Different Works in Progress

Post your tests, experiments and unfinished renderings here.
User avatar
FarbigeWelt
Donor
Donor
Posts: 836
Joined: Sun Jul 01, 2018 12:07 pm
Location: Switzerland
Contact:

Re: Different Works in Progress

Post by FarbigeWelt » Mon May 20, 2019 6:27 pm

For some reason laser do not well render in homogeneous volume and even worser in heterogeneous volume.
The only safe way to get a picture like that is BiDir at the moment.
Cascade 12, 13 BiDir Homogen.png
Cascade 12, 13 BiDir Homogen
160.8 | 42.8 (10.7) Gfp / Windows 10 Pro, intel i7 4770K@3.5, 32 GB | 2 AMD Radeon 5700 XT, 8 GB
17.3 | 19.0 ( 4.7) Gfp / macOS X 13.6, iMac 27'', 2010, intel i7 870@2.93, 24 GB | ATI Radeon HD 5750, 1 GB
#luxcorerender | Gfp = SFFT Gflops

Fox
Posts: 297
Joined: Sat Mar 31, 2018 11:17 am

Re: Different Works in Progress

Post by Fox » Mon May 20, 2019 8:53 pm

When i last played with lasers, maybe in the end on 2018, the BiDir was only one, that was able to render them.

User avatar
FarbigeWelt
Donor
Donor
Posts: 836
Joined: Sun Jul 01, 2018 12:07 pm
Location: Switzerland
Contact:

Geometrical Nonsense, openCL PGIC

Post by FarbigeWelt » Mon May 27, 2019 4:02 pm

A few first impression of my little short term project.
Street views and eagle sight of gem stone city's down town.
Best viewed open minded and mood relaxed.
Geometrical Nonsense a.jpg
Geometrical Nonsense b.jpg
Geometrical Nonsense c.jpg
Geometrical Nonsense d.jpg
Geometrical Nonsense e.jpg
I wonder how city's lime light looks like by night.


Info: openCL, PGIC, 1500 sample, approx. 19 minutes
160.8 | 42.8 (10.7) Gfp / Windows 10 Pro, intel i7 4770K@3.5, 32 GB | 2 AMD Radeon 5700 XT, 8 GB
17.3 | 19.0 ( 4.7) Gfp / macOS X 13.6, iMac 27'', 2010, intel i7 870@2.93, 24 GB | ATI Radeon HD 5750, 1 GB
#luxcorerender | Gfp = SFFT Gflops

Fox
Posts: 297
Joined: Sat Mar 31, 2018 11:17 am

Re: Different Works in Progress

Post by Fox » Tue May 28, 2019 8:39 am

I like the first 3.

User avatar
FarbigeWelt
Donor
Donor
Posts: 836
Joined: Sun Jul 01, 2018 12:07 pm
Location: Switzerland
Contact:

Re: Different Works in Progress

Post by FarbigeWelt » Tue May 28, 2019 4:03 pm

Fox wrote:
Tue May 28, 2019 8:39 am
I like the first 3.
Thank you for the feedback. They are also my favorites.
160.8 | 42.8 (10.7) Gfp / Windows 10 Pro, intel i7 4770K@3.5, 32 GB | 2 AMD Radeon 5700 XT, 8 GB
17.3 | 19.0 ( 4.7) Gfp / macOS X 13.6, iMac 27'', 2010, intel i7 870@2.93, 24 GB | ATI Radeon HD 5750, 1 GB
#luxcorerender | Gfp = SFFT Gflops

User avatar
FarbigeWelt
Donor
Donor
Posts: 836
Joined: Sun Jul 01, 2018 12:07 pm
Location: Switzerland
Contact:

Geometrical Nonsense, at night, first 'shots'

Post by FarbigeWelt » Tue May 28, 2019 6:14 pm

The glass material had to be changed because glass only lead to nothing usable.
The glass is mixed with matte (0.95 glass). Specular, transmission and matte have all the same color. I observed bright roofs but there is no light above the gem stone city. The lights emit their light vertical to the plane. I have no idea yet why the roofs are so bright.

