lots of disaplacement coming close to camera will always be faster in REYES rasterisation engines than when you have to compute all the new geo beforehand, store it, then raytrace it, and send the acceleration structure into chaos memory wise half the time.

It’s not quite as black and white as “raytracing is slower” these days in production. I don’t know why people keep harping on that string, but I suspect most never tried one of the good ones :wink:


#42

20GB per frame?? wow, some major computing work there. Can you tell how much disk space is used these days in CG production for a feature film?? lets say in “Legend of the Guardians”??, i’m just curious about those numbers.


#43

Yes, PRMan 16 is not your father’s RenderMan. Just because it hasn’t gotten a lot of buzz in these forums doesn’t mean that it’s not moving. Physical shaders, progressive ray tracing, it’s there. You can bypass the entire REYES process if you like.


#44

that is all possible with mr for example, i used alot of these things by myself.

but what you and other wrote about the complexity, thats true. you can throw in all kind of complex objects, hair, subdivs, dispmaps, alot of lights and all is rendering without any problem. lots of furry character with deep shadow lightrig is working without any problem, rendertime wise and also ram usage wise. i would get alot of headaches if i have to render this with mr.


#45

Mantra is used by a lot of larger studios to render the effects elements in shots.

One thing to remember though, is unless you have the skilled people to write the shaders your going to nonsense coming out the other end not matter what render engine you use.

:slight_smile:


#46

Mantra is used by a lot of larger studios to render the effects elements in shots.

And some other places use it to render everything.

Anyway. During the last 15 years I heard “The end of PRman is nigh!” pretty often. Let’s look again in 15 years. I say it will still be around :wink:


#47

As a frequent Mantra user, I think that Mantra and Houdini (for lighting and scene pipeline) is underrated


#48

Amen. THAT is the key feature of renderman standard, you really can do with it whatever you want…I don’t know of any renderer offering this. Adaptivity is the key to everything (evolution? hello!) and as you can do everything, adaptivity is the very strength of it.


#49

As an individual who uses PRman for Maya, I can tell you 2 things:

a) It’s a b*tch to setup if you’re not familiar with RSL (which limits alot of things you can do without the user friendly SLIM)

b) it gets the job done, no matter how complex the scene is.

The latter is why i’m sticking to renderman, and will probably continue to do so for a long time :slight_smile:


#50

Definitely, but you’re still kind of screwed if you don’t have a TD to jump in and make stuff work.


#51

Dear Bjoern,

I am not going to answer to your main question : whether PrMan is dead or not is not the point.

Obviously you still have a very naive view of rendering. As other people mentioned it before me, no renderer is perfect. Magic bullets don’t exist. People have to work hard to make things look good.

If you find it hard to work in a big company pipeline, you might want to consider working in a smaller company. I completely understand this may not be your cup of tea, but complexity comes with the territory.

All the best.


#52

haha, love rants. :smiley:
I use and used a ton of rendering engines, I have to say that the advances in raytracing in recent years have been truly revolutionary. if there’s been a revolution in 3D lately… this must be it (follows the real-time revolution, can’t wait :D).
I used to be a huge renderman fan, but let’s face it, for most people out there, small studios, freelancers, etc, renderman is overkill. it’s just too damn hard and time-consuming to setup things, especially lighting, to look awesome. yeah, renderman has stellar AA and is a fast reyes renderer, but a it’s a slow and inefficient raytracer, despite all the blah blah about it being one of the best raytracers. if you use renderman as a simple user and don’t have time to “aggressively” optimize the scene, shaders, lighting, raytracing crap, etc, you will not have a fast render.
so… let’s take vray for example, which to me seems to be an extremely efficient raytracer: no more huge time to setup lighting and optimize manually, which, for a small or a one-person studio can make the difference between can-do and can’t-do. it’s still a very fast and flexible renderer, translates geometry (in Maya) faster than any renderman I’ve used (which means the actual rendering starts sooner), has proxies that load unbelievably fast, way faster than mental ray’s or 3delight’s for example, and it’s especially… natively optimized for efficient raytracing. almost instant GI/AO… that’s pretty damn impressive.
give a large scene to renderman and you will wait forever for an AO or God forbid… area light shadows. also, all those point-based solutions are still slow to render and time-consuming to setup.
now… in a large production, like for a feature, which I’ve never worked in, things are probably different and the power and flexibilty of a prman can make a difference. for now. :twisted:


#53

Displacement and motion blur come first to make it realism.
Renderman renderer can do it best.
Time and money is not big deal for big studio.


#54

The “revolution” that has happened was cheaper, faster hardware. There’s nothing that we render today that we couldn’t have rendered ten years ago, it’s only that ten years ago it would have taken you days instead of minutes. Exponential growth FTW!


#55

The day anybody pulls something like “Avatar” with a different render will be the day we can discuss if Renderman is dying or not. All the major film studios that don’t use their own in house developed render engine use Renderman. Hardly a dying product. :slight_smile:


#56

can a raytracer render out Avatar in that time? can it render the worm robot from Transformers 3? can a raytracer render out TinTin?


#57

One renderer to rule them all!!

…And that one is Arnold! :wink:

Joking aside, and even though I do believe Arnold is the strongest renderer on the market BY FAR. There is room and a place for ALL renderers. It’s great to have competition in the market place. Look what Arnold did for prman 16 :slight_smile:

Having said that, I also agree that any renderman renderer requires way too much TD work to be usable for any mid to small shop. Yes, even v16.


#58

generell, i am pretty sure you can, question would be how much time you have to spend to seperate the shots in a way that you can render it with raytracing only. renderman needs also this seperation, but because of the heavy assets, raytracing would need much more tuning here (or seperation). but you wrote “in that time” for avatar, that would be interesting, but we have to know how all the renderings were archieved. if they really uses only reyes without raytracing for all the stuff, then i would say no, its not possible with raytracing.


#59

It took both a raytracer and PRMan to do Avatar:
http://www.tgdaily.com/games-and-entertainment-features/45279-nvidia-unveils-pantaray-engine


#60

I still can not understand how a small studio/individual can afford Renderman price. $2000/node.