HYPER GI

[Gunnah]

O_o and what does that have to do with splutt irc? This sounds like a problem with fR webboards, not splutt And about the overproud: um, we see lots of nice things in other renderers.. you really must be confusing us with cebas Trust me, I have a lot of respect for Vray, I just prefer to use Brazil.

have a good one,

G

thats simply not true!
maybe you don´t know what edwin is doing? thats the case i think.

the "problem" is that edwin has enough time to delete all postings, even while stage-1 is so late.

i think we have a few other reader of the fR forum here who will say that im right. you get banned for simply comparing vray with fR, pointing user in the direction of vray, or making comments about the total lack of informations.

only 2-3 days ago edwin deletes two postings from users who told someone to use vray when he needs a fast renderer and as longs as stage-1 is not released. i don´t saw any new postings from them and i bet the account was deleted (Frank Hams or so and another guy).

that a few splutterfishes are a bit "overproud" and can not see anything good in other products, does not makes cebas behavior any better.

[Long John]

sorry, i thought that you are a cebas guy and that you quote my posting because you wanna show, that cebas is not the only company with a problem accepting critics.

indeed, i missed that you are a splutterfish guy.

was early in the morning for me.

[Dman3d]

Not to get things going again, but

In the German engineering vein: Farfrumrendering
Sorry, had to do it

[Long John]

Dman3d thats all, but not german LOL

[Da_elf]

Two methods are combined in a unique way (classic, view-independent, finite element radiosity calculations and view-dependent stochastic shading point sampling) to provide stability and consistency in the illumination of the resulting animation........

ok. lets look for some definitions here...

sto·chas·tic ( P ) Pronunciation Key (st-kstk)
adj.
Of, relating to, or characterized by conjecture; conjectural.
Statistics.
Involving or containing a random variable or variables: stochastic calculus.
Involving chance or probability: a stochastic stimulation.

so basically their random variable involving chance or probability will give stable results. Hmmm. chances are your going to get screwed.

Just check out the sample animation rendered with Hyper-GI below. Every object in this animation is set to receive and reflect light, so every single brick will influence the GI solution......

Correct me if im wrong but this sounds like normal GI

[Long John]

do not wonder.

this whole feature preview is a joke.
i think we can only say something about stage-1 after siggraph 2003.

all information we can get now are only marketing blah blah.

[flipside]

Just check out the sample animation rendered with Hyper-GI below. Every object in this animation is set to receive and reflect light, so every single brick will influence the GI solution......

Correct me if im wrong but this sounds like normal GI

They're just trying to say that this kind of anim is difficult for a GI renderer because every brick is recieving/reflecting light. This phrase is not an explanation of their way to render GI anims.

[Long John]

it´s more or less a useless information.

ok, they could have set every n-th brick to create and receive GI but thats nothing that help understand what HYPE(r) GI is.

the other example without GI is such a stupid example, i don´t get it...
they should have shown a stage-0 animation, instead they show a animation without any GI.

the MTD preview is something you can read one old siggraph papers too.
many people asked if MTD is still slow, so a better scene with rendertimes would not be to much after 1 year.

[Da_elf]

hell, what they should have shown were render times. any renderer can do that if given the time to do it.

[Long John]

i agree.

but before cebas will show rendertimes hell will freeze.

the only thing i heard about rendertimes from a betatester was that stage-1 is more then twice as fast as stage-0.
that could mean anything.

im curious how fast MTD is (which is still in beta) the siggraph 2002 videos show a very slow MTD (keep in mind it was a multi CPU system and a untextured object). and that was two month before the first release date.