✨ Specification Features

This page describes the features of the OME-Zarr specification that are supported by ngff-zarr.

Features Overview

  • Multiscales: Support for multiscale representations of images.

  • Multiscales generation: Ability to generate multiscale representations from single-scale images.

  • Chunking: Customizable chunking strategies for efficient data access.

  • Compression: Support for various compression algorithms to reduce storage requirements.

  • Metadata: Rich metadata support, including spatial metadata.

  • Anatomical Orientation: Support for anatomical orientation metadata (RFC-4).

  • High Content Screening (HCS): Complete support for plate and well data structures.

  • Sharded Zarr: Support for sharded Zarr stores, allowing for scalable data management.

  • Format conversion: Conversion of most bioimaging file formats to OME-Zarr.

  • Tensorstore Writing: Optional writing via [tensorstore] for advanced use cases.

  • Model Context Protocol (MCP): Integration with the Model Context Protocol for AI agent interaction.

OME-Zarr Versions

  • OME-Zarr v0.1 to v0.5: Reads OME-Zarr versions 0.1 to 0.5 into simple Python data classes with Dask arrays.

  • OME-Zarr v0.4 to v0.5: Writes OME-Zarr versions 0.4 to 0.5, including support for RFC 4.

High Content Screening (HCS)

Complete implementation of the HCS specification defined in OME-Zarr v0.4+:

  • Plate Metadata: Support for plate-level metadata including rows, columns, wells, and acquisitions

  • Well Structure: Hierarchical well organization with multiple fields per well

  • Multi-field Imaging: Support for multiple fields of view within each well

  • Time Series: Support for acquisition metadata and time series data

  • Validation: HCS-specific metadata validation using the appropriate JSON schema

See the HCS documentation for detailed usage examples and implementation details.

RFCs Supported

  • RFC-1: We support and contribute to the NGFF specification via the “Request for Comments” (RFC) process described in RFC-1.

  • RFC-2: Support for Zarr v3, includingSharded Zarr stores, allowing for scalable data management.

  • RFC-4: Anatomical orientation support, allowing images to include metadata about their anatomical orientation.

  • RFC-9: OME-Zarr Zip (.ozx) format support, enabling single-file distribution of OME-Zarr datasets.

OME-Zarr Zip Format (.ozx)

RFC-9 defines the OME-Zarr Zip (.ozx) format, which packages complete OME-Zarr hierarchies into single ZIP archives. Key features include:

  • Portable Distribution: Share entire multiscale datasets as single files

  • Version Detection: OME-Zarr version stored in ZIP file comment for automatic format detection

  • Lazy Remote Access: Efficient HTTP byte-range requests for remote .ozx files

  • Compression: Individual files compressed within the ZIP archive

  • Standard Format: Based on standard ZIP file format for broad compatibility

All required and recommended RFC-9 practices are implemented in ngff-zarr.