bedroom

[metroberlin]

maybe silver can post the file, at least it´s his scene...i can´t do that. if he wants to share it, fine....later the day when i have some time i can post the settings

[hughg]

Sweet no matter what....

[silver]

hey marc, thanks again and again and again
if you want, you can post the whole file, because the same way i want to learn more, i want mates improve your knowledges, that's no problem for me, but i guess should be better you post the file because i havent a server for that, i can post only images

will try to make some corrections at the file and post again here

bye
silver

[didjaja]

Silver change *.max to *.jpg and it work fine

[silver]

hey didjaja,
problem is not the kind of file, its about your size... they accept as maximum value 800 k, and the file has 1.32k

[didjaja]

Well,compress to *.rar,the scene with dummies is little and it has to size less than 800kb.Or if it is more compress it two parts and if you need rename to *.jpg if your server only acept this archives.

[Psy]

if you want, u can use the site in my siggie to upload the file.

[silver]

hello psy,
i send you tthe file, ok. Please, post it for we learn a bit more...

Well, i follow the metroberlin settings but i really guess i have a problem about materials... as you can see... differences was not too much, but i realize something interesting points and would like to discuss about them:
time of pre-render ( to calculate irradiance) was too short : about 3 min at 400 x 300, but render final times was too long : 4: 42 hs at 800 x 600 againts only 3:00 hs( time spent byt first render after calculates) may we didnt take this time in consideration because i was rendering and doing another job at autocad... but anyway, times are too far one from another one.

folow some points to discuss about:

1. after open the file i could see that marc put big values at subdivision of vraylight: 120. This is really necessary man? because i've never had used a so high value.

2. i've just finished render the file you sent me , it takes 40:00 min. In my opinion it takes too much time if we compare with the first that i did ( 3:00 ) but the resolution was only 400 x 300 , and yours was 700 x ? ( dont remember) . however, it was amazing, exactlly what i am looking for. I didnt render it with high resolution, so i dont know if after calculate, it can be faster than 3:00 wich was time that first file took ( 3:00 at 2000 x 1500 )

times are pretty weirds, will put here only to show differences:

first render at the top of page: 3:00 h witn 2000 x 1500 after irradiance calculates and without marc settings

secound render : 4:34 h at 800 x 600 ( after calculates )

third render ( metroberlin file and configurations ) 40:00 min without calculate


Would like now to get some explanations about photons from this file. There is something that i didnt figure out: it is about distance; how can we handle with this value? if we have a bigger ambient, we put higher values to distance and density? or no? i guess this is the mainly point about photons... When i was trying to reproduce gonçalo test, with photons at first and secound bounce, i need to put 2000 at distance and 500 at density, but looking for metroberlin configurations i can realize lower values ( 1.4 at distance and 0,5 at density) so.... what is important considering when we change this values? because i really feel like a bird... lol flying all the time

sorry guys to write so much, but i need to utilize this oportunity to discuss and try to learn a bit more, cause photons are driving me crazy, hehhee
thanks in advance for metroberlin patience and for everybody could share a bit of knowledge


this is secound render, still with some dark areas... but they must be come from materials

[RErender]

If you are going to use photons it is absolutely necessary to use vraymtl. If you convert all of your materials I bet it renders in atleast half the time and look much better.

[Psy]

how/where exactly did you send it to me at?

I meant in my post to just go here:
http://vr.reaperfields.com/index.php
Click upload at the bottom, browse to where you have it saved and upload it.

But if you want, send it to webmaster@3dmaxedout.com, I switched hosts recently and hadn't yet setup my webmaster email acct. It is there now tho, so if you want, send away.

[metroberlin]

mainly your renrtimes come from the glossy reflactions and refractions. the max density controls the distance between the rays on a surface. like a grid in radiosity. higher values makes it faster, but also looses details. the search radius blurs the rays going away from the surface. mean, twice the max density makes it sharper but maybe causes some artefakes. higher values like 5 time the density makes it more smooth. you can play with the search radius even with a stored map. try as a max density the value 1-2 and search radius 5-10. this makes the stuff faster. okay 120 samples are a bit hight. that was from my last test. i normaly use these samples when i store the shadows in the irradiance map. surely you can use lesser samples. i´m finding you rendertimes quit high. as i said, on a dual mp 2400 my pic took 12 min. did you use for your high res. pic lower irradiance map settings ? for 2000x??? you dont need -3/0 ...you can use -4/-1 or even -5/-2.

mostly you have to setup ALL YOR MATERIALS in a complete new way. they cause most of you probs.

[metroberlin]

i will post a testfile later when i´m back !!!

[silver]

hello psy, i sent you just by your e-mail webmaster@3dmaxedout.com... take a look and tell me if it already arrives !

[silver]

hello marc!
thanks again by your support, about render times, when i said that i did render at 2000 x 1500, this is the resolution after calculate irradiance... indeed, i did the calculates at 400 x 300 using -3 0 .
mY PC IS AN ATHLON XP 2100, OVERCLOACKED to 2600+ , videoboard gforge ti4200 128ddr - 1.5 gb ddr 400 pc3200 and a hd sata 150 seagate 80 gb. If you said that your render tooks 12 min, i thougt difference from mine its too high ( 40 min rendering your file)

Anyway, if this was the real difference betwen a single and a dual cpu, i need to buy a dual as soon as i can would like to know if anyonelse did the test and get better results

thanks, silver

[Psy]

argh, its 8am here now and still no email.

try monster@reaperfields.com if u can

one of my many domains has got to be accepting emails....