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

#1

I (have to use…) use PRman it on big feature productions… cause thats what the studio is using.

[b]PRman and its Propblems[/b]
Why do I think renderman is dead?! Or is on its way to be dead?
 this hole baking mess in Prman sucks bigtime... not to mention all the  other pointbased features. And the lack of a good commercial implemention in to maya and all this custom tool sh**t makes the daily work an absolute nightmare with PRman.

 I can see realy no justification to use Prman anymore since arnold is  hitting finaly/soon the market. And even bevor... Vray/Finalrender (or  MR with the coreplugin) and have a solid setup for motion-blur and DOF in comp. Use Normalmaps where ever you can and don't overdoe displacement...

Of course there are situations where the comp hack does not hold up. Or Discplacement gets critical.
But most of the time its fine and the audience will never ever notice! And thats the bottom-line which matters.
And NOT if a professional sees a difference!

 Well our "lighting,lookdev / solving technical issues" ratio is about 30% / 70%
In my world before entering the big studios I consider this an absolut JOKE. Allot of times at work I want to hit my head against the wall and think: "DAAAAAM, give me a standart Maya and Mr and I solve the shots in 50% of the time..." and trust me I'm not a big fan of MR....

Anyway in the studio that I work (don’t want to piss people of, so I don’t name it…) I have allot of coworkers which had been befor at disney, dreamworks… framestore… and so on…
They are all very happy for some reason with the setup in our studio.
They even say that it is in most studios worse… WWWTTTTFFFF???

[b]My main opinion[/b]

If you need a huge R&D department to make the renderer usable for an “Artist” and after that have only a “lighting/problem solving” ratio of 50/50% or worse… then face it. Your pipeline/render is a absulut shi***tt! period.
When I work with Vray its 80/20% lighing/problem solving, or better!
And I don’t need a TD ever to support me if problems popup

[b]Opinions around me[/b]
I have also allot of friends in other studios which (have to) use Prman as well... and I get most of the time the same respons...
Please give me a f******ng arnold/vray/finalrender whatever where I don't have to worry about pointclouds,Deapshadows... and all that shit...  and I can Just work.

Sonypictures basically nailed it:" Human manpower is has a higher cost than CPUpower" Why not make the Artist life easier... get prettier images and keep the cost the same.

 [b]feer[/b]

I think there is also a big problem of a huge amount of TD’s and R&D people who don’t really want a change from there Rey-eyes pipes to full pathtracers… or ray-tracer…
cause they use it for decades… and if you just use a render that works… for the artist, you don’t need the TD’s anymore…or less ( I actually got some similar comments from coworkers who are in this position…)

What do you guys think about this…? And How do you think PRman or 3DElight will aproach this competition?
I don’t want this to be a software war! I not a fan of any renderer and I have work with all of them… I’m writing this simply out of frustration.
I’m an Artist and thats how I want to work.

#2

Wait, what ?! :surprised
I’m not a TD but I love the Mantra render inside Houdini. So you see you don’t need to be a TD to appreciate a REYES render. Even though I’m not able to take full advantage of Mantra because I lack technical skills I still prefer it over Mray or other renders.

#3

I don’t know enough about houdini on production level. But from what I have seen and heard houdini is in generell far supirior to other packages on the market. And the Implementation of Mantra is years ahead from MR or PRman in maya…

#4

Yet its not the most popular app for the same reason the PRman is not going to die any time soon…

Its just to entrenched in industry for it to disappear off the face of the earth any time soon even if there are better tools and technology…

I’d say of the next few years PRman will get features and workflows similar to arnold etc…

#5

Adrian, I’ve looked at your PF. It’s awesome and shows clearly your growing up from the start to now.
I’d like to know if you have anything rendered in Mantra and if you’d like to show me.
I’ve tried Houdini for a while, wonderful tool, but too technically oriented for me. I’ve never been able to obtain anything decent from Mantra, till, frustrated, I gave up. My fault for sure.

#6

Wow, that’s quite a rant there!

You’re quite correct of course, PRMan is a terrible renderer and people only use it because they’re afraid of change. I don’t know why we use it to be honest.

Oh yeah, that’s right, because I’m not rendering car commercials.

#7

They already are - have you seen the features that have been added to prman 16? There is a lot more physically based shading. I wouldn’t discount it yet.

A lot of this boils down to the rendering interface. I am really looking forward to trying out Katana as it should provide a lot of tools that are used in production out of the box. Plus it should make mixing and matching renderers a lot more simple.

Simon

#8

I would like to know how well Arnold deals with Fur. In my tests at least, Vray struggles with it. I can’t quite see something like Legend of The Guardians been done in a ray-tracer (yet). Happy to be proven wrong.

#9

Technique X is dead long live Technique Y.

#10

Do you mean fur for maya? Because 3ds max has harfarm http://www.cyberradiance.com/hairfarm/, which is quite a good hair solution, though a separate plugin.

#11

Well everything is difficult when you dont have motivation to learn and improve your skills.
I rarely use PRman, but its far from dead, and it is great tool, 3delight as well.

#12

bjoern ,I hope you have used Arnold before writing about it,no doubt it’s a good renderer but it’s nothing like it doesn’t have it’s issues.Sony can easily implemented it b’coz they have resources (both Technical and Financial) for it .It’s the similar case like Blue Sky’s renderer after watching their movies you can easily come up with one conclusion that it is very good but unfortunately it is not like that.
All that matters is the resources if you have them you can make a bad renderer looks good and vice versa.
I thought you are some lucky fellow (right now I am jealous of you) who gets to work with so many highly qualified people and believe me they have more experience and expertise than you, if they think that Reyes is best for the your current production than it is.
Every package has it’s negative as well as positive points it is how well you use it.I have worked on one such production where we were 6 months behind schedule and then we shifted to Reyes based renderer and we finished a month before deadline.I hope you got my point.

