Strange Material behaviour with RC1

[jitendra]

Dear Octane Team,

I have been using 2.58d until now for production and its running great.
Today I tried RC1 and found that, the rendering results are entirely changed.
All my transparent materials and metal materials are all displaying different result even with the same settings as before.
Please checkout a metal example in the images attached.

With semi transparent material this changes are real drastic.

Please HELP!

Thank you.

[matej]

A lot of things changed in Octane 3xx & RC1, from previous 2xx. Like a new color system and improved glossy materials. This changes result in old scenes not looking the same in new versions.

[abstrax]

Dear Octane Team,

I have been using 2.58d until now for production and its running great.
Today I tried RC1 and found that, the rendering results are entirely changed.
All my transparent materials and metal materials are all displaying different result even with the same settings as before.
Please checkout a metal example in the images attached.

With semi transparent material this changes are real drastic.

Please HELP!

Thank you.

There are a few things going on here and you have to look at the details:

"Saturate_to_white" is a tool to fix overexposed areas that would often end in one of the 6 combination of the RGB primaries (red, green, blue, cyan, yellow, magenta). If you set it to 1, those overexposed colours would become white instead. Your image is totally overexposed and setting "saturate_to_white" to one causes the grid of the floor to disappear.

-> When you set up materials you should make sure that the exposure is not too high.

The flatness of the metal is actually correct and caused by the introduction of an index parameter in the glossy material, which is by default 0. The "index" parameter represents the index of refraction of a coating layer which is often causing a Fresnel effect in the reflection (i.e. shallow viewing direction causing a higher reflection). An index <= 1 disables the Fresnel effect and creates a pure mirror like reflection (independent of the viewing angle). If you want to get Fresnel back, you need to set index to a value > 1.

The new BRDF model also conserves energy which means that a specular value of 1 and an index of 0 will reflect all light arriving at the surface. That's why the reflection is stronger in the new version. -> You should probably reduce the brightness of the specular value.

The higher saturation in multiple reflections is probably caused by a too high saturation and strength/brightness of the specular colour, together with the fixed colour system, which in the past didn't cover the whole gamut available.

Cheers,
Marcus

[jitendra]

Thank you very much for your detailed reply.
I will focus on your suggestions.

Dear Octane Team,

I have been using 2.58d until now for production and its running great.
Today I tried RC1 and found that, the rendering results are entirely changed.
All my transparent materials and metal materials are all displaying different result even with the same settings as before.
Please checkout a metal example in the images attached.

With semi transparent material this changes are real drastic.

Please HELP!

Thank you.

There are a few things going on here and you have to look at the details:

"Saturate_to_white" is a tool to fix overexposed areas that would often end in one of the 6 combination of the RGB primaries (red, green, blue, cyan, yellow, magenta). If you set it to 1, those overexposed colours would become white instead. Your image is totally overexposed and setting "saturate_to_white" to one causes the grid of the floor to disappear.

-> When you set up materials you should make sure that the exposure is not too high.

The flatness of the metal is actually correct and caused by the introduction of an index parameter in the glossy material, which is by default 0. The "index" parameter represents the index of refraction of a coating layer which is often causing a Fresnel effect in the reflection (i.e. shallow viewing direction causing a higher reflection). An index <= 1 disables the Fresnel effect and creates and a pure mirror like reflection (independent of the viewing angle). If you want to get Fresnel back, you need to set index to a value > 1.

The new BRDF model also conserves energy loss which means that a specular value of 1 and an index of 0 will reflect all light arriving at the surface. That's why the reflection is stronger in the new version. -> You should probably reduce the brightness of the specular value.

The higher saturation in multiple reflections is probably caused by a too high saturation and strength/brightness of the specular colour, together with the fixed colour system, which in the past didn't cover the whole gamut available.

Cheers,
Marcus