www.kraytracing.com

Janusz Biela wrote:

vgabex wrote:

Janusz Biela wrote: 4700 here, so your times render is correct.

Janusz, is your config a dual E5-2697 v4?
I'm planning to step up from my 1700 cinebench point (i7-5960X @ 4.4) to a higher level.
Can you please help with your exact pcu, mobo and ram specs?

Thank you!

This what we have in office:
CPU.png

Maybe you need for the tests the use of less expensive computers, comparable with those that have the majority of users?

Yes, but is one problem: I use Process Lasso software to control CPU affinity (https://bitsum.com/)
The problem is that, with new Xeon v4 this fantastic software has bug: Affinity is not working at all! I posted this bug on theirs forum and wait for new update of Process Lasso.
Without this feature I can not "simulate" i7 CPU.......so to receive time render on your CPU just test your CPU by Cinebench R15 divide my score (4750) to get ratio number how slower you are. I think this is the best idea because people has many different CPU.
Here is solution:
1. test your CPU by Cinebench
2. remember you score (for example 900)
3. divide: 4750:900=5.2 (around)
4. multiply my time render by 5.2

This is for K3.0? if not, how turn on QMC?

zonah wrote:This is for K3.0? if not, how turn on QMC?

This is everywhere include Native Lightwave: In both Kray: and control of noise: More rays = less noise. Unfortunately in K2 (and still K3) noise levele is static - user must prescribe static level. In future when we add PPM, VCM to the K3VPR it will be progressive (like Octane or Corona or even old FPrime, the more you wait the better image).
QMC system today is minimum for rendering. In my opinion this is future for rendering as combination of PPM and VCM.
Do not try t in Lightwave - the time render will be horrible ....

Ah ok, thanks
Making Vpr progresive with QMC like fprime , i love it!