Rainard is my favorite tone mapper. Nothing I would call straight forward usable. But hey! This Rainard algorithm rocks the colors!

openCL PGIC, OIDIN, 1500 samples, 17m28s
Geometrical Nonsense by night f.jpg
Geometrical Nonsense by night f
BiDir Metropolis, OIDIN, 1500 samples, 4h54m
Geometrical Nonsense by night f'.jpg
Geometrical Nonsense by night f'
(Still a lot to do until I am satisfied with the lights.)
160.8 | 42.8 (10.7) Gfp / Windows 10 Pro, intel i7 4770K@3.5, 32 GB | 2 AMD Radeon 5700 XT, 8 GB
17.3 | 19.0 ( 4.7) Gfp / macOS X 13.6, iMac 27'', 2010, intel i7 870@2.93, 24 GB | ATI Radeon HD 5750, 1 GB
#luxcorerender | Gfp = SFFT Gflops

User avatar
FarbigeWelt
Donor
Donor
Posts: 836
Joined: Sun Jul 01, 2018 12:07 pm
Location: Switzerland
Contact:

Gem Stone City, night street view

Post by FarbigeWelt » Wed May 29, 2019 6:15 pm

Background rendering, so time does not matter much. However, it seems that more lights require more time.
openCL PGIC 1500 samples 39m29s
Geometrical Nonsense by night g.jpg
Geometrical Nonsense by night g
BiDir 1500 samples 6h14m g'
Geometrical Nonsense by night g'.jpg
Geometrical Nonsense by night g'
BiDir 1500 samples 12h23m h' (OIDIN 84% mix)
Geometrical Nonsense by night h'.jpg
Geometrical Nonsense by night h'
PGIC 1500 samples 44m09s i
Geometrical Nonsense by night i.jpg
Geometrical Nonsense by night i
The images are far away from the point I want it. Meanwhile, I made a sketch of building what is closer to my mind view. Not sure how to apply this to the other model without working a dozen hours on the objects.
openCL 1000 samples 10m06s
Architecture-2.jpg
Architecture
160.8 | 42.8 (10.7) Gfp / Windows 10 Pro, intel i7 4770K@3.5, 32 GB | 2 AMD Radeon 5700 XT, 8 GB
17.3 | 19.0 ( 4.7) Gfp / macOS X 13.6, iMac 27'', 2010, intel i7 870@2.93, 24 GB | ATI Radeon HD 5750, 1 GB
#luxcorerender | Gfp = SFFT Gflops

User avatar
FarbigeWelt
Donor
Donor
Posts: 836
Joined: Sun Jul 01, 2018 12:07 pm
Location: Switzerland
Contact:

openCL PGIC Testing Parameters

Post by FarbigeWelt » Fri May 31, 2019 3:11 pm

First the BiDir result. I take is as reference.
Default body 2 BiDir 1200 samples 1h33m.jpg
Default body 2 BiDir 1200 samples 1h33m
Due to the bright floor the caustics are faint. One has to look good at them to notice the fine lines.
On the next picture you can see the effect of increasing the amount of photons. For this test series indirect and caustic have the same amount.
Influence of Photon amount on Caustics.jpg
Influence of Photon amount on Caustics
Naturally the higher the number the more time the render takes during session initialization.
With 20 million photons one gets an image that has a steady character although with every million photons more the difference between to renders keep even from 49 to 50 millions comparable to differences between less photons.

The next animated gifs show effect of other changed parameters.
Caustics-Look-up-Max-count.gif
Caustics-Look-up-Max-count
As you can see this parameter changes the 'noise pattern' a bit but only randomly.
(With a count above 800 the buffer size exceeded available GPU RAM of 3'422'552'064 bytes what lead to a warning before the render started (V2.2 beta 1, build 0a863a0)).
Caustics-Look-up-radius.gif
Caustics-Look-up-radius
The dots of the 'noise pattern' reduces notable by decreasing this radius.
Indirect-Cache-Look-up-radius.gif
Indirect-Cache-Look-up-radius
The animation shows that there is happening something but it is not clear to me what exactly.

Next pictures show openCL without PGIC.
openCL_Metropolis_0011.png
openCL_Metropolis_0011
openCL_Sobol_0001.png
openCL_Sobol_0001
Metropolis pattern on the body has obviously a lower resolution than Sobol pattern.

Last image has the parameters: 10 million photons, indirect cache radius 0.1, caustic cache radius 0.00001, max count 64.
2500 samples, 2m19s init, 8m24s render
Glass body, openCL PGIC 2500 samples.jpg
Glass body, openCL PGIC 2500 samples
Conclusion: In comparison with BiDir PGIC fails to render the fine caustics of this body.

