How to open OME-Zarr volumes in a Jupyter notebook
Read an OME-Zarr store with ngff-zarr: nz.from_ngff_zarr(path) returns the multiscale image, and images[0].data is the full-resolution array. In a Scigantic notebook the ome-zarr-scivis bucket is mounted at /mnt/archive, so a store path is a folder path and only the chunks you index are read.
The code on this page runs in a Scigantic hosted notebook, where the dataset is mounted read-only at /mnt/archive. The same path is in the SCIGANTIC_MOUNT_PATH environment variable and linked from ~/archive. The public source is s3://ome-zarr-scivis (AWS us-east-1, public, no account needed).
Open OME-Zarr Open SciVis datasets on SciganticOpen the sample notebook: A 3D tour of open science volumes
Code
Read one volume and show its middle slice
# OME-Zarr SciVis: read one volume with ngff-zarr and show a middle slice with the store's own contrast window.
import os, json, numpy as np, matplotlib.pyplot as plt, ngff_zarr as nz
M = os.environ.get('SCIGANTIC_MOUNT_PATH', '/mnt/archive')
store = f'{M}/v0.4/128x0/lobster.ome.zarr'
img = nz.from_ngff_zarr(store)
arr = np.asarray(img.images[0].data)
win = json.load(open(f'{store}/.zattrs'))['omero']['channels'][0]['window']
print(arr.shape, arr.dtype, 'window', win['start'], win['end'])
plt.imshow(arr[arr.shape[0] // 2], cmap='gray', vmin=win['start'], vmax=win['end']); plt.axis('off')
The starter cell from this dataset's schema card. It ran without error on 2026-10-27 in the Scigantic notebook image against the mounted data.
What the files are
The Open SciVis collection converted to OME-Zarr: 52 classic scientific visualization test volumes, such as the lobster, bonsai, aneurism and backpack scans.
Every volume is stored in several variants so readers can be tested against each: OME-Zarr 0.4 (Zarr v2) and 0.5 (Zarr v3), three chunk sizes, and single-file .ozx zip stores.
How the files are laid out
| Path under the mount | What it holds |
|---|---|
| v0.4/{64x0,96x0,128x0}/<name>.ome.zarr/ | OME-Zarr 0.4 on Zarr v2, 52 volumes per chunk-size variant, with .zattrs, .zgroup and consolidated .zmetadata. |
| v0.4/128x0/<name>.ome.zarr/scale0/<name>/ ... scale4/ | One to five resolution levels per volume. |
| v0.5/{64,96,128}x{0,2,4}/<name>.ome.zarr/ | OME-Zarr 0.5 on Zarr v3, with zarr.json and chunks under c/. |
| v0.5/{64,96,128}x2-ozx/<name>.ozx | Single-file zip stores, from 2.3 MB (lobster) to 6.6 GB (3d_neurons_15_sept_2016). |
Size, format and license
| Format | OME-Zarr 0.4 (Zarr v2) and 0.5 (Zarr v3) |
|---|---|
| Size | 1,683,024 objects, 495 GB across all variants. Single stores are small: the lobster is 56 by 324 by 301 voxels. |
| License | CC BY 4.0 unless a volume states otherwise |
| Provider | NumFOCUS (Insight Software Consortium), on AWS Open Data |
| Public source | s3://ome-zarr-scivis (AWS us-east-1, public, no account needed) |
| Notebook image | Neuroscience (ngff-zarr and zarr preinstalled) |
Things to know before you read it
xr.open_zarron a store root returns an empty Dataset, because the arrays sit inscaleN/<name>groups. Use an OME-Zarr reader such as ngff-zarr.- The display contrast window is stored under
omero.channels[0].windowin.zattrs, which the code uses for the plot.
Run it
Start a notebook from the OME-Zarr Open SciVis datasets archive page. The data is mounted when the notebook opens, so the code above runs as written and nothing is downloaded first. For a longer worked example, open the sample notebook A 3D tour of open science volumes.
Related guides
- How to browse CZ CryoET Data Portal datasets in a Jupyter notebook
- How to open the Connectome 2.0 high b-value diffusion MRI dataset in a Jupyter notebook
- How to open CAMELYON lymph node whole-slide images from S3 in a Jupyter notebook
- How to open ECMWF IFS S2S reforecasts stored as Icechunk in a Jupyter notebook
Last checked 2026-11-21. All dataset guides.