Dark side of the moon

Our forum gallery for displaying your finished renders.
Post Reply
oldladyfear
Posts: 5
Joined: Mon Apr 16, 2018 12:57 am

Dark side of the moon

Post by oldladyfear » Mon Apr 16, 2018 1:09 am

Long time lux user & lurker, decided that the new release was a good time to start getting more active in the community. So here's my attempt to recreate a classic image!

I'm amazed at how much faster the engine is - only took 20 hours to render to 5% convergence using an underclocked fx-8350 on bidir default settings, using a multiple scattering volume! One issue i've noticed though is that at high dispersion values, the RGB interpolation used to estimate dispersion seems a little off - there's far less orange and yellow compared to what a similar scene would get under the old spectral dispersion model, and it looks more like distinct r/g/b bands. I might play around with tweaking the values on my fork to see if i can get it looking a bit more accurate. Overall though i'm extremely happy with the luxcore reboot, and i'm looking forward to putting together more scenes to show off!
darkside.png

User avatar
SATtva
Supporting Users
Posts: 29
Joined: Sun Dec 17, 2017 10:27 am
Contact:

Re: Dark side of the moon

Post by SATtva » Mon Apr 16, 2018 8:06 am

oldladyfear wrote:
Mon Apr 16, 2018 1:09 am
I'm amazed at how much faster the engine is - only took 20 hours to render to 5% convergence using an underclocked fx-8350 on bidir default settings, using a multiple scattering volume!
You could've cut the rendering time by the factor of 2 or 3 (at least) by using single scattering -- without any impact to the actual image. Multiple scattering is just an overkill for this setup (it's only needed for "thick" volumetrics like liquids or smoke).

oldladyfear
Posts: 5
Joined: Mon Apr 16, 2018 12:57 am

Re: Dark side of the moon

Post by oldladyfear » Mon Apr 16, 2018 10:44 am

i know, it was overkill for the scene - the scene started out as a worst-case scenario benchmark. A re-render using single scattering only took around 4.5 hours to converge.

Post Reply