[Julian]

hey Mihai's images have something about them - the first i've seen from all the recent tests that even vaguely resemble the beta 'look' we all crave... a few more like this and my hopes might begin to climb

[jleckron]

question for Mihai (or whomever) re. the new images: You say

The bump is just a noisemap applied to this layer and no shapes necessary

- does this mean that we have now have procedurals, or just simply that you had an img with noise on it that you used for noise?

[The Pixel Artist]

With all due props to Adrian (damn good work there) that's not the "only" image Vray could make from that model. With more CPU time the GI could be better - though I don't think it's necessary, and I think the render time here would put Maxwell to shame.

And (correct me if I'm wrong) I think I see some things in there that Maxwell can't do. Displacement for the grass and roof? Alpha maps in the plants? Do I detect some procedural textures and a sky photo in the environment that's being reflects by the window glass, which has a slight "oilcan" bump effect and is not noisy?

ajlynn: Thanks man, you just saved me a lot of typing time. That was the point there.

JDhill: Sorry bud but you obviously totally missed the point of the vray test image. The whole point was to show that Vrays "unbiased" option had no problem with the same issues that Maxwell has with glass and has yet to prove otherwise. The quick test to see if anyone would be fooled by it being a Maxwell image was purely secondary for fun (I made no real attempts to make it look just like a Maxwell image). What I was really hoping was to throw down a challenge to the "A-Team" to reproduce such a simple scene and prove Maxwell could do it. Whether it could do it prettier or not is irrelevant.

And as far the example quality Maxwell image... well ajlynn said it very well again. Now for some better examples of Maxwell’s up coming competition that proves that we're going to have Maxwell’s relative quality in much faster/more functional/better programs.

Here’s a couple from Modo (sure reminds me a lot of Maxwell (or even better Arnold)
http://www.luxology.com/modo/201/img/Re ... _front.jpg
http://content.luxology.com/modo/201/im ... 1_85_2.jpg
http://www.luxology.com/modo/201/img/Re ... nPlane.jpg

Heres one for Vray using PPT (for Rhino)
http://www.asgvis.com/images/asgvis_rhino05_lg.jpg

And finally here's another from Modo to go with Toms new ones (I bet this one render a laughable bit faster)
http://content.luxology.com/modo/201/im ... 1_xmas.jpg

Oh and JDhill, if your happy with Maxwell for your production work, well, I just feel bad for how slow/unproductive your pipeline must be. And am I totally happy with Vray in mine? Nope. In fact it was this time last year that I was already looking the next level. At that time Maxwell could have been it. But NL and their marketing lied, they stole $400 dollars from me, and this year I’m much more excited by the future products of Vray and Modo. But then I would still like to get something out of Maxwell since they wouldn’t honor my request for a refund. (Boy NL if you only had given me that refund you would have been totally rid of me. Plus, I can only guess how much more than $400 dollars you’ve lost from the amount of people I convinced the your not their best option)

[Maya69]

ok pixel artist

all people says that the v-ray is very good engine .....

for the moment maxwell is not ready for production

but waiting ..... and few week you can continue a discution


for the moment you post only sterile post

salut pixel artist

tous le monde sais que v ray est un super moteur de rendu

je suis d'accord maxwell pour le moment n'est pas utilisable pour la prod

mais attendons .... et dans quelques semaines on pourra reprendre ce débat

pour l'intant tes messages ne nous apportes rien

cordialement

[The Pixel Artist]

Maya69: Actually I have been very patient with Maxwell, not only have I waited for a whole year, but I've also refrained from really saying anything through RC1,2,3,and 4. It wasn't until their non release on RC5 that I though I better give them some crap again.

But, for now I'll get back on topic....

NL/A-Teamers: can you prove (or at least show) that the currect version of Maxwell(RC5 or RC5.7 or RC2+++ whatever) can render a scene just like the Vray example I gave in a reasonable time-frame (say 4-12 hours) with simular or better image quality?

[Maya69]

i am agree with you

but wait ....

i just want says "we know that v-ray is very good ......"

best regard

[The Pixel Artist]

P.S.

Those game machines can put out some amazing images, but from what I understand they use a *lot* of tricks to do it. In other words, even if it renders almost instantly, it takes a lot of work on the front end.

Well the reason they use "tricks" is because it use is for a game. They goals are to produce the highest quality images in a 30+ FPS timeframe. These are not meant to be a HQ rendering system (impressive none the less). Luckly for us we don't need it to be realtime. A HQ rendering engine could be developed today that could provide the image quality of at least Vray and I'm sure Maxwells "reletive" quality in minutes or seconds.

