[Maximus3D]

Tom: I don't know..

jomaga: That's clean, did you render it with Vray ?

Mihai: Oops! you're right it appears to be missing the beautiful watermark i'm still working on a way to permanently remove that watermark.. the question is how...

ivox: Noooes!!

rusteberg: Wow man those shots look great! now add some noise to them so they look more realistic

koufax: impressive images, that multilight color thingy looks real useful!

Here's one i had cooking during the night, it ran for something like 9,5 hours or so.. i can't remember what SL it hit but it doesn't matter. Too bad that bloody logo is in the way and covers just the part of the image i wanted to see one hdr is used for illumination, that's all. Scene is the same from Modo as in my previous tests.


/ Max

[sandykoufax]

thanks Chris, btw where is your Bruce Lee?

Max, thanks. I like your orange juice scene very well.

rusteberg, the hotel scene looks very nice.

renato, may I ask you how did you setup your scene? looks very interesting.

and here is multi-layered emitter material test(though ugly scene setup). I really like this new layer system.

[Maximus3D]

Hah, funky looking materials Koufax how did you do that ? was it difficult to do ?

I just had to test my photostudio props to see how well they worked with 2.0 and although this is a very early and rough test i have to admit that i'm impressed how well and also how fast it works! considering that i'm using thinSSS for the cloth with a 0,5mm thickness (awesome parameter btw) and a mirror like lampcone for the lamps, something which would have been impossible to do with 1.7.1 since i would lit the entire scene with caustics. Now it renders just fine oh my.. the possibilities this opens now.



These results are still pretty noisy but with more work spent on refining it it will work much better and render cleaner. Still as a early preview i'm bloody impressed. And that does not happen often.

/ Max

[tom]

Max, what about your glass+liquid geometry setup?

Tom: I don't know..

Has everything happened so sudden?

Rusteberg, the scene and those cars are way too realistic. Except, maybe they'd look better with lower gamma.

Sandykoufax, thanks for posting that nice example of Multilight in color.

[RK_art]

Maximus3D wrote:
The sofa looks awesome!

Quick question.. where are the sss presets hidden these days ? i can't seem to find them

/ Max


sss is in the main material database

But there is only a small fraction of the original amount of SSS-presets that we had with V1.7
And btw., the SSS-'wizard' is totally useless as it has no SSS-presets to choose.
Also I cannot find the SSS-conversion-option of V1.7 although is still descriped in the V2-manual.
And finally, the integration of the SSS-section in the BFDS makes it more confuse to use SSS.
What BFDS-settings should I choose when I only want a SSS-Layer ?
From my first view on V2, there are a lot of drawbacks with using SSS compared to V1.7.
I would kindly beg for an urgent update of the V2-upgrade.

[tom]

But there is only a small fraction of the original amount of SSS-presets that we had with V1.7
And btw., the SSS-'wizard' is totally useless as it has no SSS-presets to choose.

We had many complaints about the implementation of SSS in 1.7. Many users didn't find the old wizard useful and
some of them found the presets limited or in lack of compensating the absence of a practical wizard. Now with new wizard,
it's easier to build the desired look. You only need to pick the color, set the thickness of object/effect and the density.
You can also find a new SSS preview scene in MXED.

Also I cannot find the SSS-conversion-option of V1.7 although is still descriped in the V2-manual.
And finally, the integration of the SSS-section in the BFDS makes it more confuse to use SSS.
What BFDS-settings should I choose when I only want a SSS-Layer ?

A perfect conversion is not quite possible as now the parameters are merged with BSDF and the approach is slightly different.
Now, Absorption is replaced with Transmittance/Attenuation pair and you don't need to switch to an additional SSS panel.
The SSS section in the manual has a decent in depth step by step explanation for usage and parameters.

From my first view on V2, there are a lot of drawbacks with using SSS compared to V1.7.

I understand it's never easy adapting to changes and it makes you feel disappointed but, eventually you will never like to go back.

[Mihai]

Rusteberg, where can I buy that car?

Many cool renders, great to see you've adapted quickly to V2

[Bubbaloo]

Rusteberg, those are AWESOME!

