Impressive! The simulation and the rendering are flawless. I could even hear the sound of the waves, crashing on the rocks. Can you please tell me what are the specs of your rendering hardware, and the time it took you to simulate and render this video? Is the renderer Mantra? What were your target resolutions, for the container (simulation) and the camera (renderign) ? Did you add some extra compositing ? Sorry if I'm being too curious, I just really like this work!
Thanks a lot :) I used our renderfarm in our school, base fluid layer I tiled it in 2 and it's about 100 minute each for one tile with 32G RAM 32cores render nodes; for simulation, there are about 34M particles in base FLIP and about 3 min per frame, Xeon E5-2630 v2 @ 2.60 GHz cores. Rendered in Mantra, and particle separation 0.02. Not sure for camera if you means the camera region in particle surfacing or sampling in mantra node? I did some color correction but the compositing is only those layers.
Thank you for the answers. I think it's always a good thing to have an idea of the cost of a simulation, and that's something that is missing in most CGI breakdowns. Concerning the camera resolution, I was precisely talking about the output resolution parameter, in pixels.
Seeing renderings of this quality level is always for me a reminder of how powerful Houdini is, (in the good hands). Well done!
Thanks! The samplings for render varies a lot for volumes and particles, for particles only I have 5,5 and min 6 max 12, with volume sample volume shadow 0; for volume the overall sample are smaller but the volume sample, limits and stochastic sample higher (sometimes 6 - 8). Hope it helps :)
Thank you for the extra information. There is usually a lot of tweaking there before obtaining the desired look. I did forget to ask you that. I was actually refering to the final size of the frames that you used at rendering time (like for example 1920x1080 px). It may be obvious from the video size displayed on Vimeo, but I've seen some artists using different sizes when they need extra details. Also, thank you for being that fast at answering.
COMMENTS
Fractal_By_Nature 7 years, 6 months ago |
Impressive! The simulation and the rendering are flawless. I could even hear the sound of the waves, crashing on the rocks.
Can you please tell me what are the specs of your rendering hardware, and the time it took you to simulate and render this video? Is the renderer Mantra? What were your target resolutions, for the container (simulation) and the camera (renderign) ? Did you add some extra compositing ?
Sorry if I'm being too curious, I just really like this work!
lvwanqi 7 years, 6 months ago |
Thanks a lot :) I used our renderfarm in our school, base fluid layer I tiled it in 2 and it's about 100 minute each for one tile with 32G RAM 32cores render nodes; for simulation, there are about 34M particles in base FLIP and about 3 min per frame, Xeon E5-2630 v2 @ 2.60 GHz cores. Rendered in Mantra, and particle separation 0.02. Not sure for camera if you means the camera region in particle surfacing or sampling in mantra node? I did some color correction but the compositing is only those layers.
Fractal_By_Nature 7 years, 6 months ago |
Thank you for the answers. I think it's always a good thing to have an idea of the cost of a simulation, and that's something that is missing in most CGI breakdowns.
Concerning the camera resolution, I was precisely talking about the output resolution parameter, in pixels.
Seeing renderings of this quality level is always for me a reminder of how powerful Houdini is, (in the good hands).
Well done!
lvwanqi 7 years, 6 months ago |
Thanks! The samplings for render varies a lot for volumes and particles, for particles only I have 5,5 and min 6 max 12, with volume sample volume shadow 0; for volume the overall sample are smaller but the volume sample, limits and stochastic sample higher (sometimes 6 - 8). Hope it helps :)
Fractal_By_Nature 7 years, 6 months ago |
Thank you for the extra information. There is usually a lot of tweaking there before obtaining the desired look. I did forget to ask you that. I was actually refering to the final size of the frames that you used at rendering time (like for example 1920x1080 px). It may be obvious from the video size displayed on Vimeo, but I've seen some artists using different sizes when they need extra details. Also, thank you for being that fast at answering.
Please log in to leave a comment.