maxwildersmith/splataverse-ortho

Splataverse Berlin

Berlin, Germany. The L0 manifest lists 22,045,669,352 Gaussians, 719 source survey tiles and 139 full-resolution storage chunks. All 139 listed chunk files are present and nonempty (211.86 GB); the manifest's chunk and source-L0 counts agree. This is a legacy native-density release, not a newly rerun balanced conversion.

Approximate source-tile footprint: 93.9 km², calculated as the union of source-tile horizontal bounding rectangles in EPSG:25833. The much larger 720.2 km² outer scene envelope (25.95 × 27.75 km) includes gaps. Neither value measures occupied mesh surface or municipal coverage. Coarse L1/L2 payloads have been removed; L0 is retained. This metadata audit checked manifest consistency and file availability, without independently decoding all payload headers.

Area provenance · City metadata · All real-city datasets

Real textured photogrammetry meshes converted analytically to Gaussian splats, with the original native-density L0 release with all 139 full-resolution chunks retained.

Brandenburger Tor from Pariser Platz (street level)	Reichstag from Platz der Republik (street level)
Berliner Dom from the Lustgarten (street level)	Fernsehturm from Alexanderplatz (street level)
Olympiastadion and Westend	719 source tiles listed by the manifest, top-down, north up

All renders are of the splats themselves (no mesh), drawn with an out-of-core LOD renderer: per view it frustum-culls tiles, picks each tile's level by distance (full resolution within 250 m of a street-level camera), and alpha-composites far-to-near passes when a view exceeds one GPU's budget (the street views are 120–200M Gaussians each, on a 24 GB GPU).

Source: the textured Berlin mesh tiles from TUM's MovingDrone dataset.

Releases

folder	area	Gaussians	files
berlin/	Mitte (Berliner Dom, Museumsinsel, Fernsehturm), 1.46 × 1.46 km, 17 tiles	517M	4 chunks
berlin_city/	719 source tiles; ~93.9 km² tile-bounds union	22.0B	139 full-resolution chunks (211.86 GB); coarse LODs removed

berlin/ layout

berlin/
  scene.usda        # one Xform per chunk with a `splat:file` asset path + `splat:gaussianCount`
  manifest.json     # CRS, origin, frame, per-chunk counts and bounds, source tiles
  chunks/chunk_XXX.spz

Chunks. Splats from all source tiles are merged and then re-split. Each split cuts a block at the median Gaussian count along its longer horizontal axis, and splitting repeats until every chunk is at most 10 GB of uncompressed data. At 56 B per Gaussian (14 float32 values: position, scale, rotation, opacity and degree-0 colour) that is up to 178M Gaussians per chunk, so each chunk fits a single-GPU VRAM target when loaded. scene.usda places the chunks, and manifest.json gives each chunk's bounds for streaming or culling.

Frame. All values are in metres and Y-up, with x = east, y = up and z = south. They are given relative to origin_utm_enh in manifest.json, which is in ETRS89 / UTM 33N (EPSG:25833). To convert back: E = x + o[0], N = -z + o[1], H = y + o[2]. The same origin is stored in scene.usda's customLayerData.

berlin_city/ (full city)

Same chunking rule and frame convention, at city scale. SPZ positions only cover ±2 km, so every chunk (and every LOD tile) is stored relative to its own origin, and scene.usda places it with a float64 translate relative to one city origin (manifest.json: city_origin_utm_enh). Coarser overview files and their USD layers have been removed to retain the finest city representation.

Format. SPZ v3 files with degree-0 spherical harmonics. Positions are 24-bit fixed point with 12 fractional bits, which gives 0.24 mm precision over ±2 km.

Conversion

Each Gaussian is placed analytically on the textured mesh surface by the splataverse mesh-to-splat converter. There is no optimisation step and no rendering loss. Each texel becomes a flat Gaussian on its triangle, coloured with the texel's colour.

Licence and attribution

The source meshes come from the MovingDrone dataset (Dhaouadi et al., TUM). Check the original dataset's terms before redistributing. Terms for the textured mesh layer had not been confirmed when this was published. The other geodata in that dataset (LoD2, DSM, orthophotos) is listed under dl-de/by-2-0.