Page 11 of 17

Re: Possible in Virtuality only

Posted: Fri Aug 24, 2018 11:18 pm
by FarbigeWelt
Hi kintuX, thank you much for your story.

Sure, for other eyes you can pass through or below a rainbow, but you cannot see it with your eyes and you cannot look at the back side of a rainbow, could you?

Is HDR the key to hyperspeed preview rendering?

Posted: Fri Aug 24, 2018 11:24 pm
by FarbigeWelt
Well, just load the scene and look at the Samples/s in view port render with CPU or even better GPU.
Rotating Cube.zip
Rotating Cube
(1.3 MiB) Downloaded 206 times

Re: Test scences GPU Path vs CPU BiDir

Posted: Fri Aug 24, 2018 11:30 pm
by kintuX
FarbigeWelt wrote: Fri Aug 24, 2018 11:18 pm Hi kintuX, thank you much for your story.

Sure, for other eyes you can pass through or below a rainbow, but you cannot see it with your eyes and you cannot look at the back side of a rainbow, could you?
Use a drone. Also, mirrors, prisms can do miracles... you can look at the back of your head even. :lol:
I know, i'm just messing around... i have a dislike for the philosophy & principle stating: "If i can't see it, then it doesn't exist."

Keep it up ;)

Re: Test scences GPU Path vs CPU BiDir

Posted: Sat Aug 25, 2018 9:46 pm
by FarbigeWelt
kintuX wrote: Fri Aug 24, 2018 11:30 pm
Use a drone. Also, mirrors, prisms can do miracles... you can look at the back of your head even. :lol:
I know, i'm just messing around... i have a dislike for the philosophy & principle stating: "If i can't see it, then it doesn't exist."
Well, in this case thanks for opening my eyes. :D Obviuos, you're good at finding solutions. But drone or camera aren't your eyes and you'll need to place quite a bit of mirror in the right position and a good optical device to see yourself below the rainbow or the back side of it. It sounds like a very sophisticated solution, not sure if light's intensity and shimmering air allow a handy solution. ;)

Actually, just a few words may place you below and around of one of the brightest, most colorful and clear rainbows with tones from dark, deep red over jumping yellow to vivid green into heavenly blue and warmest violet you've ever seen. :D

Glossy transluscent 1. Study

Posted: Sun Aug 26, 2018 12:19 am
by FarbigeWelt
Glossy translucent is a rather complex material. Let's start with the first tricky settings.
Glossy TL 1
Glossy TL 1
Glossy TL 2
Glossy TL 2
Glossy TL 3
Glossy TL 3
Glossy TL 4
Glossy TL 4
Glossy TL World
Glossy TL World
Glossy TL HDRi
Glossy TL HDRi
Glossy TL Area Light
Glossy TL Area Light

GPU Path and CPU BiDir together

Posted: Sun Aug 26, 2018 6:05 pm
by FarbigeWelt
GPU Path in view port and CPU BiDir can run together.
GPU Path in view port and CPU BiDir together
GPU Path in view port and CPU BiDir together
Spot the differences of both render modes in the glasses' shadows.

Re: Test scences GPU Path vs CPU BiDir

Posted: Sun Aug 26, 2018 6:39 pm
by B.Y.O.B.
If you use an intersectible light source (e.g. an area light), Path engines will also be able to render the caustics (with much worse performance of course, compared to Bidir).

Re: Test scences GPU Path vs CPU BiDir

Posted: Sun Aug 26, 2018 7:16 pm
by wasd
FarbigeWelt wrote: Sun Aug 26, 2018 12:19 am Glossy translucent is a rather complex material.
It's not complex. It just isn't translucent. The only tricky part in it is how with rather thin volume and absolutely smooth surface it scatters light so evenly.
Here be the glossy translucent material transmitting 97% light on the left and glass on the right.
Both has heterogeneous volume with 0.001 scattering scale.

Study: HDRI, Sun, Dispersion, Noise

Posted: Sun Aug 26, 2018 10:39 pm
by FarbigeWelt
On Saturday I observed a rather amazing render speed using HDRI. The scene was a very simple cube. Today, I rendered a bit more complex scene and compared render speed of different HDR images and light sources. The result was it does not matter much for speed what light I used.

Just a few impressions of my study. Render speed range was 750 to 900 kS/s. Sobol is slower than random, result was about the same.
Disp Flint Glasses and RGB Glossy 300 Samples not clamped
Disp Flint Glasses and RGB Glossy 300 Samples not clamped
Disp Flint Glass and RGB Glossy clamp 46313
Disp Flint Glass and RGB Glossy clamp 46313
I've tested different clamp values between not clamped an clamped with 46313. If the number is to small caustics disappear and the dispersive glass gets reddish.
Disp Flint Glasses and RGB Glossy HDRI 300 Samples 781 kSps clamped 1939100
Disp Flint Glasses and RGB Glossy HDRI 300 Samples 781 kSps clamped 1939100
The used HDR image is black with a very bright spot. This spot acts as light source. The pictures shows caustics too. Difference to sun is the color tone and the spot on the glasses.

Because all pictures with clear glass and caustics are noisy even with 300 samples I tried the denoiser again.
Probably I use it the wrong way by taking the picture from the denoiser channel.
Disp Flint Glasses and RGB Glossy 75 Samples 1939100
Disp Flint Glasses and RGB Glossy 75 Samples 1939100
Disp Flint Glasses and RGB Glossy 75 Samples denoised channel search 8 scales 4
Disp Flint Glasses and RGB Glossy 75 Samples denoised channel search 8 scales 4
Disp Flint Glasses and RGB Glossy 300 Samples denoised channel size 8 4
Disp Flint Glasses and RGB Glossy 300 Samples denoised channel size 8 4
Obviously the denoiser does not like the black spots on the walls.
Denoiser works well for cleaning noise on glass.

Re: Test scences GPU Path vs CPU BiDir

Posted: Sun Aug 26, 2018 10:43 pm
by FarbigeWelt
wasd wrote: Sun Aug 26, 2018 7:16 pm Here be the glossy translucent material transmitting 97% light on the left and glass on the right.
Both has heterogeneous volume with 0.001 scattering scale.
Thank you for the illustration.