[ludi]

Rivoli:
Did whiskey left the board?

[rivoli]

well, it's just another way to put it. i don't want to get into that matter another time, it's been discussed already and i don't see any point in starting it over again. but anyway, he kinda committed a "forum suicide", didn't he? still i do really miss him hanging around here.

[daros]

In my experience sun light goes trough glasses since version Alpha1.33.
The problem was the much bigger sampling level needed and, as Rivoli demonstrated, the extesion of the scene, and the image protion and image centering accupied by the caustic.
The other limitation is that if you use "render region" the caustics are not correct. This rendering is made on a single CPU in 140 hours with Alpha 1.33, no noise reduction.

[daros]

Hi MVerta, this is a very old post of mine focusing the issue.
http://www.maxwellrender.com/forum/view ... ight=glass

[daros]

Hi buffos, i's difficult to speak with you but i try.
they are other commercial rendering engines that claims to be able to doe caustics, rendering, dof and many more. For example Lightwave.
The problem is that you can spend 100 hours of work for a 1000 hours rendering and you don't will obtain this quality in a Lightwave Rendering.
But Newtek claims their rendering engine is able to doo that.
And now?

[rivoli]

then its is almost as if it does not pass.

david rendered that with the alpha, which was slower. slowness can't be cosidered a bug with maxwell, it's one of its main features.

[aitraaz]

well that pic is nice (and your work is great ) - but the issue is that for an 130 hr rendering, the noise/blurred glass is a bit problematic. I'm no expert, but i think something like lightscape would do better (in less time) with the same situation. Of course, once maxwell (and maybe it already has) resolves all these glass issues, so much the better...(no dof in lightscape, grant you...)

[daros]

Here is one of my first post comparing Maxwell to Lightwave.
http://www.maxwellrender.com/forum/view ... =lightwave

So the fact is that Maxwell is able to simulate caustics through glass. The only problem is the time. However it is faster as other commercial softwares like Lightwave and probably as Vray if you use Path tracing with surfaces with diffuse reflectance, caustics, 16 light bounces, aberration, DOF, vignetting, and a area light for the Sun.

[The Pixel Artist]

Just to requote myself

Mike,

Well really if you want to end this debate once in for all you also need to prove that the lighting resolves not only throught the glass but also as seen through the glass.

For example: (very old image)


Here are a couple test scenes (in a zip file). If you can render these without problems, well, there yet may be hope for this product.

http://www.thepixelartist.com/storage_f ... scenes.zip

As you guys have point out again very well it's not just simply does light/sunlight (not that in reality "sunlight" is something different) passes thought a transparent mass "once" but in a way that accurately simulates reality (with near infinite passes).


So far Maxwell has been far from realistic in this manner, despite whatever their marketing claims or deceptive test images you show us. IMHO any modern rendering system is fairly worthless without the abilities to "display" this type of realist light transmission, in a timely manner of course (whether calculated in a biased or unbiased way, as either way is still a poor hack to simulate reality)

Mike, again we appreciate your efforts (I'm sure they were genuine), but unfortunately this scene you're given us is a poor test for what we're concerned about. So with that said I challenge any you fab "A-Teamers" to render these two test scenes I've provided in RC-5 in which ever RS core you think represents Maxwell best capabilities (RS1, RS1+, RS2, or RS2++++get-er-done++++). Actually it would be cool if you could render them with all your RS's so we could actually see the difference.

So A-Team, got the cohunes?

[daros]

For Adam.
I know, but my answeare was for buffos affirming "well you say that the light comes in but i will need 200h to render, then its is almost as if it does not pass".

I was demonstrating that other packages clamis to doo some things they don't do under Buffos point of view.
An normal user can say: ok Vray makes beutiful images but if i need to learn all the possible fakes for two years it's better for me to use Maxwell.
That is exactly wat happen.

[rivoli]

anyway, even if mike was not aware of all the sun/glass stuff cos he never really look into them before, i'm pretty sure other testers (tom, thomas, mihai...) are. so i guess they will look into them, if they hadn't already, and won't realease another version with the same old problems.

[tom]

I'm here adehus, as usual.

[mverta]

Hey guys -

Just got up; been reading through the last couple of pages, and learning more about this issue than I'd ever have cared to

I'm already talking to the men behind the curtain to find out once and for all what the story is on scenarios beyond what I set up, and I'll try to get clearance for a full description that addresses of all your concerns (I've noted them).

I don't know if I'll get it, especially because I'm supposed to be testing about a million other things right now, but it's clearly an important issue to you guys, and at this point, I'm kind of curious.

Stay tuned.

_Mike

P.S. I doubt I'll be around much this week, but I'll try to be. I just have more to do than usual, so don't sweat it if it takes me a bit to reply.

[Frances]

I noticed the note "5 hrs Rendertime. No optimization" on Mike's render. If that refers to optimization of the engine, then that would explain the long render time. I'm not convinced that I should remain unconvinced.

I'm not throwing confetti, but I see progress.

[tom]

Hey Tom... I meant I was hoping that you would say something regarding the sun-glass issues, and if you think that the rs1+ handles it better than rs1.

We'll see and when we see, you'll know.