FurryBall 4 - new speed compare (unbiased rendering) V-RAY, Iray, Mental ray, Arnold

#81

again… nice render times… but could you post also your CPU configuration? I think its also very important to compare with other render times. thanks!

#82

i7 3930K stock speed

#83

I have an i7 3930K running at stock speeds as well. However I don’t think it has a big impact on the render times when using Redshift. CPU does play a role when rendering with Redshift, but it’s a very very small role in most cases. You will probably get similar render times even with much cheaper CPU:s.

I did a test with this benchmark scene. The idle CPU usage was between 0-5% and during rendering it was between 8-20%. Average was probably around 12-13%. I could do a more precise test and close chrome and other apps, but I think this gives a pretty good idea.

#84

FurryBall 4.6 test - illuminated only by HDRI, brute force - 10min render time on Titan. (Max ray intensity parameter used for reduce noise).
2000 rays.


https://scontent-b-fra.xx.fbcdn.net/hphotos-frc3/t1/1796455_10152182201242570_307554373_n.png

#85

Can you post the scene in 3dsmax for comparison?

#86

or share scene for reproduction in other engines as well?

#87

It’s just few balls, toroid and plane and HDRI map - 10 min works. :wink:
It will be faster to make it yourselves ;-))

#88

thing is to get identical scene and look for better comparison.
everything can affect, poly number, texture size/resolution etc

#89

Mental ray reporting, 3min 45sec

#90

Thanks for posting, but it looks like just one light bounce (not 4)…
Could you also send your CPU?
Thanks

#91

I rendered it with photons and FG, interiors should never be done with FG only. So 10 bounces, not 4.
i7 3770K@4.5Ghz

Btw. iRay wants to play also, pretty much the same result in the same time (16 minutes, as your time with iRay), but on a GTX 650, not a TITAN.
[img]http://i.imgur.com/MJBTik0.jpg[/img]
#92

Yes, this Iray render is much more “correct” to me - your previous Mental ray render image looks little bit strange to me (no shadow under balls - specially fur, no reflection inside the red ball, almost no shadows on the walls…).
This is why I was asked about number of rays.

Your Iray render looks little bit more noisy than ours Iray render, but hard to say. There is huge differences for final tuning.

http://furryball.aaa-studio.eu/aboutFurryBall/compare.html
If I compare for example our two FurryBall images on top (3 min vs 9 min) the result is NOT 3x better, even NOT 2x better, but time is 3x worse… :cry:

#93

Might be time to update the chart in the first post, a lot of it is out of date.

Also seems a little silly to compare the latest GPU to a 5 year old CPU. :rolleyes:

#95

Thank you for sharing this amazing comparison.

https://www.laptopinspect.com/
https://www.laptopicker.com/