> This is __really__ stunning work Thank you :) > Is it a consequence of input photo quality, or something else? It's more a consequence of spatial resolutio

> This is __really__ stunning work Thank you :) > Is it a consequence of input photo quality, or something else? It's more a consequence of spatial resolution: the bigger the space, the more voxels you need to maintain a fixed resolution (e.g. 1 mm^3). At some point, we have to give up spatial resolution to represent larger scenes. A second limitation is the teacher model we're distilling. Zip-NeRF (https://jonbarron.info/zipnerf/ https://jonbarron.info/zipnerf/) is good, but it's not _perfect_. SMERF reconstruction quality is upper-bounded by its Zip-NeRF teacher.