[kraemerJK]

Maybe we can all agree on : It should get faster. I for one thing do not want to go biased. That IMHO would mean I had to put up with all that light faking here and there to make things look just right. So if the render times improve with the next update I for one thing would be happy.

And the client stuff is not THAT easy to handle, because clients are used to a "last-minute-changes-are-possible" attitude. Just last week I had a guy who wanted (10 minutes before printing btw) to change all facades on all views. Even with the fastest render engine out there that would be next to impossible. If my clients go for maxwell quality, I distinctly put in the estimate the rules that apply. Quit simple rules actually : NO CHANGES POSSIBLE AFTER THIS OR THAT DATE. That tends to work.....

Still, faster is always nice, but not If we compromise quality. And I would love to see all those promised features and stuff work before we all send NL down the biased road in order to have both systems next to each other...

[lllab]

Hi thomas,
no i havent seen the thread on vray forum, the car thing is what i tell my customers normally.

they also want highest quality at no price;-) still i want a good price for good quality.

maxwell really feels fine on my setup, but i have plenty of cpu power. there are people even rendering faster like daros- so it is at the end just a matter of cost/money...

therefore i compare maxwell with a ferrari-it doesnt have a fiat engine at all!, it has the best-see the result. and if you have the cpu power it is really fast, also for interiors. render on a decent small office renderfarm setup( a bunch of quad core pcs) and you have renderings in high print quality in 3-4 hours-indoor. on crazy setups in like 30 minutes.

as maximus said, i also agree, it would be ridiculous to lower quality to be faster. the same quality with higher speed seems wishfull but i fear is very unrealistic.

of course i understand you, but i really feel vray p.e. is the right thing for the people wishing for fast biased-good solutions.

cheers
Stefan

[thomas lacroix]

yes illab, thats why i check your website often for news

both rendering engine interest me, but the fact is that i 'd like to use maxwell more than i do actually, but i cant as its still to long on my new ( not anymore in fact) dual dual core opteron, ( i use to have aP4 2.4 Ghz for rendering , so already decreased the time for a render by 5 to 6 time but it still not enough, i cannot get a decent interior render in less than 20-30 hours and cannot afford another computer )

[lllab]

well, as i said i understand you. and thats why i use vray or FR for things that need speed.

cheers
stefan

[thomas lacroix]

got your point illab, just waiting a maxwell update or any news concerning your plug for cinema4d

cheers

i'll still continue to run test to see whats possible thought

[Ernesto]

Of course we would like an instant render at infinite speed, but:
I love Maxwell quality.
If there is a way to make it faster without compromising actual quality, I am sure next limit will do it.
As for the speed of current version (1.1) In my oppinion it is on the limit.
I will try to explain it better: It is possible to do serious work, although it is slower compared to the competence, but It could be handled. In case of a client that wants a faster render, and cannot see the diference, go ahead and use other package!
Anyway having said all this, if it could be twice faster I would be more happy!
But quality if first!

Ernesto

[JTB]

I would like to support my opinion from the 1st page. I think there is a missunderstanding here.
NOBODY wants Maxwell to go biased.
ALL of us want Maxwell to be faster.
So, what if we can have the choise between Maxwell and biased Maxwell?
That means having one application, one material library, one model file, etc. BUT!!! we can CHOOSE (<=the key word) If we want Maxwell to make us a perfect solution or if we want Maxwell to make a quick solution with acceptable or even pretty results.

That's what I'm saying. That's what I would think ideal.
I still think that many of our clients don't understand the difference between a biased and unbiased render, and most of the times they don't care. For them it's just a nice picture or not...

[-Adrian]

Understandable but don't you think they simply reached the limit of how fast they can go without bypassing the necessary calculations? I'd say they probably spend a lot of thought about optimization. Sure they could try to make use of EMT64/AMD64 and whatever SSE2-4 additonally have to offer but then people would again be complaining because Maxwell won't run on their last gen PCs and MACs.

So what will "just make it faster please!" accomplish

If we want Maxwell to make us a perfect solution or if we want Maxwell to make a quick solution with acceptable or even pretty results.

How does RS0 perform for you?

[JTB]

IMHO maxwell can be faster. I don't care if they name it optimization or RS2 or whatever. Every application can be improved.
I am not complaining, wishing and looking for improvements is not complaining, I knew all about Vray and the other renderers I decided to buy Maxwell, it was not a birthday present. I have supported NL with some of my posts during the RC difficult days.
I am just looking for a quicker tool (even if it is cheating a little) for my work.

RS0 might work I don't know yet but I think it wouldn't be a big difference.

[lllab]

JTB i think all understood what you mean.

if you know how to make maxwell faster maybe tell nextlimit;-)

...no honestly, i am very sure they do their best. they want to sell their product and it would sell more if it would be faster with the same quality.
i am very sure nexlimit does their best. if it would be so easy as you suggest they would have done it.

stefan

[shaun]

I agree 100% with JTB. We need more speed in some way or the other.

I hate 'car anologies' but what's the point of having a ferrari in the garage but one cannot put any fuel in it to show it off. I've had Maxwell for almost a year now and it hasn't earned me money yet. It's just a fancy toy to play with at the moment for me.

I know NL works hard and I do appreciate it but we do need RS2 (I'm not complaining, its a wish). For me Maxwell needs to increase its speed 10 fold to become usable. Not everybody has 20 supercomputers in their offices so we somehow need speed improvements in the software.

Thats my 2 cents.

[Miles]

I hate 'car anologies' but what's the point of having a ferrari in the garage but one cannot put any fuel in it to show it off. I've had Maxwell for almost a year now and it hasn't earned me money yet. It's just a fancy toy to play with at the moment for me.

What's the point of all these car analogies?

If you had a Ferrari in the garage it would just be a fancy toy to show off with

[shaun]

If you had a Ferrari in the garage it would just be a fancy toy to show off with

That's true

[lllab]

"cannot put any fuel in it to show it off"

..well you can, just buy enough cpu power. i have made several projects with maxwell, which brought my manyfold back my maxwell investments.

you know how fast maxwell is on a quad woodcrest- it is already quite nice.

of course this makes maxwell a rather expensive solution. but in a few month a stanard PC will as fast as the new quads now, and most people will be happy with speed i guess.

the spectral renderengines is new technology that also needs very new hardware at the moment. but what is expensive now is stndard hardware soon.

my thougts
cheers
stefan

[shaun]

Not all of us are lucky enough to be upgrading their computers every six month for a R20000 system a pop. Especially if one isn't getting the jobs for it.