PDA

View Full Version : What are the biggest hurdles in wanting to learn Renderman shader writing?


Njen
12-14-2005, 02:05 PM
Knowledge for writing Renderman compliant shaders on the net is few and far between, and most sources of information assume one has some previous experiance in computer science or math, which can be hard for non-technical artists wanting to learn shader writing.

I am seeking to speak with artists who have a non-technical background, that had wanted to learn how to write Renderman comliant shaders only to be stumped by a problem that was seemingly trivial, but you could not figure out either on your own or by research. For example maybe you:Didn't know where to obtain a renderer?
How to install a renderer?
How to properly configure a renderer?
Didn't know what a rib file was?
Wanted to render something but couldn't?
Didn't know how to write a shader?
Didn't know how to compile a shader?
Didn't know how to apply a shader to an object to render it?
Don't know enough math to solve syntax problems in your shaders?
Don't know how to format shaders?
Never get what you expected from shaders?
I would love to know what problems people experiance in trying to learn shader writing for Renderman, no matter how embarrassing those problems are or have been!

For example, way back when I was attempting to learn shader writing, I read in the documentation for prman, that to render something, all you need to type is 'prman foo.rib'. Now I had no idea *where* to type that line in. I opened the prman.exe application program and typed in the line, only for it to give me an error. I tried all sorts of things until I eventually worked out where and how it was supposed to be entered.

I am seeking to speak with artists that have been through a similar situation to the above. Even experianced shader writers that can remember any similar experiances when they were learning shader writing are encouraged to get into contact with me! So please send me all of your stories about what part of shader writing you:Found difficult
Found lacking in information
Found information hard to understand
Simply could not understand!
I am looking for specific examples just as much as generalities. Please email me at:

rsl@motionmagnetic.com

Thanks for your time everyone!!!

ThE_JacO
12-15-2005, 06:12 AM
do you intend to write a renderman for beginners book or something?

Carina
12-15-2005, 10:40 AM
Do people with non-technical backgrounds generally involve themselves in writing shaders?

(actually wondering, not saying they don't..)

tweeeker
12-15-2005, 10:23 PM
TBH I found shader writing very hard to get into at the time I wanted to learn. At that time I was an artist with a non technical background. A couple of years later Advanced Renderman was published. I bought it, read it about 10 times, and then everything was much easier. I find it amazing when people say they want to learn renderman and rsl, but haven't bought a copy of ARMan.

T

-Vormav-
12-15-2005, 11:02 PM
Example source code. If you've worked with the language enough, then you can just read a simple explanation of nearly any function, and just go ahead and use it. But when you're new to it, sometimes you need to see some example code of how other shaders have made use of certain functions, or it just comes down to trial and error . Finding the source code you need isn't always so easy, though.
And if you don't have access to something like Renderman Artist Tools, setting up a pipeline for working with a renderman complient renderer can be a real pain in the ass.

Those are the main hurdles I've come across. Everything else has been relatively easy. Although I guess that if someone wasn't very good at math, picking up on how shaders work might be a bit more difficult (although the math involved rarely gets complicated).

rendermaniac
12-16-2005, 09:08 AM
Definitely the most useful information is practical examples. As well as ARMan a resource that I found incredibly useful was Steve May's rman notes http://accad.osu.edu/~smay/RManNotes/rmannotes.html as it actually take you through the process of achieving specific effects (even if pretty simple by themselves).

The RenderMan Accademy http://www.rendermanacademy.com/ tutorials pretty much cover a lot of the simpler stuff on your list and it's pretty easy to work out.

http://www.fundza.com/ also has good information - mostly to do with Slim rather than RenderMan in general.

The most pointless information is tutorials on setting up ambient occlusion and secondary passes. If you are searching for RenderMan tutorials then they will be what you find most of.

More examples of complex BRDF's would be useful as well as different ways of procedural pattern generation.

Simon

neuromancer1978
12-22-2005, 05:07 AM
(I know he said to email him...and I might...) ANYWAYS, I actually started getting Renderman a few years ago, once I realized that there were free Renderman compliant renderers out there. Grant you I still suck at writing shaders, and mostly rely on Shaderman to create them for me...got so used to a GUI. Of course I learned over time what does what in an RIB, and slowly learning SL. I am going to college next summer for computer science, intend on doing the Technical Director route, so learning programming is kinda important. Problem is I suck ass at math.


Anywho......

kpwproductions
12-23-2005, 07:39 AM
When I first started out using Renderman - what was most difficult was understanding how Renderman could be used as a real tool for 3D. I remember sitting in front of the documentation thinking "well, okay - RSL is basically C-style programming, no problem there... but, what do I do with these .sl files after I finish writing them?" After awhile, I learned that you had to compile the .sl files to .slo files in order for them to work... but then ran into the problem of how to then use the compiled shader files in my 3D package.

At least for me, it was a process of running into blocks in the road, sitting around trying to figure them out - then move on after the problem was solved. I did learn the process well through this, but many times just some clear information should have saved me a lot of time.

