-
Notifications
You must be signed in to change notification settings - Fork 40
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
How much frame rate can be achieved in blender #3
Comments
Hi, let me provide some details to answer this question. "Evaluating" a test image in Blender simply boils down to rendering the frame in question. In contrast, NeRF, or in this case Instant NPG, first trains a model to learn the 3D scene, and is then capable of predicting an image, given a camera. Depending on your hardware, the rendering parameters and the scene complexity (geometry and materials), Blender can take minutes to hours to render a single frame, especially if using Blender's path-tracer, Cycles. When rendering small scenes (simple geometry and materials) with Blender's real-time rasteriser, Eevee, Blender usually performs quite fast. But even with Eevee, rendering a single frame can still take from seconds to minutes. Instant NGP on the other side, will take from a few seconds to a very few minutes to train its NeRF model, and is then capable of predicting a test image way faster than almost any renderer. Therefore, when testing on hundreds of images, Instant NGP will be drastically faster than Blender. |
Thanks your answer. |
!!! sounds great! I will try it |
In blender, will the test efficiency be higher than ngp?
The text was updated successfully, but these errors were encountered: