parallax mapping Fstorm

General computer graphic news and papers.
lighting_freak
Posts: 234
Joined: Thu Jan 18, 2018 6:02 pm

Re: parallax mapping Fstorm

Post by lighting_freak »

Hello,
lacilaci wrote: Thu Jun 25, 2020 10:50 am Don't get me wrong.. I don't care about bidir
BiDir is still the only engine that seems to be able to manage complicated light paths...
Bidir isn't often used so it is not a lost
I guess is often used by people that care about the physics based realism.
OS - Windows 7 X64
CPU - Intel CORE i7
GPU1 - Variants of notebook card from nVidia
GPU2 - Variants of notebook onboard card from Intel
Lux - Latest possible relaease
User avatar
lacilaci
Donor
Donor
Posts: 1969
Joined: Fri May 04, 2018 5:16 am

Re: parallax mapping Fstorm

Post by lacilaci »

lighting_freak wrote: Wed Jul 01, 2020 5:13 am Hello,
lacilaci wrote: Thu Jun 25, 2020 10:50 am Don't get me wrong.. I don't care about bidir
BiDir is still the only engine that seems to be able to manage complicated light paths...
This is just plain wrong. In fact it is exactly opposite since pathtracing+lighttracing+caustic cache can aslo do sds caustics(underwater, reflected caustics etc) which bidir can't. It also makes other features that would make the renderer more flexible very complicated to add

Here's an example what I wouldn't be able to render with bidir: download/file.php?id=6109&mode=view
lighting_freak
Posts: 234
Joined: Thu Jan 18, 2018 6:02 pm

Re: parallax mapping Fstorm

Post by lighting_freak »

Hi,

with complicated light paths I meant something like that:
Lichtleiter_neu.jpg
I've added the blend-file to that post.
If you'd like you could try to find a solution to reach this result without biDir.
It would be really helpful for me to find something to speed those kind of visualisations up.

Thanks in advance.

BR
Attachments
LL2.zip
(1.83 MiB) Downloaded 307 times
OS - Windows 7 X64
CPU - Intel CORE i7
GPU1 - Variants of notebook card from nVidia
GPU2 - Variants of notebook onboard card from Intel
Lux - Latest possible relaease
User avatar
lacilaci
Donor
Donor
Posts: 1969
Joined: Fri May 04, 2018 5:16 am

Re: parallax mapping Fstorm

Post by lacilaci »

lighting_freak wrote: Thu Jul 02, 2020 11:11 am Hi,

with complicated light paths I meant something like that:

Lichtleiter_neu.jpg

I've added the blend-file to that post.
If you'd like you could try to find a solution to reach this result without biDir.
It would be really helpful for me to find something to speed those kind of visualisations up.

Thanks in advance.

BR
Not sure what's the problem, just switch to gpu and enable lighttracing.
This way I went from about 163K rays/s to 5M rays/s in your scene.
lighting_freak
Posts: 234
Joined: Thu Jan 18, 2018 6:02 pm

Re: parallax mapping Fstorm

Post by lighting_freak »

Hi,
lacilaci wrote: Thu Jul 02, 2020 12:22 pm Not sure what's the problem, just switch to gpu and enable lighttracing.
From now on you're my personal hero. A test scene runs much quicker than before and reaches the good result in a time that I could get little animations with a single weekend of rendering.

Thank you very much.

BR
OS - Windows 7 X64
CPU - Intel CORE i7
GPU1 - Variants of notebook card from nVidia
GPU2 - Variants of notebook onboard card from Intel
Lux - Latest possible relaease
User avatar
lacilaci
Donor
Donor
Posts: 1969
Joined: Fri May 04, 2018 5:16 am

Re: parallax mapping Fstorm

Post by lacilaci »

lighting_freak wrote: Mon Jul 06, 2020 2:08 pm Hi,
lacilaci wrote: Thu Jul 02, 2020 12:22 pm Not sure what's the problem, just switch to gpu and enable lighttracing.
From now on you're my personal hero. A test scene runs much quicker than before and reaches the good result in a time that I could get little animations with a single weekend of rendering.

Thank you very much.

BR
I'm not sure if there isn't an unresolved problem with animations and lighttracing.
viewtopic.php?f=4&t=2138&hilit=flickeri ... g&start=40

If I understood the problem correctly, there might be a problem with some flickering because lighttracing might reach different sampling/noise level with the same frame than pathtracing and in some frames it could do problems. I guess this is an issue that still needs to be addressed.
Post Reply