Performance & large data

BrainDeer is built for large neuroimaging datasets. Three mechanisms keep the UI fluid:

• Volume proxy: large volumes are fully loaded into memory only when needed (e.g. the visible slice).
• Surface LOD: surfaces render at different levels of detail depending on zoom/camera.
• Tract handling: large .trk files load efficiently; for very large bundles, check the load timings.

Media for this page

• [ ] Before/after comparison: load time with and without proxy (numbers from scripts/load_timing_howto.md)
• [ ] Screenshot: large dataset running smoothly in 3D

Practical tips

• Use .nii.gz instead of .nii to reduce disk load time.
• Many overlays at once? Deactivate invisible layers instead of deleting them.

See also

• Data model
• Troubleshooting