Render subdiv settings

Started by digitalguru, May 19, 2025, 08:30:39 AM

Previous topic - Next topic

digitalguru

For my animation I've been testing still frames prior to rendering a whole sequence (will be testing a segment overnight) but just wondering about some of the render subdiv settings, particularly remembering the animation checker feature that used to change some of those settings.

Reading the docs I see a few salient settings - from the docs:
QuoteDetail jittering: For still images you should keep this turned ON. However, there is a mistake in the renderer that causes this to change from one frame to the next in an animation, so for animations you should switch it OFF. The Animation Check button (on the render node's Advanced tab) will check this setting and offer to turn it OFF.
QuoteJitter shading points: Jitter shading points chooses a random point on each micropolygon as the point where lighting and shaders are calculated. The results of those calculations are used to colour the whole micropolygon. Jittering provides a more natural image, but because of a mistake in current versions the jitter is different on each frame so it is a source of unwanted noise in animations. The Animation Check button (on the render node's Advanced tab) will check this setting and offer to turn it OFF.

I'm highlighting these as the help mentions "because of a mistake in current versions the jitter is different on each frame" etc and I wonder how relevant this is today? Don't know if the those doc are up to date with recent versions of TG. I'm currently on 4.6.31.

The last time a rendered an animation was 4(!) years ago and it looked good with these settings:
fully_adaptive = "1"
microvertex_jittering = "1"
force_all_edges = "0"
detail_jittering = "0"
detail_blending = "0.5"
displacement_filter = "1"
jitter_shading_points = "0"
ray_detail_multiplier = "1"
stabilise_ray_detail_in_motion = "1"
ray_detail_stabilisation = "3"
override_size_of_subdiv_cache = "0"

I'll do a test with default setting vs the last render settings used.
Just wondering if these params are still relevant for the newer version of Terragen (4.6.31) vs the older version 4.4.18.1, probably yes, but still wanted to check (as I may farm this render out).
f.y.i - render test segment will be where the camera pans the scene at a medium speed.

digitalguru

#1
I rendered a short section last night - results were interesting

no popping or artefacts in the non animation setting render - the only difference I see is in the shadows. The animation settings had shadows with a softer falloff, whereas the non animation had a harder terminator. Looks better to me and in comparison the non animation version shadows now look a bit "blobby"

non-animation setting:
fully_adaptive = "1"
microvertex_jittering = "1"
force_all_edges = "0"
detail_jittering = "1"
detail_blending = "0"
displacement_filter = "1"
jitter_shading_points = "1"
ray_detail_multiplier = "1"
stabilise_ray_detail_in_motion = "0"
ray_detail_stabilisation = "3"
override_size_of_subdiv_cache = "0"

animation settings:
fully_adaptive = "1"
microvertex_jittering = "1"
force_all_edges = "0"
detail_jittering = "0"
detail_blending = "0.5"
displacement_filter = "1"
jitter_shading_points = "0"
ray_detail_multiplier = "1"
stabilise_ray_detail_in_motion = "1"
ray_detail_stabilisation = "3"
override_size_of_subdiv_cache = "0"

non animation settings render:
non_animation_settings.jpg

animation settings render:
animation_settings.jpg

Dune

Interesting. I like the animations render better, so maybe for stills those settings would be good to try. Is there a difference in render time?

digitalguru

Agree, but on closer examination, the fake stones lose shadowing, which is a shame as I like what it was doing on the larger features, though I suspect it is technically not correct.
The animation feature that changed the shadows was Stabilse ray detail in motion. I'll experiment with that a bit more, but I suspect I might have to go back to the previous "blobbier" version :)

Render times were pretty much the same for both renders.