FPS vs RenderTime

Started by PhilippeV8, March 16, 2011, 06:12:30 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

PhilippeV8

Is there a clear relation between the FPS we get in the benchmark and rendertime ?

If I get 5 FPS and render a scene in 30 minutes, does this mean that on a machine getting 10 FPS that same scene would render in 15 minutes ?

guest84672


PhilippeV8

Regardless of using realtime render or the other ?

jeffw

I don't understand FPS. I thought frames per second referred to motion graphics but
Keyshot uses the term for single image rendering. If you are just making one image where
are the frames? When I render I set the image size and the DPI and it takes
X amount of time render. I don't see how FPS relates to the rendering process.

guest84672

Higher FPS means faster rendering. It certainly is no indication of how long your rendering will take.

jeffw

So if FPS only affects rendering speed it must be controlled by hardware and not something one can vary in Keyshot. If you want higher frames per second you need modify your computer.

guest84672

Materials, ray bounces, ground shadows, detailed shadows, ground reflections, shadow quality, detailed indirect illumination, indirect ground illumination affect FPS, and therefore render times.

PhilippeV8

I understand the confusion for the term FPS .. since FPS is most used for frames .. which indicates it's a movie with a # of pictures rendered per second.  In case of a single image render I would rather expect people to talk about pixels per second; no ?

hop

Right - FPS is completely pointless, this isn't a video game. All I am interested is what renders fastest because it affects my workflow - I don't need to spin it around fast, I need to have it photorealistically render fast. Benchmarks should all be a render time for a scene at some standard resolution and settings.

Would be very cool if keyshot embedded the settings and render time as EXIF data in the final jpegs

Robert V.

#9
I disagree that FPS is pointless, your PC still needs to calculate during the "realview/time render"
Comparing the FPS from 1 scene (the camera) with all users, does tell us which system is the fastest.

Looking at rendertime (everybody is going to test render 1 scene) you will get different numbers, but not a different outcome.

However, the benchmark is useless. It's quite obvious that the fastest processor will give you the best result.
It's just fun to see where your PC/notebook stands.

(ok, if you start comparing the same processor with different mainboards/memory etc. could give you info on which components work best together. But this would take a lot of keyshot users!)

guest84672

@hop - once again, if your FPS number is low, your render times will be longer.

But I like your idea of adding a render time.