Thanks! Would shooting a series of sharp 20+ MP stills instead of video for the source result in a production ready 4K NerF model or is there something else in the translation process that is holding it back from full quality? If so, would adding a final ML/upscale pass solve it?
