search

ome / ngff

Next-generation file format (NGFF) specifications for storing bioimaging data in the cloud.
https://ngff.openmicroscopy.org
Other
113 stars 38 forks source link

Z-downsampling metadata #25

Closed will-moore closed 2 years ago

will-moore commented 3 years ago

It's possible that we may generate multiple "datasets" for a given OME-Zarr Image, e.g. with and without down-sampling in Z. It could be useful to have a way to identify these.

See https://github.com/hms-dbmi/vizarr/pull/71#issuecomment-760261353 "will there be any way to distinguish between the downsampled-in-z version of a dataset and the non-downsampled version? The shape of the zarr array, I presume? But will there be any metadata about it? Just curious........."

@joshmoore "One thing we could do at the moment (pre-lightsheet changes) would be to have a naming convention for the multiscales themselves. Another option would be to use the metadata for the method of the downsampling, but we’d need something well-defined to say “in-z".

ilan-gold commented 3 years ago

Just out of curiosity @will-moore @joshmoore, what are the "lightsheet changes?"

joshmoore commented 3 years ago

Hi @ilan-gold. The next body of work is focused around having better support for large 3D volumes. We'll likely use the opportunity to make the first breaking changes to the metadata (0.1 --> 0.2) so if there's something that's needed (e.g. the way to detect the z-downsampled multiscale zgroup as you were asking in https://github.com/hms-dbmi/vizarr/pull/71#issuecomment-760261353) then it would be good to roll them into the "lightsheet changes".

ilan-gold commented 3 years ago

Gotcha @joshmoore, I know @will-moore and I talked on image.sc about scaling as well, derived from the reported physical size, so that would be great too in the next change. We are very interested in this 3D stuff as well so I'll keep my ear to the ground.

constantinpape commented 2 years ago

The downsamplling is now specified using coordinateTransformations:scale in v0.4; so the downsampling across all spatial dimensions is transparent.