PhotonGI cache

Discussion related to the LuxCore functionality, implementations and API.
User avatar
lacilaci
Donor
Donor
Posts: 1969
Joined: Fri May 04, 2018 5:16 am

Re: PhotonGI cache

Post by lacilaci »

Yes it's all great and even some bias is great. I'm personally pretty confident with the feature by now, but for new users it might be important in the future to automatize what can be automatized. Or some quickstart-guide to point people to when they ask would be also good.

Another small annoyance that comes to mind is "usage threshold scale" the name doesn't seem correct to me. Wouldn't it be more acurate as "usage distance scale" or "distance threshold scale"? Or is even scale a proper name? isn't this a multiplier?
User avatar
Sharlybg
Donor
Donor
Posts: 3104
Joined: Mon Dec 04, 2017 10:11 pm
Location: Ivory Coast

Re: PhotonGI cache

Post by Sharlybg »

Fair Fight without OIDN in the mix. PGI vs Path :


Path 1mn GTX 1080ti

Fair Fight Path.jpg

PGI 1mn GTX 1080ti

Fair Fight PGI.jpg
PGI Debug
Fair Fight PGI debug 10cm.jpg
Support LuxCoreRender project with salts and bounties

Portfolio : https://www.behance.net/DRAVIA
User avatar
Sharlybg
Donor
Donor
Posts: 3104
Joined: Mon Dec 04, 2017 10:11 pm
Location: Ivory Coast

Re: PhotonGI cache

Post by Sharlybg »

And PGI with Normal Angle of 90°:

Light is Back :D

PGI 1mn 1080ti + 90° normal angle
Fair Fight PGI + 90°.jpg

Wonder why is that and why normal angle default is so low if it can improve thing at 90°. is there any Lost :?:
Support LuxCoreRender project with salts and bounties

Portfolio : https://www.behance.net/DRAVIA
User avatar
lacilaci
Donor
Donor
Posts: 1969
Joined: Fri May 04, 2018 5:16 am

Re: PhotonGI cache

Post by lacilaci »

@Sharlybg try also using 10-15cm lookup and usage threshold of 1 and 90 degree normal angle. Should be much faster even.
User avatar
Dade
Developer
Developer
Posts: 5675
Joined: Mon Dec 04, 2017 8:36 pm
Location: Italy

Re: PhotonGI cache

Post by Dade »

It looks too dark to me without 90deg normals, I will investigate more the topic.
Support LuxCoreRender project with salts and bounties
User avatar
Sharlybg
Donor
Donor
Posts: 3104
Joined: Mon Dec 04, 2017 10:11 pm
Location: Ivory Coast

Re: PhotonGI cache

Post by Sharlybg »

lacilaci wrote: Tue Feb 26, 2019 8:17 pm @Sharlybg try also using 10-15cm lookup and usage threshold of 1 and 90 degree normal angle. Should be much faster even.
Done here is the result : Note i just increase a little bit the exposure in color management
Set.png
Set.png (31.91 KiB) Viewed 4585 times

Luxcore 2.2 1 + 15cm.jpg
Support LuxCoreRender project with salts and bounties

Portfolio : https://www.behance.net/DRAVIA
User avatar
lacilaci
Donor
Donor
Posts: 1969
Joined: Fri May 04, 2018 5:16 am

Re: PhotonGI cache

Post by lacilaci »

I think for max size a 1000 or 2000 should give same result but faster precomputation
User avatar
lacilaci
Donor
Donor
Posts: 1969
Joined: Fri May 04, 2018 5:16 am

Re: PhotonGI cache

Post by lacilaci »

I think all materials should build cache for diffuse rays.

Also cache size and photon count should be calculated to fill some criteria, maybe coverage in screen space or something, instead of guessing how much photons to cast and how big cache should be

lookup size should be 15 cm hardcoded
usage threshold should be 1 hardcoded
glossiness of 0.1 should be hardcoded
normal angle should be 90 or some other fix for dark rendering should be presented as default/hardcoded value

Basically what I'm suggesting is, fill screen with cache entries of 15cm radius and use 1 for usage threshold scale by default. For glossy rays use 0.1 threshold and for diffuse use 0.0

The thing that needs to be discovered here is of course, how to automatically determine when you have enough photons.
Fox
Posts: 437
Joined: Sat Mar 31, 2018 11:17 am

Re: PhotonGI cache

Post by Fox »

Why to hardcode lookup size at 15cm?
I use 0.5 cm for small scale scene, the total scene has the size of 20cm.
User avatar
Dade
Developer
Developer
Posts: 5675
Joined: Mon Dec 04, 2017 8:36 pm
Location: Italy

Re: PhotonGI cache

Post by Dade »

lacilaci wrote: Tue Feb 26, 2019 8:51 pm Also cache size and photon count should be calculated to fill some criteria, maybe coverage in screen space or something, instead of guessing how much photons to cast and how big cache should be
Screen coverage would be a good criteria, the problem is in having a constant radius: one pixel covers 15cm at 1m of distance and 150m at 1km (random numbers, just to give you an idea of where the problem is).
Support LuxCoreRender project with salts and bounties
Post Reply