BTW if your studio is in Nepal than you really are one lucky fellow .

#13

LOL! It’s not going away.

All renderers have their faults. Lets just say there were some thing in Arnold that I thought was better in say Mantra or Renderman with the proper setup. The key really is explicit control. The strength of prman is that you have that control but you have to implement it.

I’m starting to see why prman is so widely used in film. It’s ability to handle large scale data and scale that up is second to none. It covers every primitive with mature approaches towards rendering them efficiently, including volumes and now point clouds.

I’m seeing there are gaps in functionality within MR, Arnold, V-ray, and whatever other renderer is out there that every single-user desktop warrior raves about. Some rave without even putting through a production encompassing the myriad of challenges that comes with that!!!

I’m not saying one is better than the other. Sometimes it’s nice to click a few buttons and solve your problem that renderer xyz solves well. The difference is what do you do when the problem is beyond the scope of your renderer? Wait for the dev to update (could be waiting for nearly a decade in MR’s case), or soldier on and roll your own solution?

PRman is here to stay. The best looking trailers in regards to quality were ILM or Weta’s and they were all rendered using renderman.

-Lu

#14

The first rule of Arnold: You do not talk about Arnold.

#15

I’d love to know more about how Arnold works under the hood. Renderman is well documented and I personally love how it handles scenes. I really wish 3delight, or any renderman translator, was available for 3dsmax.

#16

displacement would make a big difference in this shot

#17

And they even didn’t need raytracing for this… why not to use displacement then?

#18

Do you mean fur for maya? Because 3ds max has harfarm http://www.cyberradiance.com/hairfarm/, which is quite a good hair solution, though a separate plugin.

If we had to use a seperate renderer for Guardains to just do feathers, I could imagine it being quite painful.

#19

OP seems to confuse REYES and rasterisation with forward path tracing, raytracing etc. with the difference between PRMan and VRay, as if things were always mutually exclusive.

Rasterisation is due to take the back seat to raytracing and bruteforce at some point, REYES might or might not become too much of peg leg to support that, but none of it means PRMan as a product is going to die soon (it sure as hell isn’t dead already).

Have a look at PRMan 16 and the roadmap for future releases, they are catching up to the above, and know they need to to both keep their movies looking on par or above par, and to keep it a viable product and enjoy the high refinancing of their RnD they have enjoyed insofar.

The TDs, shaders, and pipeline points aren’t really worth commenting on.
The moment you confuse renderman (a standard) with PRMan (a product) and even worse think there’s such a strong coupling between technique and brand or technology and standard description, common sense and credibility take a jumping leap out of the window.

Forest for the tree IMO.

P.S.
No, not a renderman and even less a PRMan fan.

#20

Different renderer’s are around for different purposes. PRMan was designed for film use, others such as Maxwell were not. Renderman is far from dead and if anything gaining popularity simply because back in the early 90’s when Renderman pretty much redefined visual effects, nobody heard of it. Now that Pixar is a mega house, as opposed to the early 90’s when they were developing software and making commercials for Listerine, it has gotten a lot of attention. Just web presence alone is a huge factor.

Studios that roll their own renderer do so because it works in the pipeline, it was built for their films so they know what it is capable of, can add to it, optimize it and fully push it to it’s limits. They do not need to pay for license fees, technical support, or the upgrades when they come about. It’s all internal so that cuts out that expense. They are also years worth of R+D before anything is in “production status”. This is what separates Pixar, Dreamworks and Blue Sky. They all do the same thing, render frames fast and with the highest visual quality. They differ in the code. When it comes to code and how different the rendering spec’s are, it does not matter to the general public because they will never know the difference, let alone why. Even if explained, only us CG geeks understand let alone get emotional over render times, memory use, cache types, displacements, lighting methods and so on and so forth.

Renderman has always been about getting the best visual quality vs. render time for film and until the added ray tracing and point based stuff, has relied on tricks to get that photo real quality. Jurassic Park allowed us to believe that those dinosaurs were real, that was at Renderman’s early years really, since it has been nearly 20 years since that was released. There was no global illumination, ambient occlusion, point clouds or any of the other illumination methods have come since. That was artists and technicians working to get the best look for the film, for the shot. Thousands of man and computer hours for a second or three of visual effects, just in the rendering end. It may not have been physically accurate, in terms of rendering and lighting but it made you believe for a brief moment in time that what you were looking at was not computer generated. Then the next 20 years of CG visual effects happened and Pixar had to keep pushing the envelope, improve the speed and capabilities, features and ways to improve the lighting while at the same time keeping render times as low as possible which is what PRMan was known for.

Think about this, when lighting directors are making decisions to set up lighting, in real life not CG, they commonly fake lighting situations to accentuate the look of the film. This is no different than Renderman technical directors using various tricks or hacks to achieve a look, even if it is not physically correct lighting, which in certain renderer’s is how they operate. While it may be accurate and very useful for arch viz, this is not the case when it comes to film frames where time literally equals money. Arch viz really requires realistic, physical lighting because well we live the real world where light defines everything we see, lighting can affect mood so these people that design such structures need to see how light will affect the visual aesthetics of said structure. Film directors want to achieve the same but in a dramatic and emphasized way and if it has no basis in reality well be damned with realism, we all know this.

Renderman is far from death, it just finally has a little competition that is all and in the visual effects world, a tool is a tool, even if you need to use another renderer for a shot, then so be it. In the long haul however, Renderman has proven itself over and over that it can get the job done, even if you need to hack it, it can get the job done fast and still has much room to grow. All the others are newbies comparatively.