render benchmark

geli.holte_53070dbc82

Hi,

once in the old forum, there was a weblink to a scene for a render benchmark for different 3D Programs, like Carrara, C4D and Max and so on.

Now someone a site where I can found a scene for a render benchmark?

Thanks
Geli

wetcircuit

Boy oh boy..., That sounds like a fun way to spend the afternoon pushing the render button.... I hope you win!! ;)

geli.holte_53070dbc82

Sometimes, if the animation tools in Carrara do not work like I want, I consider to change the 3D Program.

I know the rendertimes in Cinema 4D and they are not acceptable, but maybe another 3D program works better for animation and render fast as in Carrara.

Geli

wetcircuit

Is just a small joke...., we would all love the fastest renderer possible, but there are SO many variables that could help render *your* scenes faster.... Shadow buffers, texture baking, rendering foreground and background separately and compositing in post.... And the REAL time spent is the time you are infront of the computer. Once you hit render and step away, the numbers are less important. Overnight is still overnight... The interface is honestly more important, how productive you are when you are using the program...

Rendering a test scene on different computers in different programs is imho completely academic - it only proves THAT particular scene , which is usually a lot of bounce lighting calculations on architecture, the very opposite of your clever animation in the contest....

Art is not the simple crunching of numbers on computer chips, art is usually the elegant process of overcoming obstacles...

Also the cost of a second renderfarm computer is probably LESS than the cost difference between Carrara and the next capable competitor.... A fast PC with a low-end videocard (best for renderfarming) is a few hundred dollars...

But it is interesting to hear that C4D is still slower..., that was my also observation years ago too.

evilproducer

The scene you are referring to goes back to C6 if I recall and it's not optimized for anything after that. It also used a lot the renderer's bells and whistles, such as full GI, which is not what you'd want to use in an animation. As Holly mentioned there are other factors which a test scene doesn't take into account and that's hardware configuration for one. There can be significant differences in machines with identical processor. For instance, if machine A has a new hard drive and machine B has an older, fuller hard drive it could slow down swapping data to the scratch disk. RAM and the type of RAM are also an issue. If all things hardware are equal, then there's the system's OS version, what programs are running in the backgound, etc.

In my opinion, it's a glorified pissing contest. To see how other software compares, try their demos or trials (if available.) That's the only true way to see if something works better for you.

WendyLuvsCatz

I would love to see a software pissing contest %-P
Blender fluids would prob be neck to neck with Houdini but our metaball particles do not make a bad looking yellow stream, it is just the inability to fill the cup except by faking it that is lacking, we can be just as splashy!
to enable Youtube posting, it would have to be a cocked leg dog or spraying cat as humanoid urination is a Youtube no no!

wetcircuit

The things I learn at the forums... LOL!

JoeMamma2000

IMO, the real benefit of render benchmarks is that they can help you decide whether that brand new super processor that just came out is worth the investment.

Often you get guys jumping up and down about how incredible the new 48 core ultra processor is, and later find out that it will only cut your render time in half, because for some reason the software doesn't take advantage of whatever magic the processor is doing, or whatever reason stuff has for not working like you'd expect.

Which is fine if you're doing 6 hour renders, but if most of your renders take 4 minutes, is it really worth paying $1,000 for a new machine? In that case, benchmarks can be helpful.