Page 1 of 2
Bounce Torture Test
Posted: Mon Aug 01, 2005 11:07 am
by tom
Pure maxwell, no post process...
MAXWELL 1_2_2a_beta ( Engine core: rs1c )
[ Physically and Unbiased Based Rendering ]
## Scene Information ##
11 meshes ( 29220 triangles )
## Render Information ##
*settings:
resolution: 800 x 600
*render time: 42 hours 54 minutes 16 seconds
target sampling level: 50
sampling level reached: 19
MaxwellBench 1.0: 9.138018
Posted: Mon Aug 01, 2005 11:14 am
by Duncan
looks good Tom, but damn thats a long time for such a low res.
What are the specs on your machine ?
Posted: Mon Aug 01, 2005 12:12 pm
by tom
Thx Duncan,
Sampling Level 19.67 is not so easy for such a scene.
Remember, it has unbiased bounce for not an easy condition
You can find the config details in texts above and on image...
Posted: Mon Aug 01, 2005 12:27 pm
by jomaga
Very nice, Tom! and very clean
btw: Why only "MaxwellBench 1.0: 9.138018". Your computer should go higher. I´ve seen results about 15-20 with computer like yours.
Posted: Mon Aug 01, 2005 12:28 pm
by Nicolas Rivera
Hi Tom, very nice image, a really tortuosa image.
But Tom, if you are now part of the NL team, you most be in the payroll, so please by a new machine, or ask Victor to send you at least the mother board with 2 Xeon procs.
I'm so desperate to change my G4 800mhz, that i'm planing on saleing one of my baby's a custom 62 stratocaster.
Posted: Mon Aug 01, 2005 12:41 pm
by tom
Well, it's not a poor machine Nicolas.
Gigabyte 915P Duo Pro 800/1066 FSB LGA 775
Intel Pentium 4 3.2 Ghz HT LGA 775
1 GB RAM, DDR2-533Mhz
I know the problem is Maxwellbench shows 9.138018 and seems very low.
But I'm not sure about this value right now. Seems like it's getting down by
approximation in time. I'll try to explain this later...
Posted: Mon Aug 01, 2005 12:46 pm
by Nicolas Rivera
Ok, its just that i see that using Xeon procs. is the best way to go with maxwell??? i'm i wrong, i'm realy considering buying a PC this time, but i have no idea if its better to make my own machine or buy it from a big ass company, what will you recomend??
What does this mean???.....Gigabyte 915P Duo Pro 800/1066 FSB LGA 775?
Thank you Tom.
Posted: Mon Aug 01, 2005 12:48 pm
by tom
Sure it's not as powerful as Xeon
but I said "c'mon it's not that poor"
Hehe... anyway Xeon's are all right, keep em
Posted: Mon Aug 01, 2005 12:50 pm
by def4d
Hmm sorry , i missed something : what is the MaxwellBench, and where do you find this?
Posted: Mon Aug 01, 2005 2:07 pm
by mgroeteke
tom, very interesting and pleasing for a test image!! please can you share the material settings? this render seems to be apparently totally unaffected by the 'blur behind glass' problem, also the reflections seem to be much more intense than in the glass i can produce on my setup. but maybe it's the better max integration, i suppose...
Posted: Mon Aug 01, 2005 2:09 pm
by abgrafx3d
Nice test Tom!
Posted: Mon Aug 01, 2005 2:20 pm
by mane162
Beautiful Tom
mane162
Posted: Mon Aug 01, 2005 2:56 pm
by DELETED
DELETED
Posted: Mon Aug 01, 2005 3:43 pm
by Hervé
Hey Tom, please... be decent... stop posting photos... !
Posted: Mon Aug 01, 2005 3:47 pm
by mgroeteke
whiskey wrote:
also the reflections seem to be much more intense than in the glass i can produce on my setup
this depends on the angle of the light hitting the glasssurface resulting in total reflection...when we render watches, at steep angles the glass totally reflects the light..at shallower angles, you can partly see through..and by looking perpendicular to it you see almost no reflection (of course this depends on where your emitters are)
EDIT:
actually the other way round..sorry
yes, this actually is the fresnel effect, but on mac and LW i cannot really control overall reflectivity...