Render gaussian splats relative to the camera independent of (large) global coordinates#356
Open
pmi-gbs wants to merge 1 commit into
Open
Render gaussian splats relative to the camera independent of (large) global coordinates#356pmi-gbs wants to merge 1 commit into
pmi-gbs wants to merge 1 commit into
Conversation
pmi-gbs
force-pushed
the
relative-spark-renderer
branch
from
c23fb8e to
e541fe3
Compare
|
If this works, it will solve a lot of issues |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
I've had issues with big global coordinates (which happens e.g. when using WGS84 cartesian CRS). This happened within the discussion in WilliamLiu-1997/3D-Tiles-RendererJS-3DGS-Plugin#2 where even parenting the Renderer to the camera didn't help.
I think we can generally fix these floating point quantization issues by having the renderer really relative to the camera automatically (so that all the math is done in f64, instead of what I suspect f32 in the shader currently, but I haven't carefully checked this TBH).
So this should get rid of all these issues (including within the old renderer I think), so that the user doesn't have to care anymore about these float quantization artifacts.
As I'm maintaining Open source projects myself I want to disclaim this was heavily assisted by AI to be fair, I tried my best to understand the changes, and avoid slop. But well I'm not (yet) deep into the codebase, and I think the changes are compact enough to be worth putting into a PR. I did instruct that it should avoid allocations, so this resulted in a few drive-by fixes, which I think are fine (in case you're wondering).