OTOY Forums

Hi,

i am getting strange artefacts (negative values) in rendering glass materials (Material_Specular), and i cant nail it down what causing the issue.
path tracing kernel, texture environment (polyhaven .hdr).
octane 2024.1.2.0, houdini 20.5.445

Please can you supply a sample?

It is generally from something that is "overbright" in this case probably some pixels or parts of the hdr too. Sometimes slightly blurring the hdr in PS helps.
Alternatively, you can reduce your GI clamp value way down.

hi ricky,

thx for the reply.
i also thought about the hdr as the cause of the issue, but switching the env to daylight, gives me similiar issues.
the GI is already clamped to 10 (instead of 1e-10 as the default) and setting this way down to 1 doesnt help...

by changing the material to a standard helps to get less to none visible artefacts, but they are there in the AOVs (reflection + refrection).

i can provide a sample file asap.

The "solutions" are provided on this page below.

Long story short, it's "expected", not unusual (and common with digital cameras as well). Easily manageable via AgX OCIO or Octane's own improved processing (or in the hands of a knowledgable digital imaging technician / engineer / colorist). Not ACES.

➜ OctaneRender® “Color Management”.
This example from the Caustics page from the site, features a similar scenario, specular refraction dispersion-enabled caustics.
No before/after available, but that's the result it would yield (particularly the "stars" looking "glare" on the left and right base of the shader ball):

From the same link above, an extreme stress-test example, this time before/after: