Page 2 of 2

Re: the light source inside the glass "NOISE"

Posted: Mon Oct 13, 2014 8:04 pm
by Karba
Goldisart wrote://////

Tell me that I don't sleep ....?!?!?
Tell me why I bought a graphics card for $ 2000 ?!?!?!
.... You need urgently to persuade me not to buy a render farm for Vray :twisted:
Just go and use vray.
looks like you use vray network rendering with 4 pc. Why don't you use octane network rendering with 4 pc?

Re: the light source inside the glass "NOISE"

Posted: Mon Oct 13, 2014 8:21 pm
by glimpse
Goldisart, when You're comparing different engines, please compare same scene with same circumstances,
'cos You are not making a point..just a lot of noise =)

You're not sleeping 'cos You don't know the engine, don't know it's strengths & weaknesses,
& why did You buy a card for 2000$ ??? - well, I guess You don't know about hardware as well..

so don't go around the forum with Your offensive tones,
'cos nobody is going to help if You act like this..

there is plenty of guys around who has similar kind of problems, but before they make some kind of accusations,
firstly the assume that they don't have enough competence & start looking in the forum for information & solutions..

if they don't find a solution, then they ask for help..

Re: the light source inside the glass "NOISE"

Posted: Mon Oct 13, 2014 8:28 pm
by glimpse
what is Your final scene? what is Your target? what effect are You trying to achieve?

balance Your lighting, optimise mesh emitters, then start with other issues - now we don't know what You're doing..no technical info, no target presentation is known.. - You might not need to get rid of noise in Your wips as final image will have completely different lighting.

start from Your target. Deconstruct entire scene. Try to identify what is causing the noise exactly.. - fro msimple prinscreens we can say much about Your scene - put wire frames etc..

Re: the light source inside the glass "NOISE"

Posted: Mon Oct 13, 2014 8:56 pm
by Goldisart
Thank you that you started talking to me!!!!! I broken the translator --- will try in some way to Express my opinion. I understand how this is funny to read but I don't know English. 1. I have great respect for your team. The team who created this project. Octan render!!!! 2. I always love the render and make it a decent well-paid commercial projects!!!! THANK you !!! I raised this question on noise only because I want to see octane as a universal render and not to jump from one renderer to another. If there is a problem with the glass and my commercial project begins to rotate due to the fact that I have to wait 20-30 minutes to 1 picture and I was planning on 3-6 I get into the knockout. The client says I want that would showcase glowed = all I in knockout I have options: 1. to reduce IOR - but it is not glass 2. falloff opacity - but the noise does not disappear 3. falloff reflection is not bad but it's not glass 4. PMC - Yes but the knock-on time .... if this question is not to decide and write I'm just going to warn customers and to make adjustments at the time of my commercial projects

Re: the light source inside the glass "NOISE"

Posted: Mon Oct 13, 2014 9:10 pm
by Goldisart
GTX 780 - 6GB + GTX 780 - 6GB + GTX 590 = three video cards must be replaced at least 9 CPU - I only have 4 cpu 4*8 = 24 cores easy to skip through the glass light - and it turns out that there are a bunch of fireflies due to the reflection..... Yes and the client wants to see pure gloss - this is a problem.. I don't know whether it is possible to make a fake octane - such as it is implemented in CPU rendering). If not, PMC or more render time me time - time is money ! Octan render animation allows me to do very quickly!!!! it quickly loads the large scene - when the render farm still think. In octane is very much a plus for it and love it

Re: the light source inside the glass "NOISE"

Posted: Mon Oct 13, 2014 9:36 pm
by akira
I replaced one of the mesh emitter with 3 octane lights, and set the lights' samples to 10000(max number)
And use linear response curve for the imager setting.
The result is obviously better than before.
mesh_emitter.JPG
p.s. the unit setting of your file seems wrong, the "Display Unit Scale" of unit setup panel is not the "real" unit settings.
The "System Unit Setup" button is where we set the unit of the scene, it's millimeter in your file. It should be centermeter in this scene.
That's why you had to lower the power of lights in this scene, because everything is 1/10 to their real size.

Re: the light source inside the glass "NOISE"

Posted: Mon Oct 13, 2014 9:54 pm
by Karba
akira wrote:I replaced one of the mesh emitter with 3 octane lights, and set the lights' samples to 10000(max number)
And use linear response curve for the imager setting.
The result is obviously better than before.
mesh_emitter.JPG
light samples don't make any sense here. You reduces noise for one light source but increase the noise for other light because they get less samples.

Re: the light source inside the glass "NOISE"

Posted: Tue Oct 14, 2014 12:16 am
by akira
Karba wrote: light samples don't make any sense here. You reduces noise for one light source but increase the noise for other light because they get less samples.
Ouch, you're right.

Well at 16k samples/pix(path tracing) i can get rid of the noise, it looks more like film grain than noise.
mesh_emitters_16000samples_PT_glossy64.PNG
p.s. Mr. Goldisart about the shell modifier, you need to check the "Straighten Corners" option at the bottom of the modifier panel to keep corners straight.

Re: the light source inside the glass "NOISE"

Posted: Tue Oct 14, 2014 2:37 am
by coilbook
Karba wrote:
akira wrote:I replaced one of the mesh emitter with 3 octane lights, and set the lights' samples to 10000(max number)
And use linear response curve for the imager setting.
The result is obviously better than before.
mesh_emitter.JPG
light samples don't make any sense here. You reduces noise for one light source but increase the noise for other light because they get less samples.
nice to know. i always crancked them all to 10,000 each i did not know you can take samples away from other

Re: the light source inside the glass "NOISE"

Posted: Tue Oct 14, 2014 5:00 am
by Goldisart
I think that this is not a problem of light .... more it seems that this is the "problem of reflection" (a distinctive feature of the photographic grain). the reflection in the glass synchronously with the reflection of the floor which is usually for the GPU. When the lights are turned on it starts to produce noise === reinforcing it several times and begins to intensify film noise. in this situation it is necessary to wait until the environment becomes clean