They have non-nVidia rendering, but only for their biased engine, no raytracing kernel. People need to accept that this is a limitation of AMD hardware/drivers at the moment. Production features just aren’t there, as evidenced by the lack of higher end features on AMD by nearly every GPU renderer available.
Wow - big discussion here! Thanks. Many point of view to think.
Btw, we offering also student license for 99,- , non-commercial price for 299,- and rental from 3-12 month.
FurryBall also support rendering over network and standalone - (you can export whole animation with EVERYTHING and render it on separate computer without Max or Maya. It’s great for studio use and HUGE scenes.)
Does this handle free renderers?
FurryBall: “Are Blender users open to pay for plugging??”
If they offer a good Renderer we are open to pay for plugging, training, software and more, the question is what advantage Furryball have to offer to us?
a lot of people doesn’t know about FurryBall, so this thread is perfect to know more about this Renderer and how it works And the possibilities of integrate with blender.
somebody that have maya or max can make a comparison vs cycles about speed???
Yes, we will be happy for speed compare.
Here you can download test scene and try (sorry Maya and Max only, but somebody can convert). According our test is FurryBall the fastest.
http://furryball.aaa-studio.cz/aboutFurryBall/compare.html
Speaking of uninformed, you are wrong about it being available for commercial usage and neglected to mention that it leaves a watermark on renders in the free version. Given that you already know that it’s also limited to three hours a day as well, it should have dawned on you that I wasn’t speaking about anything like the limited usage trial version that they offer. And you could have come to the conclusion that I did in fact look at the website based on the other points I made, but apparently you’d rather jump on your high-horse and ride off into the sunset.
Anyway, good to know that they do have a non-commercial version. I must have glanced over it what with it being buried under all of the rental versions.
Thanks for the update and the scene. I’ll be doing my own comparisons and I’ll upload the results (and the converted .blend) here when I’m done. I’ll do FurryBall, Cycles, BI, Thea, Arnold, Lux, and Yafaray. If anyone wants any others tested just let me know here.
I’m getting a file open failed error with the .max file, haven’t tried the maya file yet. Any ideas?
Also, running the benchmark crashes my display driver on “render shading”, so I can’t even get results from that.
EVGA 560Ti, driver 314.07
If you have Nvidia 560Ti (suppose only 1 GB memory) it’s so few GPU RAM for serious raytracing. This is on edge to crash ;-( Minimum is 2 GB for work.
Also your rendertime, even if it work will be so slow: (75 sec, compare 23 sec on Titan)
http://furryball.aaa-studio.cz/aboutFurryBall/benchmarks.html
Max scene is from 3DS Max 2014.
Hi FurryBall
a friend exported your test scene to Thea but he told me that the displacement map is missing, could you please post the displacement map ?
something about the comparison renders on your web page:
I think it would be good if you post the original files you used to make all the comparison renders from the other render engines. first of all it is important that everybody can test those files and see that everything is “correct”. in case you made a mistake with one or more render engines, those that are more familiar with those render engines can correct the mistake and so the comparison will reflect reality a bit better.
the Maxwell example is particularly bad … I don’t think its to difficult to get the sun at a similar angle then the furryBall scene … with Octane you even let the sun out and only sky light is illuminating the scene (no wonder it seem to be the fastest as no sun light bounce was calculated).
now about the feature set you post … why is sss and instancing not on the list ? what about relight the scene after finishing the render (Maxwell and Thea feature).
and now something I find very “curios” … on the web page you are talking about “unbiased rendering” … but caustics are not supported.
and then you setup the test scene with very low tracing depth etc.
one thing is rendering in progressive mode and another is being unbiased … those things can confuse people very easy.
I would be very carful saying that FurryBall is the fastest render on the market … you could have a “unexpected” surprise 
Greetings Patrick
1.) Yes there are.
2.) In fact yes I have so it appears you’re the one expressing uninformed opinions.
3.) Derogatory? Really? Do I need to call Whine Wahn Wahn and have them send a WAAHHHmbulence? 
I’m kidding, I’m kidding. I’m not trying to be insulting. I’m just saying I think it’s a might on the overpricey side.
4.) I would buy FB if it was a one time cost, even if that one time price was $1000 or if they had a rent program that was more affordable.
5.) That depends. If they lowered their prices, they might make up for what they would have lost in volume of purchasers. I work with a large number of people and we’ve had this conversation many times before, and not just about FB. We’re small studios and individuals. We can’t always easily afford to shell out 1300$ year after year when we’re shelling out as much on other things that are more important.
I don’t know why in the world you’re getting all hissy at me. I don’t think anything I said comes even close to being insulting… maybe a little sarcastic, sure but derogatory? Pft. Maybe I spent too much time in NY. Shrug
Anyway, no one likes a Negative Nancy so I won’t keep harping on the point. I think it’s too expensive but if their business model works and they are successful then, well, good on them.
Patricks:
Please use any procedural displacement texture - we used checker.
Also 100% similar light direction and same hair shape is not so much important - this benchmark is about speed.
Please be free to set your favorite renderer better, if you are able. We will be happy. We are not 100% hardcore users of all those renderers and as it’s written, we used almost default settings in all renderers. Please send us after your results.
Also many others features than SSS, instancing and relighting are not at the list, but FurryBall support all of them btw.
Hi, I am a Octane user and use the new addon also.
This addon is based on the Cycles addon but change Blender source code too.
They use also a “Render Server” which communicate with Blender over TCPIP (local network), this is closed source.
This is a huge project and Otoy has a full time developer for it, he work for months now on it.
There is also a exporter based on the python obj exporter in Blender, a Octane user developed this in a few days.
May somebody could post a .blend from the FurryBall test scene than i can try in Octane.
Cheers, mib.
Very nice comparison, page like that makes me want to buy it. I too look forward to m9105820’s tests…
Just want to throw my 2 cents in that there are open source purists, but blender is such a good app that some of us use it out of preference. We are the ones that are not opposed to buying apps that help us get the job done if they’re worth the money. I was under the impression that octane and cycles were the main GPU contenders… Thank you for posting the thread, I was unaware of your renderer and am much more aware of it’s capabilities… thanks.
Might be waiting a bit for tests… I’m unable to get the benchmark to run on my GPU, and even relatively small scenes that render with lots of memory to spare in Cycles/Octane/iRay/VRay are causing some issues with the FurryBall demo version.
Please believe our tests - we already tested Octane, Vray and FurryBall just test Blender Cycles for CPU and GPU for this scene.
Any rendertimes times in Cycles or Lux?
Or did anybody convert the scene to Blender?