2011.09.21-Big Images & OME-TIFF
Attending: Melissa, Jean-Marie, Andrew, Josh
Josh: HDF
- tongue-in-cheek
- would be better but not really relevant
- since we don’t have anything ready.
- Melissa: Java libraries are also
- http://lists.openmicroscopy.org.uk/pipermail/ome-devel/2009-February/001150.html
- http://www.openmicroscopy.org/site/community/minutes/conference-calls/2007/2007-12-27_Project-wide
Melissa: our pyramid files are following the TIFF spec
Andrew: are we clear what we need to try this?
- Melissa: the existing metadata is enough
-
Success is:
- bfconvert .svs to .ometiff
- if that works, then we should be fine.
- Tile size is stored in the TIFF header
- We let the container formats handle chunking/tiling.
Jean-Marie: cF Imaris.
- Offer of help from Marius
- Need to be careful to not overwhelm libtiffs eventually.
Andrew:
- BinData? The files would be huge.
-
Melissa: OmeXmlWriter does NOT support writing chunks.
- There’s a ticket for it.
Next steps:
- Josh: so we’re saying our pyramid files should become the “standard”?
-
Melissa: pyramids don’t have OME-XML block yet.
- Josh: right but add on export
- Andrew: use the metadata-only companion file?
-
Melissa: do we want to require the full pyramid? Or optional?
- Andrew: optional since others will make a mess of it?
- Josh: it’s about consumption
- Melissa: to strongly recommend we should push into BF
- Andrew: so no model changes?
- Melissa: for full pyramid solution
would help to have a field with info on the resolutions
Josh: with the current pyramid files and no extra info...
- Andrew: could you tell it was a pyramid?
-
Melissa: visually, yes.
- You could check aspeect ratios...
Andrew: i.e. no flag in TIFF to say, “pyramid”
- IDs are the sequence of images that make the pyramid, etc.
Andrew: short-term top-level annotation to describe pyramids.
-
Josh/Chris: modify exporter to inject OME-XML
- Melissa: filename of .ome.tiff should just work
-
Josh/Chris: modify exporter to inject OME-XML