banding issue in reflections

[garytyler]

I've had this issue a few times and this isn't the best example but it is the most recent. In the bottom right and left corners of the frame, you'll see banding in the light reflections. I've tried various values of the Ray Epsilon setting and none of them help. The lighting was done with the also included HDR file created with HDR Studio. The reflections are from the very diffuse lights you see in the file. Also, the banding exists just the same in TIFF of PSD exports so I don't believe it is a compression issue.

Anyone have another idea where this comes from? I've gotten it with other HDRs and with regular Octane lights. Maybe it isn't an Octane issue but a C4D issue?

Thanks!

[aoktar]

try float buffer and a 32bit depth file format

[garytyler]

Is that the same thing as an object buffer?

Everything above was rendered in 32 bit. I also tried 16bit but got the same results.

[atome451]

Here in pink: Float(tonemapped) or Float Buffer

Render Buffer Type

[garytyler]

Thanks! Know where I can find some documentation on this setting? I have read about it before but can't seem to find it. Not sure why.

The names of settings seem to vary from version to version, the "help" interface doesn't have a search function and the 2 different manual links I have seem outdated.

[aoktar]

Hi,
it's a live product which is still changing. Updating the docs are big job without fixing the features. But i think that most parameters are self explanatory, don't suppose there is big secrets. Also you can ask here what do you want to know.

[garytyler]

Hi,
it's a live product which is still changing. Updating the docs are big job without fixing the features. But i think that most parameters are self explanatory, don't suppose there is big secrets. Also you can ask here what do you want to know.

Can you provide a brief breakdown of scenarios that would benefit from each of the 3 buffer settings? I was using the 8 bit buffer when I was getting banding, just now tried Float (tonemapped) and it rectified the issue! Thank you for that.

[aoktar]

First option is 8bit buffer. It's same as in Live Viewer. Because live viewer uses also 8bit buffer due performance. I think this is very clear.

Second is float(tonemapped). It has 32bit pixel depth for each color channel as you know. Also reflects the tone mapping of camera imager. And output values are limited between 0-1. values than bigger 1.0, will be clamped to 1.0.

Last is float buffer. This is raw colors of renderer. Colors will not be clamped in this mode. So you can have values which are bigger than 1.0. Also gamma is 1.0 for colors. Even there is no any camera imager effect. Best saving options for float buffer are 32bit tiff, exr or psd.

On you render, that was the problem. 8 bit depth is not enough to fill the semitones of close colors. I hope all is clear. Don't forget that practice is your friend.

[garytyler]

Thanks for the information. Sounds like tonemapped is the way to go. I think I used to do that and just forgot to check that setting. It seems like tonemapped would be a better default setting than 8 bit.