A LOT TIME WITHOUT LIGHT Tweet

[senorpablo]

Hi paulforgy,
https://render.otoy.com/forum/viewtopic.php?f=30&t=3001&p=29248&hilit=system+unit#p29248
https://render.otoy.com/forum/viewtopic.php?f=80&t=64441&start=30#p329566
Regards

I did a quick test. In one image, the system units was meters, in the other, inches. In both cases, the plane and light are the same size and power. Since light falls of with the square of the distance, it would be noticeable if Octane didn't properly handle system units other than meters. I think it's safe to say that for purposes of lighting and object sizes at least, Octane handles things correctly regardless of Max system units. A similar test could be done with material settings to see if they are effected by system units.

Meters:

Inches:

[paride4331]

Hi paulforgy,
this test is obvious, you said they have the same size, so what is your test? Power is in watts but adjusted to the surface, if surface brightness is on.
Scale issue are if you think working with system unit instead of another or when system unit is not considered.
This happens importing assets, with materials (transmission, medium, displacement ...), with camera values (DOF ...), post processing.
using unbiased render engines, unit of measurement is not very important and many users who have experience with them, have render artifacts with biased render engines.
Regards
Paride

[frankmci]

What I wanted to express is that the first image with more objects and a single light input that is a black body does not generate noise and the second image with the same parameters and fewer objects and everything in the same proxy, generates more noise and therefore the increase of the maxsamples

Ah, ok. I understand. I agree with paride4331, that your geometry position relative to the light source has a huge impact on this kind of very diffuse lighting environment. With the curtains and the sofa so close in front of the light source, you need many more samples to evenly illuminate the interior. I just did a quick mockup of your scene, and the difference in noise levels is quite dramatic, based on curtains and sofa in front of the window or on the other side of the room. To achieve similar noise levels for both layouts, I had to increase the sampling rate and render time (and light power) to a similar degree that you did.

So, in short, your render is behaving as it should, as far as I can tell.

Here are two renders, same settings, same render time, the only things different are position of the geometry and the power of the area light to keep overall illumination close to the same. I used very low sampling levels to exaggerate the issue.

[senorpablo]

Hi paulforgy,
this test is obvious, you said they have the same size, so what is your test? Power is in watts but adjusted to the surface, if surface brightness is on.
Scale issue are if you think working with system unit instead of another or when system unit is not considered.
This happens importing assets, with materials (transmission, medium, displacement ...), with camera values (DOF ...), post processing.
using unbiased render engines, unit of measurement is not very important and many users who have experience with them, have render artifacts with biased render engines.
Regards
Paride

The test is that system units aren't a factor in terms of lighting and geometry. Doesn't matter if they're set to millimeters or miles. You were suggesting in other posts, that Octane only worked correctly when the Max system units was set to meters. You said: "Octane Render works in meters, then system unit scale 1 unit = 1 meters..." This is wrong. If Octane didn't correctly interpret system units, there would be a clear difference in light falloff in the renders I provided. There isn't. This is because the power setting of the light didn't change, and neither did the size. What did change was system units. If Octane only worked with system units set to meters, there would be a huge difference as there are almost 40 inches per meter. Since light falls off with the square of the distance, the visual difference would be enormous--much larger than 40:1.

If there are settings in the Octane camera, materials, or kernel settings, for example depth of field or density, that are dimensional, and the Octane plugin isn't correctly interpreting the values from Max, that's an issue with the Octane plugin--an oversight. Max provides everything needed internally for developers to correctly deal with both system and display units of any variety.

Having objects and scenes built to scale and the correct size is important, but you can do that in any system unit setting, not just meters.

[merid888]

hello, thank you for commenting, a single light input that is a black body is the same used in the second image with the same parameters and the proxies are independent in all objects

[paride4331]

Hi paulforgy,
this test is obvious, you said they have the same size, so what is your test? Power is in watts but adjusted to the surface, if surface brightness is on.
Scale issue are if you think working with system unit instead of another or when system unit is not considered.
This happens importing assets, with materials (transmission, medium, displacement ...), with camera values (DOF ...), post processing.
using unbiased render engines, unit of measurement is not very important and many users who have experience with them, have render artifacts with biased render engines.
Regards
Paride

The test is that system units aren't a factor in terms of lighting and geometry. Doesn't matter if they're set to millimeters or miles. You were suggesting in other posts, that Octane only worked correctly when the Max system units was set to meters. You said: "Octane Render works in meters, then system unit scale 1 unit = 1 meters..." This is wrong. If Octane didn't correctly interpret system units, there would be a clear difference in light falloff in the renders I provided. There isn't. This is because the power setting of the light didn't change, and neither did the size. What did change was system units. If Octane only worked with system units set to meters, there would be a huge difference as there are almost 40 inches per meter. Since light falls off with the square of the distance, the visual difference would be enormous--much larger than 40:1.

If there are settings in the Octane camera, materials, or kernel settings, for example depth of field or density, that are dimensional, and the Octane plugin isn't correctly interpreting the values from Max, that's an issue with the Octane plugin--an oversight. Max provides everything needed internally for developers to correctly deal with both system and display units of any variety.

Having objects and scenes built to scale and the correct size is important, but you can do that in any system unit setting, not just meters.

Hi paulforgy,
I do not say it's right or wrong, I said, guys pay attention to unit size, Octane thinks in meters, that's all.
I think it is just right to have my workflow consistently on the scene. We can say that it would be better if Octane adjusted its parameters to 3dsMax unit size and maybe in the future it will be.
But today, be careful and remember this rule.
Regards
Paride

[senorpablo]

Hi paulforgy,
this test is obvious, you said they have the same size, so what is your test? Power is in watts but adjusted to the surface, if surface brightness is on.
Scale issue are if you think working with system unit instead of another or when system unit is not considered.
This happens importing assets, with materials (transmission, medium, displacement ...), with camera values (DOF ...), post processing.
using unbiased render engines, unit of measurement is not very important and many users who have experience with them, have render artifacts with biased render engines.
Regards
Paride

The test is that system units aren't a factor in terms of lighting and geometry. Doesn't matter if they're set to millimeters or miles. You were suggesting in other posts, that Octane only worked correctly when the Max system units was set to meters. You said: "Octane Render works in meters, then system unit scale 1 unit = 1 meters..." This is wrong. If Octane didn't correctly interpret system units, there would be a clear difference in light falloff in the renders I provided. There isn't. This is because the power setting of the light didn't change, and neither did the size. What did change was system units. If Octane only worked with system units set to meters, there would be a huge difference as there are almost 40 inches per meter. Since light falls off with the square of the distance, the visual difference would be enormous--much larger than 40:1.

If there are settings in the Octane camera, materials, or kernel settings, for example depth of field or density, that are dimensional, and the Octane plugin isn't correctly interpreting the values from Max, that's an issue with the Octane plugin--an oversight. Max provides everything needed internally for developers to correctly deal with both system and display units of any variety.

Having objects and scenes built to scale and the correct size is important, but you can do that in any system unit setting, not just meters.

Hi paulforgy,
I do not say it's right or wrong, I said, guys pay attention to unit size, Octane thinks in meters, that's all.
I think it is just right to have my workflow consistently on the scene. We can say that it would be better if Octane adjusted its parameters to 3dsMax unit size and maybe in the future it will be.
But today, be careful and remember this rule.
Regards
Paride

Agreed that one should always be conscious of units--very important. Thank you for the practical example. The Max plugin for Octane should definitely either: respect display units, or clearly label dimensional parameters as such, whether it's meters, etc.

[merid888]

hi every one, add image with new parameters of path tracing
regards