Show/Hide Toolbars

Clarisse 4.0 User Guide

To denoise final renders, Clarisse provides a special image filter called Optix Denoiser. Please remember that the Optix Denoiser runs exclusively on NVIDIA GPUs. If you plan to render your images over a render-farm without NVIDIA GPUs, the image filter is going to fail. We recommend you instead to disable this filter and use CDenoise command line tool on the rendered EXR image sequence.



It is highly recommended to render images with an anti-aliasing filter set Box with a filter size of 1.0 for optimal results.


To add the image filter just select the 3D Layer you wish to denoise and using the Attribute Editor under Filters attribute, choose Add > Optix Denoiser. The filter comes with several options to control the behavior and quality of the denoiser:





Control the strength of the denoiser output. It basically performs an over operation from the denoised output over the input image. When set to 100% you get the full output from the denoise image. When set to 0% you get the original input image.

Use Albedo

Use the albedo AOV to guide the denoiser to improve the results. Note that sometimes this can introduce some artifacts leading to an image degradation.

Use World Normal

Use the world_normal AOV to guide even further the denoiser. Note that you can only enable normals if the Raw Color is enabled.

Albedo AOV

Choose the name of the image buffer you wish to use as albedo

World Normal AOV

Choose the name of the image buffer you wish to use as world normal


By default, the albedo and the world_normal AOVs are not automatically added to the image. You must output albedo and world_normal from your renders using the AOV Editor. These extra AOVs generally help the denoiser to improve the quality of the denoised output. However, in some cases, these AOVs can degrade the quality of the denoised output so uncheck them to see if the result is improved. We recommend to first uncheck Use World Normals and then Use Raw Color.


Raw Output Render

Raw Output Render


Denoised Output

Denoised Output


Copyright (C) 2009-2019 Isotropix. All rights reserved.