EDIT: I found the reason for the missing caustics. The calculation of the dispersion is wrong for this object, it is too high.
160.8 | 42.8 (10.7) Gfp / Windows 10 Pro, intel i7 4770K@3.5, 32 GB | 2 AMD Radeon 5700 XT, 8 GB
17.3 | 19.0 ( 4.7) Gfp / macOS X 13.6, iMac 27'', 2010, intel i7 870@2.93, 24 GB | ATI Radeon HD 5750, 1 GB
#luxcorerender | Gfp = SFFT Gflops

User avatar
FarbigeWelt
Donor
Donor
Posts: 836
Joined: Sun Jul 01, 2018 12:07 pm
Location: Switzerland
Contact:

openCL PGIC without Dispersion

Post by FarbigeWelt » Sun Jun 02, 2019 9:20 am

Body BiDir 1500 samples.jpg
Body BiDir 1500 samples
Body openCL PGIC 1500 samples.jpg
Body openCL PGIC 1500 samples
This object is obviously not a good example for caustics with PGIC.
160.8 | 42.8 (10.7) Gfp / Windows 10 Pro, intel i7 4770K@3.5, 32 GB | 2 AMD Radeon 5700 XT, 8 GB
17.3 | 19.0 ( 4.7) Gfp / macOS X 13.6, iMac 27'', 2010, intel i7 870@2.93, 24 GB | ATI Radeon HD 5750, 1 GB
#luxcorerender | Gfp = SFFT Gflops

User avatar
FarbigeWelt
Donor
Donor
Posts: 836
Joined: Sun Jul 01, 2018 12:07 pm
Location: Switzerland
Contact:

openCL PGIC Caustics with Lenses. I like :-) !

Post by FarbigeWelt » Wed Jun 05, 2019 10:31 am

EDIT2: Thanks to Fox Node wrong and missing camera volume I was able to fix the BiDir image. Therefore I replaced it in this post to give the correct aspects.

First I start with the less spectacular BiDir image.
Approx. 1000 samples, depth 25, metropolis, denoised 95% (mix 95% OIDN 10% raw).
Lenses BiDir 1000 samples OIDN 0.95 3h55m CPU 6of8.jpg
Lenses BiDir 1000 samples OIDN 0.95 3h55m CPU 6of8
The next picture is based on the same file but rendered with openCL.
Lenses openCL PGIC 1000 samples OIDN 0.95 0h11m04s init 2m28s.jpg
Lenses openCL PGIC 1000 samples OIDN 0.95 0h11m04s init 2m28s
Okay, BiDir is more colorful, but look at the render times. openCL is so amazing fast.

Further images rendered on GPU ( all following images have an external air volume with 1.5 absorption)
openCL, at least 1500 samples, depth 24 total path and glossy the others 12, denoised 90%.
Lenses openCL PGIC 2500 samples r 0.001 OIDN 0.9.jpg
Lenses openCL PGIC 2500 samples r 0.001 OIDN 0.9
Lenses openCL PGIC 1500 samples r 0.01 OIDN 0.9.jpg
Lenses openCL PGIC 1500 samples r 0.01 OIDN 0.9
As one can see caustic lower look up radius leads to smoother caustics in the case of the scene.

Just for the memories, some pictures without denoiser.
Lenses openCL PGIC 5000 samples, r 0.0001.jpg
Lenses openCL PGIC 5000 samples, r 0.0001
Lenses openCL PGIC 5000 samples, r 0.01.jpg
Lenses openCL PGIC 5000 samples, r 0.01
Rendering time openCL approx. 25 to 60 minutes depending on samples (on 2 Hawaii GPU).

Conclusion: If an object leads to hard caustics like a 'burn glass' PGIC is able to do render a very good job. And in contradiction to BiDir it renders inter- and or intra-mirroring effects or what ever leads to the much more interesting images of the lenses. And what counts even more: PGIC is much faster than BiDir (at least on my microsaurier CPU).

(As once written a combination of openCL and BiDir would be great. I think this is what PGIC is for to a certain good amount.)

EDIT: By the way, openCL without PGIC looks like.
Lenses openCL 2500 samples.jpg
Lenses openCL 2500 samples
Last edited by FarbigeWelt on Thu Jun 06, 2019 4:03 pm, edited 1 time in total.
160.8 | 42.8 (10.7) Gfp / Windows 10 Pro, intel i7 4770K@3.5, 32 GB | 2 AMD Radeon 5700 XT, 8 GB
17.3 | 19.0 ( 4.7) Gfp / macOS X 13.6, iMac 27'', 2010, intel i7 870@2.93, 24 GB | ATI Radeon HD 5750, 1 GB
#luxcorerender | Gfp = SFFT Gflops

Post Reply