Wave Optics Kernel

[ValorCatStudios]

https://youtube.com/@mandyxia8428?si=jyLOK4jYcdkD4ANj

Would LOVE to someday soon see a "Wave Optics" kernel in Octane! Can't think of anything more scientifically accurate than the literal simulation of light waves (Interference, diffraction, and polarization)!

[Bulwerk]

+1

[nuno1980]

We don't need this new kernel but yes the path tracing kernel is supported by birefringence (polarization... included but future), diffraction (future but available on Cinema 4D)... materials.

[ValorCatStudios]

Polarization is only 1/3 component of the physical properties of lightwave interaction. Interference and diffraction also should have a tremendous effect on how light wraps around an object.

And I would argue scientific accuracy has been a major part of OTOY's DNA (I've watched every "state of otoy" video under the sun!)

L1: Direct lighting
L2: Spectral Brigade R&D
L3: Spectral Pathtracing
L4: Photon Tracing
L4: PMC (Metropolis Light Transport)
L5: ???? Wave Optics perhaps?

I love a diversity of options that literally no one else offers!!

[nuno1980]

...
L3: Spectral Pathtracing
L4: Photon Tracing
L4: PMC (Metropolis Light Transport)
L5: ???? Wave Optics perhaps?

I love a diversity of options that literally no one else offers!!

FALSE!

I correct your errors:
L3: PMC (for few various caustics missed - OR since v3.00 is caustics later than v2.25 at same S/px)
L4: Photon Tracing (for variable caustics (new) missed and/or glitched (learn - Variable caustics (new): Path Tracing vs Photon Tracing), if the glossy ground has roughness with less than 0.3 or 0.4)
L5: Path Tracing (spectral... included) (or Wave Optics? Dunno due to the lack of our test) but unfortunately we need very, very long number of samples per pixel with very long render time for very good caustics at our sun source.

L5 is highest realism of lighting level.

[nuno1980]

No error to correct as only the name of the Kernels were mentioned.

But... I think @ValorCatStudios will make mistakes.

That's the nature of path tracing and the case for all uni-directional Monte Carlo path tracers.

Well, bidirectional path tracing (PT) is quick caustics but... indirect caustics aren't quick. Bidir PT is as faster as uni-directional PT for indirect caustics.

[nuno1980]

That's prejudice and a bold assumption.

The discussion around caustics and PT is rather futile as Photon Tracing is available. It covers most end-user use cases in a relatively quite elegant and examplary way.

Ok but I continue to use PT kernel for highest realism of lighting for no any limitation as universal rays because PHT kernel and Wave Optics(?) can't render any light reflected and/or refracted at glossy and/or specular non-roughness material(s). PHT only(?) may not render any variable caustics, if the roughness of glossy ground has less than 0.3~0.4.

Can you understand this?

[nuno1980]

In any way, Octane caustics solutions — both in PT and Photon Tracing — are exemplary, efficient and versatile.
There little to nothing more needed at this point, besides the expected and typical constant improvements.

I've knew this but I changed the kernel from PHT to PT since 2022 because the brightness issue for caustics only isn't fixed at PHT yet for 1 year ago -> check my «"Allow caustics": wrong brightness on caustics at Photon» topic. And, at PHT kernel, the variable caustics may be glitched on *glossy* ground with the roughness of less than 0.3~0.4.

[james_conkle]

Does anyone have an in-render way to get similar wave effects that come from heat in the air? Similar to the OP reference.