[philip99]

..render with 2.0 using 1.7 file...only at SL14

[rusteberg]

thanks guys!
@ mihai, you're not the first person to ask where to get one ... you'll have to ask it's owner (see below)

car and this guy are from old mental ray project. mental ray lighting & material setup = too long. maxwell lighting & material setup = too easy (although i do wish the max plugin had/will have image preview support when tweaking brightness/contrast/saturation through maxwell material parameters in max material editor. nonetheless, the new max plugin ROCKS!)

these rendered @ 3min. First scene with car was around 45min and closeup of hood was several hours (the closeup of hood was almost too clean! i almost felt like i had to add some noise/grain to it and also felt compelled to desaturate a little ) I'm STOKED!

[rusteberg]

@ tom, i think you are right. (see below) the actual hotel scene hasn't been touched from 1.7 and the only materials that seem to not be optimized (not cleaning up as fast) are the whites (notice the white reliefs on underside of marquee). Although even whites created in 2.0 seem to take longer to clear than other materials.

for instance, here's a test of the car with white backdrop using hdri and several staged emitters. i can't find a spec of noise on any material for car, yet the same can't be said for white backdrop. i've tried white lambertian, roughness between 90-95, and values between 221-255. they all seem to take longer to clear than other materials? although the clay model test i had posted earlier cleared up quite fast with physical sky environment?




also, i don't know what all has changed in the camera model with 2.0, but the images seem to be A LOT sharper, i virtually see no need for sharpening in post.

[tom]

@ tom, i think you are right. (see below) the actual hotel scene hasn't been touched from 1.7 and the only materials that seem to not be optimized (not cleaning up as fast) are the whites (notice the white reliefs on underside of marquee). Although even whites created in 2.0 seem to take longer to clear than other materials.

for instance, here's a test of the car with white backdrop using hdri and several staged emitters. i can't find a spec of noise on any material for car, yet the same can't be said for white backdrop. i've tried white lambertian, roughness between 90-95, and values between 221-255. they all seem to take longer to clear than other materials? although the clay model test i had posted earlier cleared up quite fast with physical sky environment?

This is an illusion It's not because the base is slow, it's because the speculars are always faster in rendering reflections. They are not wasting time with calculating the microsurface perturbance like rough ones does. Also, this has always been the same and nothing new with 2.0. Furthermore, it's faster... Here's a simple same rendertime comparison between versions:

1.7.1__________________________________________2.0

also, i don't know what all has changed in the camera model with 2.0, but the images seem to be A LOT sharper, i virtually see no need for sharpening in post.

Let's call it v2 magic.

[pavel59]

@ tom,... the only materials that seem to not be optimized (not cleaning up as fast) are the whites (notice the white reliefs on underside of marquee). Although even whites created in 2.0 seem to take longer to clear than other materials...

I agree,same happens to me. Metals and dielectrics clear at a very fast speed. WHOW.

BTW: very nice images Rusteberg, I think I remember the model from some articles or tutorials. Very nice one, excellent modeling and great concept.

I attach here some sample images from a work in progress on which I did some tests, the first two images were only to set the sink metal, the lighter is closer to the real one. Both show artifacts from poor mesh settings.
The third one shows improved mesh settings, however there's still evident noise on the countertop border (sink hole) and on its reflections in the sink.

The metal was clear in approx. 6 min. on a Dual Xeon quad core machine running Win7 64 / 8GB Ram.
The three images were rendered in approx. 16min, I think it was SL 11 or more, don't remember and approx 345 bench. Resolution was 2048x1886

The two sets were just quick tries with some more elements in the scene. Res. was also 2048px. 120 min. I was mainly testing light settings which was less than obvious, quite different results from rel. 1.7
I didn't pay too much attention to some details (you can see the wrong texture settings on chairs and the floor is also pretty bad, an old image and something happened to the bump as it shifted a bit, have to investigate what happened), I was more interested in see how light worked and about noise reduction.
Noise reduction is pretty good, illumination and tone balance need more work to get used. I'll post something more interesting soon.

BTW: do you have any suggestion on light mateials (whites and light tones) ? How to setup for a better performance ?

Paolo

[rusteberg]

thanks pavel. yes, from this model, found in this scene, which was from this silo thread, which i'm going to rework in maxwell at some point...

[pavel59]

YES, that one. I remember that. Beautiful. Evrything was so nice with that good little hironic falvor .... GREAT !
And the car, it would have been pèrfect for Roger Rabbit.

I just installed SIlo, it's a great tool. I'm not into polygonal modeling but I think I'll spend some time on it, it surely worths.
It's very powerful and friendly at the same time.

have a nice time and keep doing well.

Paolo