• Interview with SplutterFish
    18 September 2002 | Leonard Teo
     Image: One of the beautiful images adorning the SplutterFish gallery, by Calibre Digital Pictures.

    SplutterFish has finally unleashed Brazil r/s v1.0 upon the 3D world. In development for a year and a half, the Brazil rendering system has gotten a large interest and following from the 3ds max community through its open-testing philosophy, where users could download latest test software for their own personal testing.

    It all started at Blur Studio, where Steve Blackmon and Scott Kirvan had been coding and testing various rendering technologies for production purposes. In 2000, Blackmon wrote the beginnings of a new core rendering engine with global illumination (GI) experimentation as a render effect. Called “Ghost”, the plug-in was made free for download on the Blurbeta website, a Blur Studio web repository with dozens of invaluable production tools.

    Steve Blackmon and Scott Kirvan, the original founders of SplutterFish (photo: Leonard Teo, SIGGRAPH 2001)


    As the 3ds max community started using Ghost, the response was phenomenal. Images poured in from all over the world -- beautiful pictures that displayed new levels of lighting and photo realism that just wasn't possible from most renderers at the time. In early 2001 Kirvan and Blackmon formed SplutterFish as the vehicle to turn Ghost (later named Brazil r/s) into a production-grade rendering system.

    We speak to Scott Kirvan about the release of Brazil 1.0 and Splutterfish’s involvement in porting Joe Alter’s Shave and a Haircut to 3ds max.

    3DF|CGN: Brazil has been in development for quite a while. How did the project end up sprawling out over such a long period?

    Kirvan: This is really an interesting question. Actually, Brazil has only been in focused development for about a year and half. It has an extremely rich feature-set, created with modern software design methods from the ground up. Brazil has many of the same features you expect to see in high-end renderers that have been around for 10 or 15 years. From my point of view, I think it's really astonishing that something this complex was written so 'quickly', especially when you see what it can do and how stable it is, based on its performance and stability. It doesn't perform like a 1.0 piece of software. We wrote Brazil to be used in production, by people who require reliable, functional tools to do their job. As a new company, we feel it is imperative that we establish our reputation as a company with high standards, and quality software that artists can depend on.

  • 3DF|CGN: Many other companies have or are attempting 3ds max renderers too such as FinalRender, VRay and the late Entropy. What was the impetus for developing Brazil as a plug-in renderer for 3ds max?

    Kirvan: For a long time Steve and I had been searching for better a production renderer for 3ds max - one that worked with the existing toolset of plug-ins, one that could be counted on to finish the job, one that could produce stellar imagery, one that was usable and controllable by artists and not just Technical Directors. We had tried everything out there and there was nothing living up to what we needed. Had there been, we never would have taken on this development challenge. Even back in 1998, we were playing with different rendering technologies including GI, but it was in 2000 that the real work started and the huge buzz about GI rendering in 3ds max got going. Once we got into this, it became obvious that we were not only serving our own needs but the needs of many artists who were working under the same constraints. The impetus was and still is, the need for real production quality rendering in 3ds max - quality, performance, reliability, etc. Brazil is the only renderer that has taken that mission seriously and is the only renderer available that truly lives up to that challenge. Our focus has shifted from the needs of our studio to the needs of a much larger audience, and it's become obvious that this need for a better production renderer extends beyond the max community.

    3DF|CGN: Was there much pressure to release Brazil sooner because these renderers were hitting the market?

    Kirvan: We don't deny that certain external pressures existed, but at SplutterFish we are driven by our own internal standards, and don't have to answer to the demands of shareholders or marketing departments. People asked us to release the software so they could start using it, which is quite understandable, but we weren't willing to release a premature product. Around the time of SIGGRAPH last year we were feeling some pressure to release early, but we were not comfortable with arbitrarily slapping a "version 1.0" on the software and trying to collect as much money as we could before anyone figured it out. Instead, back in November 2001 we started the pre-release beta sales program so that those who wanted to buy early, honestly knew what they were getting. We made it clear that we would not release Brazil r/s until it was complete, stable and above all, production ready. And that's what we have done.

    3DF|CGN: Are there plans to port Brazil to other software/platforms such as Maya? What about a Linux render farm version?

    Kirvan: Sure. There's obviously a need for Brazil r/s on other platforms. We get e-mails constantly asking about all of them. We have planned to do cross platform / cross application support since the beginning, so that the animator can work wherever they're comfortable and ultimately get the same quality rendering. We don't have a timeframe for any of these yet. As information becomes available we'll be sure to let everyone know.


    3DF|CGN: Brazil is obviously targeted at the high end production level. From the feedback from customers, how does it compare to PRMan and Mental Ray? Should it be compared?

    Kirvan: Should it be compared? Probably not at this point, but inevitably it is. Those renderers are much more mature than Brazil and are, in many ways, much more technically evolved. They have benefited from many years of advanced development and some of the best minds in the field. We get a lot of positive feedback from our customers comparing Brazil to these renderers, but in a more established facility with a fully integrated PRMan or MR pipeline, Brazil would probably seem more limited since presently it only runs in 3ds max. In terms of 3ds max and VIZ rendering, however, Brazil's integration is much cleaner and the third party compatibility simply cannot be matched by a renderer that has to export to an intermediate format. SplutterFish is a young company and Brazil r/s is a young product. It would be an honor to be held on par with those renderers and with some luck, some day it will be an appropriate comparison.

    3DF|CGN: The announcement highlighted that Splutterfish would be porting Joe Alter's Shave and a Haircut to 3ds max. Will this work with the default scanline renderer? Does Brazil work with Shag:Hair?

    Kirvan: Yes, Shave will work with scanline. SplutterFish is spearheading development for the port of Shave and a Haircut to 3ds max. The project is not directly related to Brazil r/s at all. Obviously, we'll have the opportunity to maximize the features and capabilities of Shave when the renderer is Brazil, but we won't be limiting any functionality and such, as a penalty for not using Brazil.

    Brazil does indeed work with Shag hair as well as hundreds of other 3ds max plug-ins. We have gone to extreme lengths to make Brazil work seamlessly with all the third party plug-ins. There are, of course caveats because in some rare cases third party plug-ins do something that assumes or requires that you use the default scanline renderer in max, but so far those cases are few and far between and can usually be worked around in one manner or another. I'm confident that Brazil r/s has better third party compatibility than any other max renderer available.

    3DF|CGN: Now that version 1.0 is released, will the public tests continue?

    Kirvan: We plan to always offer some sort of free version of our renderer. Our clients will now participate in beta testing some of our newer features, but the public testing is always good to make sure our core systems are still as bullet-proof as possible. [3DF|CGN]

    Talk about this on CG Talk >>

    Related Links

    << Previous page (1 of 2) 

blog comments powered by Disqus