BTW, the point on hardware rendering - have you seen nVidia Gelato?

Yes I have, quite familer with it. Good example but not very good. Better examples would be Pixars latest "relighting" tool or Wetas "Jet" version of the Massive program. These are again just a sign of whats to come very soon (and should already have). If there isn't a HQ option in the next 1-2 years then I hope to provide one in the next 2-3 (hopefully business goes good this year, exciting things to come) Again just image what would be possible with a quad SLI nvidia 7800 system, and Gelato could still turn into something good.

[Robbe K]

OK Guys! Relax a bit!

I am a VRay user, both Max and Rhino and I am happy with it! Really!
I mainly use VRay for animation
For stills, I think the setup time/final output time relation is not in balance! Meaning: It takes a long time to setup a scene and to render it!
My experience with Maxwell BETA is, that I can setup a scene within an hour and have a really good picture the next day with proper lighting, no light leaks and all that hoela that you can have with VRay, if not setup properly! And Yes: I am an experienced VRay user!
Botomline: both great, but for me in different ways!

I know => we are all not happy with the current situation concerning Maxwell, but come on guys! It is just a render engine hehe

[Maxer]

Even if it could be developed, how much would it cost?

[The Pixel Artist]

Adam: Evidently not, because it's more than possible. That doesn't mean that there isn't some out there secretly developing it now. As much as things change in this industry it's movers and shakers have always had the problem of not seeing the obvious and simple ways to change it for the better. A good example of the rare times it does would be Sketchup, Modo, and even Maxwell at one time. The thing is there has been nothing technologically stopping them from being made even 5-10 years ago, it only had to be invented and implemented. Same as there's nothing stopping us from inventing the software/technology of 5, 10, 100 years from now "today" other than thought.

Maxer: Why would it have to be any more expensive than what we have now? The point would be to use existing graphics hardware (not custom hardware like past attempts, cough, cough, Pure... crap)

[max3d]

This topic is competing with an off-topic topic called the mother of all topics. (challenge: make a coherent sentence while using 'topic' more than I just did). We cover everything but the taste of whiskey

I agree with pixelartist that you could have written some renderers years before people did. The research papers which are the underlying logic, are in most cases years and years older than the actual commercial renderers. However you would go broke if you did as the hardware capabilities would prohibit the usage of your product. So, it's as always a question of timing.

Maxwell seems to be right on time (pun intended) as 2006 will be the year that dual core, very fast cpu's are readily asnd cheaply available. For stills at least, the promised RS2 when distributed will be useable in production (and deliver exceptional results with minimal user fiddling).

I understand that the current delay's are very disturbing and I completely agree as you all prepaid. However if you were a new prospect looking for a renderer and not in a hurry I would certainly shortlist Maxwell for mid 2006 succes.

What, pixelartist, is wrong with Pure hardware? I never tried it, so I'm really interested.

(btw I would suggest that Mike starts a new 'Amazing Mike' thread with a short resume as this thread keeps on drifting)

[Eric Lagman]

This thread has been destroyed, and I just contributed to the destruction by typing this sentace pointing it out. I go back to my hole now..........

[mverta]

No, no... any thread that's like 40 pages long has to have some organic ebb and flow to it to survive. Don't sweat it. If it really becomes ungainly I'll lock it and we'll start over. But right now it's like a soap-opera: you never know what cliffhanger and drama is right around the corner.

Play on. I have to go back to my testing hole.

_Mike

[The Pixel Artist]

I too have wasted way to much work time today so one last quick reply...

Adam: "On what do you base that conclusion?"

Well… on my fairly extensive knowledge of where computing hardware is today and where it's been. But the history and details of what makes it more than possible today (and quite practical for a few year) is a long topic not meant for today or this thread. Quick hints; 128 bit depth, programmable hardware, long/unlimited shade lengths, distributive tile rendering (current trend).

Max3d: "What, pixelartist, is wrong with Pure hardware? I never tried it, so I'm really interested. "

Oh boy, really, really long topic there. Again too long for here and today. Plus it's been beaten to death. It could be a 40+ page thread on it's own. We'll just say it's not the best option for cost vs. quality reasons.

Ok, sorry, last one for the day (I'm sure many will be happy though)

[ludenhud]

The fastest way of finding bugs would be to release a NOT FULLY WORKING VERSION of maxwell, with a CLEAR underlined text saying that it's for testing propose only. You beta testars (atleast mverta) are doing a remarkable work. But It's taking too much time for you to go through all sections of it's capacity.