napari
The napari-openreadout plugin opens microscopy files in napari without converting them first.
pip install napari napari-openreadoutnapari slide.svsYou can also use File › Open in napari, or viewer.open(path, plugin="napari-openreadout") from Python.
The plugin registers for the image formats OpenReadout reads, including CZI, ND2, LIF, Olympus OIR, OIB, OIF and VSI, whole-slide SVS, NDPI and QPTIFF, Imaris, OME-TIFF, OME-Zarr, MetaMorph, Hamamatsu DCIMG and the electron-microscopy formats MRC, DM3/DM4, SER and EMD.
What you get:
- Each image in the file (a CZI scene, an ND2 position, a LIF series) is its own layer.
- Each channel is a separate layer, named and coloured as in the file. RGB images stay one layer.
- Layers are scaled by the pixel size in micrometres (and the time step in seconds), so 3D views and the scale bar are correct.
- Layers are lazy dask arrays. napari decodes a plane, or for large tiled planes the tiles in view, only when it displays it.
- Files that store a pyramid, such as whole-slide images, open as multiscale layers.
Without napari, napari_openreadout.read(path) returns the same layer data.
If another installed plugin also claims an extension, napari asks which reader to use, or you can pass plugin="napari-openreadout".
From Python, without the plugin
The Python package gives you the same lazy arrays, if you want to build the layers yourself:
import napariimport openreadout
f = openreadout.File("run.czi") # keep it open while the viewer is upim = f.images[0]ps = im.get("physical_size", {})viewer = napari.Viewer()viewer.add_image( f.to_dask(0), # (T, C, Z, Y, X) channel_axis=1, name=[ch.get("name", f"c{ch['index']}") for ch in im.get("channels", [])] or None, scale=(1, ps.get("z", 1), ps.get("y", 1), ps.get("x", 1)),)napari.run()For a pyramid, pass f.pyramid(0) with multiscale=True instead.
From an export
An OME-TIFF or OME-Zarr export opens with napari’s own readers or with plugins such as napari-ome-zarr. Use this when you want a file to hand to someone else:
openreadout export run.czi -o run.ome.tiff