Cinema 4D living room

[strangevisitor]

some objects are out of scale. the balcony door in your picture has a 110 mm / 220 mm ratio.
there are doors like this, but visually more convincing would would be a 80mm / 220mm ratio.
measuring the couch height by your windows/doors , the couch has a height of 100-120 mm,
something in between 70-80 mm is more common. the same for your wall lights.

the purple vase in the left foreground tilts your picture to the left / the eye is attracted to this
region.

[kubo]

wow, radiance is right, there is something very wrong with that vase, could it be the IOR? it looks like it has an IOR of 1, it should be around 1.5 or higher.
Strangevisitor, it could be a low ceiling flat, and thus the strange proportions, if it's an old building (40 years or so) and specially a penthouse, it could easily have 2,30 or even 2,20 m inside height, I've seen quite a few in my country.
And katerlake, good job, add more detail, and you'll be there in no time.

[buda]

bad that the noise for a long time away. the difference between the first version and second to none. With this algorithm, there is no significant advantage in GPU rendering. Starts fast, but gets rid of the noise is very slow ... in Maxwell can get the same picture of the noise on the CORE I7 and probably for the same time..

[strangevisitor]

bad that the noise for a long time away. the difference between the first version and second to none. With this algorithm, there is no significant advantage in GPU rendering. Starts fast, but gets rid of the noise is very slow ... in Maxwell can get the same picture of the noise on the CORE I7 and probably for the same time..

i am a bit unsure about this whole gpu based unbiased rendering , but i highly doubt you can render this

http://www.refractivesoftware.com/forum ... d=3220&t=1

in maxwell/fry/indigo/thea in ~17 min. the limitations of gpu rendering is imho not speed but space.
VRAM is a very limited ressourece. because of this implementation of features like micro polygon
displacement or subsurfce scattering is promised by random control/ refractive software, but the
quality / the fact of implementation is unsure imho. the speed is real, the question is, will it be
production ready without buying gcrads for 3000 € and more ?

unbiased gi gpu renderers are ike prototype cars in the public - 50% of the body is camouflaged atm

[radiance]

bad that the noise for a long time away. the difference between the first version and second to none. With this algorithm, there is no significant advantage in GPU rendering. Starts fast, but gets rid of the noise is very slow ... in Maxwell can get the same picture of the noise on the CORE I7 and probably for the same time..

buda, i'm getting a bit annoyed with your constant negative comments.

maxwell render has been in development for 6 years now by a company with several developers. offcourse it's got more effective algorithms, it also costs an arm and a leg.
it's also WAY slower for outdoor scenes or product viz.
the fact that octane can match it's speed for interiors with a much simpler algorithm only shows what the future will bring, when octane is 3-4 years old.

i suggest next time when you make comments you think twice before writing dumb posts.

Radiance

[kubo]

Look, Maxwell, and the other unbiased cpu based renderers are good, really good, nobody will deny that, but the thing is (price aside, althou that's a big difference), that they are no match with octane, which like radiance said is just in it's beta, I'm not saying this as a fan boy, I can perfectly see the fireflies camping my renders as anyone elses, but those are easy enough to get rid off and I can have a production render in under 30 minutes with good resolution (MLT as radiance has stated time after time will solve that in interiors), that I can hardly get with any of the best, plus is easier to navigate, make changes and preview saving lots of time to get the picture right.
All my clients want not only perfect results but ridiculus deadlines, with lots of changes in the way, the usual call I get 1 day before deadline is "hey, I've been thinking, we should change the total heigh of that building, plus I think that the north facade would look better if instead of wood paneling we used monkey heads on a stick",2 hrs later they call to make those monkeys fur blue....
Well as you can imagine, for me unbiased renders were totally out of the question up until now. I can indulge all those changes with the superb quality of an unbiased renderer thanks to octane, and I save huge amounts of time in tinkering with all those pesty options that plague any biased renderer, plus I get shots that I would never have found before, just because I can perfectly see them before hand.
In the short time since I bought the license I've used octane for 3 pro-bono jobs, plus 2 profesional jobs (besides my personal projects) and I have added it to my workflow with almost a flat learning curve and gotten better and faster results than before, true enough there has been one, with over 1M instances between cars people and trees, hogging all my 12 Gb of ram in a CPU redender for a 8000+ resolution, but even then, while those were rendering, I had octane render a different project, both in the same machine, at the same time. So fire away your resentment all you want, but for me Octane is the future, while the other unbiased are dying dinosaurs, and if you don't believe me check fryrender and arion, and see wich of those 2 are going to survive.
Cheers, and sorry for the long post
Also katerlake sorry for highjacking your post for an OT rant

[FamilyGuy]

It seems to me that the problem with your render is a glass material in windows and a very small openings by which the light penetrates room. I did some test with Octan and try to simulate Spot Light by making small holes in the ceiling. After 16000 samples I still had noise. I think the emitter material and light portals will solve the problem.
But until that you shold delete glass material and give a little more time to Octan.

[buda]

bad that the noise for a long time away. the difference between the first version and second to none. With this algorithm, there is no significant advantage in GPU rendering. Starts fast, but gets rid of the noise is very slow ... in Maxwell can get the same picture of the noise on the CORE I7 and probably for the same time..

buda, i'm getting a bit annoyed with your constant negative comments.

maxwell render has been in development for 6 years now by a company with several developers. offcourse it's got more effective algorithms, it also costs an arm and a leg.
it's also WAY slower for outdoor scenes or product viz.
the fact that octane can match it's speed for interiors with a much simpler algorithm only shows what the future will bring, when octane is 3-4 years old.

i suggest next time when you make comments you think twice before writing dumb posts.

Radiance

I do not write negative, I am writing about the obvious problems that I see. And you do not accept criticism appropriately and go to the person.

[Chris]

But your statement is wrong Buda, thats probably why

Do you think Maxwell only uses pathtracing for its rendering algorithm?
It doesent, it uses both MLT and bidir pathtracing.

So you cant really compare them for now.

And you cant really deny the parallel processing power of gpu`s compared to cpus.

[radiance]

let's get back to the topic pls,
we're supposed to be giving tips to katerlake for his interior render...

Radiance