melissalinkert
commented
1 year ago Usually channels (or Z sections) being detected as timepoints means that there is not enough metadata in the TIFF file to set the dimensions correctly. Bio-Formats defaults to treating all planes in a TIFF file as timepoints, in the absence of any other metadata.
This typically isn't a bug in bioformats2raw, and most often isn't a bug in Bio-Formats either. You'll need to make sure that the original data is written in a way that Bio-Formats can detect that the 5 planes are channels.
One option is to write each channel as a separate TIFF file (e.g. bigtiff_c0.tif, bigtiff_c1.tif, ...) and then use a .pattern file to group them together - see also https://github.com/glencoesoftware/bioformats2raw/issues/130.
Another option is to add channel metadata when writing with tifffile. Writing either ImageJ-style metadata or OME-TIFF should work, and https://github.com/cgohlke/tifffile#examples has examples of both. ImageJ-style is usually simpler if you just want to record that the planes are channels, but don't need to specify channel names, colors, etc.
Hi, we currently encounter a problem using bioformat2raw to convert multi-channel tiff with more than 5 channels.
The problem is that it will be converted to multi-timestep instead of multi-channel, and from the METADATA.ome.xml we found that sizeC=1 and sizeT=5 instead of sizeC=5 and sizeT=1.
What we did is to use the command line
bioformats2raw bigtiff_5c.tif bigtiff_5c.tif.pyramid --max-worker 32 -p.We are using tifffile (https://github.com/cgohlke/tifffile) library to combine different stain images(tif) all together to a large array [1, 5, 1, x, y] where 5 is the channel and save it to a single .tif and than using bioformat2raw to convert them.
With this pipeline, we can make it work for the channel less than 5.
Do you have any suggestions or workaround?