Indigo Renderer Forum

Houdini Terrain generation.
Found lately some fine tunning.
9x9 km. 64 million quads.
More memory is needed for a first plane!!!

Pure colors.

Wow, cool.

yonosoy wrote:Houdini Terrain generation.
Found lately some fine tunning.
9x9 km. 64 million quads.
More memory is needed for a first plane!!!

Nice!
I did try it also some days ago. ^ . ^

Thank you!

38 h for 3800 spp. Still very dirt ...

A revision from the serie A.
18 million instances. 3h of rendering for 450 samples pixel.
Just playing ... Seems like a little mockup.

145 millions of polygons in a very big patch.
Not finished, loose power.
The curvature of Earth, ...

I expect a more consistent serie, called "Situations".

Startpoint!

^U^ YES!
Very nice looking...


Would you care to elaborate, describe some more about the process? /in short will be just fine

Is it all procedural: terrain, clouds, ocean... from houdini?
How, since no plugin is available?
Can't (or can you) import .abcs & .vdbs, so most likely another app is used - which is the mediator?

TIA & stay strong

Hi burnin. Very glad by your interest.
Give me some weeks to report a detailed post, actually I´m refining some questions.

So many thanks for your words, Will try to stay stronger. Things go better lately.

Edit: wip of the next two; bidirectional, mie cumulus. Fits with supersample 8.

I need 50-80 hours to clean each 5K image... or maybe more...

Working in this one...
A remastered of a old one try.
From a 512x512 pixels map, all procedural. Mont Ruapehu, in New Zealand.

(burnin, sorry. I´m not able to work vdb´s inside cinema but abc´s YES. Just fusionate them, in the same way like a cinema 4d scene, the option is not in the import panel. The clouds are code from Ctzn (adapted to scale) and for the oceans, the only option with 32 GB of ram is a bump (also I can´t work with curved oceans think, almost at the moment...).

Situations_002.
Not finished, only 2400 samples per pixel in 20 h.
Want another isolation with better shading for print. Edited: denoised and downsized. Another promotion... training...
Study of Mont Ruapehu and Mont Ngauruhoe with pure colors. Edited_2: Final run. Situations_003

After five years of test, feels that it´s right and fits.
Five non perfect spheres with increased precedence !!!

A wip of today (resized from HD to 25% after two hours of process).
Remember somehow Rothko´s compositions.

Will post this in a higher resolution soon.
Thanks in advance.

A-MA-ZING!

Thank you very much, thesquirrell!

Bidirectional path tracing.
Real tabulated spectrum for absorption and SSS in the ocean, Cauchy on.
About 60 h in the laptop (an i3 of two cores and 4 threads at 2.1 Ghz), 3400 spp.
Chroma noise filtered.

This system only works in v3.8. Five non perfect concentric spheres:

AtmoSphere, precedence 1.
CloudSphere, precedence 2.
AtmoSphere, precedence 3 (same definition of the material, 2 km below).
OceanSphere, precedence 4.
EarthSphere (2 km underwater).

The difference with mie cumulus noise is evident.

Hope this system will work some day in v4 (with clamp contributions will be perfect).
This kind of challenge is exciting to me.

Custon atmospheres work with 5 spheres (4 perfect and the ocean non perfect).
Metropolis bidir.