[big K]

yes you are right. but i need to render it again anyway for the blurry reflections.

[The Pixel Artist]

Hi all, nice images. Those Maxwell ones are beautiful, love those blurred reflections and caustics.

BigK I thing photorealism-wise your Maxwell one takes the cake so far (kind of neck and neck with Thomas's then hdesbois's.)

Thomas your Vray one is absolutely gorgeous, Id say the prettiest of the group. But then again 14+ hours... ouch. (Well I guess that’s better than 19+ and 21+ hours).

I gave it a quick go last night, but I was going for something hopefully a little faster. Technically I probably don't qualify for this challenge because I didn’t render it with DOF, did it in post this morning instead (so add 20mins of total post time on that). Setup was about an hours worth of messing before bed and hitting render. It ended up rendering for a bit over 3 hours, but I actually rendered it out at 1600x1200 so I could do better post work on it and then resample it back down to 800x600 (to retain more fine details). Ended up using IR Map and lightcache. I suppose though that if I have rendered it out at 800x600 it would have been a bit under an hour. I made a mistake on the caustics though, forgot to change a setting, so they could have been better (chalk one up for a more simple setup program)

I did start a test with DOF and a more brute force setup and extrapolated it would have taken around 4-6 hours, but it probably would have been prettier. I also think the new adaptive PPT would probably do an incredible job in a 5-10 hour range, but then I don't have access to that yet (please render Gods give us 1.5 soon).

Anyhow fun challenge, looking forward to more.

P.S. Oh specs... Max/Vray P4 laptop @ 3.4ghz w/HT and 1gig ram.

[dutch_designer]

Working on a Brazil render now, 30million caustics photons... eeks..
I'm a newbie when it comes to photon mapping, so far I've never really needed it and just did everything in the easy peasy QMC.
So whatever the result will be, I'm sure people more familiar with Brazil can easily outgun my image

edit: ok well my entry can be ingored when I finish it, I'm not even using image sampling and its taking me too much time to figure out all these settings... i'll take maxwell over anything else any day when it comes to these kinds of images.

[dutch_designer]

Ok, try not to shit your pants, I only learned the basics of photon mapping this morning and I'm confused, there's sliders for bounces and whatnot everywhere and I'm thinking there's a whole lot of redundancy going on in the configuration.. so I'm not gonna let this stuff mess with my head any longer and give it a rest..
If anyone wants the .max file (requires Brazil), let me know...

[ivox3]

hey dutch .........too late !

j/k ........

[dutch_designer]

hey dutch .........too late !

j/k ........

Too late? aaaarghhh

[arch4d]

ok, here is my try with maxwell:



rendered 1 hour on the below mentioned machine...

[Thomas An.]

Hi Pixel,

Thanks for your test, man

Let me tell you ... I have put a little over 150 hours studing the behavior of Vray IR map and LC based on the this scene. So I have been in your shoes on this one ... it looks so close ... but ... to get from this point of dotted caustics to a completely smooth state will bring up the time substantially.

So far, I think the best method of doing this scene with Vray is to use photon mapped caustics and bump up the light photon subdivs to 15000 or 20000. The draw back is that with photon mapped caustics it is no longer possible to do region render tests (you have to render the whole scene every time just to check if its ok and have to wait for the darn photon map to compute; and that can be a couple of hours on it own) ...

[Thomas An.]

so here are my tries on my PIV 3.4 GHz

cinema 4d 2h35min (have to change the blurry reflections though and something is a little off with the caustics)

Yup, caustics need some work

maxwell SL 18 in 19h45min (setup in really no matter of time ) DOF is a little strong (should have used fstop 5.6 instead of 4)

This one is looking good !

i tried in fR2 for cinema, but something didn´t work out. so if holger reads this. please tell me how you got caustics for the area lights.

Thanks for giving it try

[Thomas An.]

Hey Dutch

Yeah a Brazil version would be great (I rarely see Brazil in action)

But, man whats up with the antialiasing ?

[Thomas An.]

arch4d,

Which file did you download for this try ?
I see those rings are not smooth (there is distortion like in Mihai's) ... so I better check that file to see what is up with the mesh.

-

[mverta]

If I never have to wait around for another photon map - whose values I just guesstimated on - to calculate, it'll be too soon. Come to think of it, if I never have to do anything beyond press "render" and get photoreal GI again, it'll be too soon. If Maxwell does nothing else, please God let it set the bar for how easy setup should be!

Looking good all around, guys... great idea, T!

_Mike

[dutch_designer]

Hey Dutch

Yeah a Brazil version would be great (I rarely see Brazil in action)

But, man whats up with the antialiasing ?

Hey man, yeah I disabled the antialiasing to speed up the render preview.. since I only started bothering with photon mapping since this morning I'm so lost in all the settings and what they do and what they mean.. I must've been messing with this for a good few hours before I gave up.
It sure puts me back on my feet and shows how awesomely simple Maxwell is and the output quality it offers..
My pc isn't very slow and I can't imagine it takes that many photons to get a good result, but I have no idea where to start to get better results.

[Thomas An.]

Hey man, yeah I disabled the antialiasing to speed up the render preview.. since I only started bothering with photon mapping since this morning I'm so lost in all the settings and what they do and what they mean.. I must've been messing with this for a good few hours before I gave up.
It sure puts me back on my feet and shows how awesomely simple Maxwell is and the output quality it offers..
My pc isn't very slow and I can't imagine it takes that many photons to get a good result, but I have no idea where to start to get better results.

Cool
Looking forward to your results

[Thomas An.]

Hi everyone,

I would like to point out that the images should be in "Final Quality" and artifact free. In other words ... something worthy for a client to put in product catalogs.