Revisiting portals and glass

[treddie]

Was just curious...I remember a thread stating that portals with glass was a feature not "yet" implemented. I assume this means that there are plans to include it soon, or no?

And concerning the old thread that shows how portals are designed to work, the discussion placed high importance on the fact that a reverse ray from the camera goes out into the scene looking for a light source. If it cannot find one after a reasonable amount of searching, it aborts that ray and starts with a new one. My question is that since an HDR "dome" is essentially a huge 360 degree light source, does a portal have a much easier time finding a "light source"? After all, the light source is basically everywhere.

[treddie]

Anyone?

[gabrielefx]

for me Octane light portals are useless.
I always use transparent lights behind every window aperture. I can tweak the power and color.

We must think like real photographers. Think what is done when a movie is shooted in a studio. You see discharge, quartz, halogen lamps everywhere...on the back of the windows and camera, on the ceiling, etc.

[treddie]

By "transparent" do you mean an emitter with opacity = 0?

[gabrielefx]

By "transparent" do you mean an emitter with opacity = 0?

yes

[treddie]

That's a great idea.

I have a special case though where it wouldn't work...I'm still playing around with virtual cameras with multiple lens elements, and beyond maybe 12 or so elements, Octane just can't push enough light through them all from the environment. That's critical because it's the rays coming from the environment that I need to project onto the image plane.

Beyond 12 or so elements, Octane can never get beyond a very noisy image, even incomprehensible, without reaching time-to-completion first.

[gabrielefx]

That's a great idea.

I have a special case though where it wouldn't work...I'm still playing around with virtual cameras with multiple lens elements, and beyond maybe 12 or so elements, Octane just can't push enough light through them all from the environment. That's critical because it's the rays coming from the environment that I need to project onto the image plane.

Beyond 12 or so elements, Octane can never get beyond a very noisy image, even incomprehensible, without reaching time-to-completion first.

I don't think that the light can pass through 12 layers of glass, you need special softwares for physic experiments.

[treddie]

Well, a group of 8 lenses works great, but 22 was impossible except for a couple of spots showing up on the image plane. 19 was slightly better and 12 actually worked out fairly reasonably. Problem with the last one is that I couldn't get it to focus...It was like looking through Hubble after it first went up. So I calculated some rays through it and found it couldn't focus with the data in the patent. So it looks like there were one or more typos in there.

There is a second issue as far as Octane is concerned. Some rays will just pass right through the lens groups and hit the image plane directly. Others, of course will bounce around and reflect off of all the lens surfaces to varying degrees before they reach the image plane. I found that I had to optimize the "lifetime" of bouncing rays in the 8-element lens to get as many of the reflections as possible. In the 12-element lens, I'm not sure much if any of those bouncing rays will make it all the way in.