Clarity is what is needed most for proper teaching.

yadi2001my
01-04-2006, 02:07 AM
has anyone still got BMRT 2.6 that similar to RenderMan.. can u share wit me :)

neuromancer1978
01-07-2006, 04:19 AM
I wouldn't even try to use BMRT sir.... it's an obsolete program, there is no support for it since Pixar sued the programmers. If you want to use a Renderman compliant renderer go with AQSIS or Pixie. They both are very good free renderers. Pixie is not as updated as often as AQSIS but Pixie does have all the features of BMRT. AQSIS is updated often and I use it heavily. I actually use AQSIS to write shaders, use it as a preview renderer. I use Pixie with Maya through Liquid to make final images.

Again...forget about BMRT. The only thing it is usefull for is to be a collectors item someday.

K. Scott Gant
01-11-2006, 10:20 PM
Pixar never sued the programmer of BMRT over BMRT. It was one guy, Larry Gritz, who wrote it as a thesis project. Larry went on to actually work for Pixar after making BMRT for a number of years leading their rendering research group, and is also the co-author of the book mentioned above, Advanced Renderman.

Larry stopped maintaining BMRT because he never really intended it to go on as a product anyway and quite frankly it was slow and not a high-quality renderer by todays standards...but that's more my opinion than anything, your miliage may vary. Pixar doesn't sue just people that make Renderman renderers as there are a number of open source projects right now and also commercial products that adhere to the Renderman specs. Larry left Pixar out of frustration into where the software was heading along with a couple of other programmers and started their own software company, to develop the new Entropy renderer...which itself is a Renderman renderer.

Here's where the lawsuit comes into play...Pixar sued them all for misappropriated trade secrets they had aquired while they worked for Pixar. They settled out of court and the terms were the death of Entropy and the all-but-dead-anyway BMRT.

It's a real mess of a tail...and I used to know a website that had a timeline as to the downfall of BMRT and Entropy, but I can't find it now.

But anyway, look into the open source Renderman renderers. Pixie and Aqsis leap to mind.

neuromancer1978
01-12-2006, 06:05 PM
Pixar never sued the programmer of BMRT over BMRT. It was one guy, Larry Gritz, who wrote it as a thesis project. Larry went on to actually work for Pixar after making BMRT for a number of years leading their rendering research group, and is also the co-author of the book mentioned above, Advanced Renderman.

Larry stopped maintaining BMRT because he never really intended it to go on as a product anyway and quite frankly it was slow and not a high-quality renderer by todays standards...but that's more my opinion than anything, your miliage may vary. Pixar doesn't sue just people that make Renderman renderers as there are a number of open source projects right now and also commercial products that adhere to the Renderman specs. Larry left Pixar out of frustration into where the software was heading along with a couple of other programmers and started their own software company, to develop the new Entropy renderer...which itself is a Renderman renderer.

Here's where the lawsuit comes into play...Pixar sued them all for misappropriated trade secrets they had aquired while they worked for Pixar. They settled out of court and the terms were the death of Entropy and the all-but-dead-anyway BMRT.

It's a real mess of a tail...and I used to know a website that had a timeline as to the downfall of BMRT and Entropy, but I can't find it now.

But anyway, look into the open source Renderman renderers. Pixie and Aqsis leap to mind.


Ah.... I didn't know the specifics of the lawsuit. There doesn't seem to be much info on this piece of history anymore, maybe it's like one of those deep dark family secrets... do not speak of it, but we all know it happened. I am aware of the open source Renderman compliant software, I myself use AQSIS and Pixie on a regular basis. I just thought BMRT had proprietary code from PRman itself. Like I said I never got into BMRT much, and when this lawsuit came around I was not aware of it untill way later, so I guess that was ignorance on my part, reading info from unreliable sources.

Anyways. I myself just bought Rendering for Beginners. I had been wanting to get it since August when I spied it in a bookstore shelf. I guess it's a late Christmas present to myself. I was up till 2 AM reading it, great source of material. I am no newbie to SL, but I still have MUCH to learn.

rendermaniac
01-12-2006, 10:06 PM
There is no big secret - there is more information about the Exluna vs. Pixar case here: http://www.renderman.org/RMR/OtherLinks/blackSIGGRAPH.html. This is also why Nvidia's Gelato cannot be RenderMan complient. I don't think Pixar had a very good case, but they definitely had more money at the time.

Ironically the Gelato API is a lot more innovative than the RenderMan spec. The main thing crippling it is the need for at least an FX based graphics card. You aren't going to be able to shoehorn one of those into a blade renderfarm node anytime soon!

Simon

playmesumch00ns
01-13-2006, 09:55 AM
Yeah it's such a shame, as a developer the API's wonderful. The shader linking stuff in there's well thought out too :) I see aqsis have implemented something similar.

CGTalk Moderation
01-13-2006, 09:55 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.