mac4DSTEM

Version 1.0 — available now

mac4DSTEM

Four-dimensional STEM analysis, native on Apple silicon.

A macOS application that takes 4D-STEM datasets from raw detector frames to calibrated strain, orientation, and phase maps — interactively, against a live view of the data.

Download for macOS

Version 1.0 · 5.1 MB · signed and notarized

  • macOS 14+
  • Apple silicon
  • Metal & MLX
  • GPL-3.0
  • Free

Try it — real data, in the page

A real 4D-STEM scan of a NiCu alloy. Every scan position holds a full diffraction pattern, and each space produces the other: the detector mask on the left integrates to the virtual image on the right, and the region you pick on the right integrates to the pattern on the left. Computed in your browser.

Loading the dataset…

The workspace

The mac4DSTEM strain workspace. A convergent-beam diffraction pattern with detected Bragg disks and lattice fit overlay sits beside the resulting epsilon-xx strain map, with dataset metadata, calibration, and fit diagnostics in side panels.
Strain mapping on an experimental Si/SiGe dataset. The diffraction view shows measured disks, the local lattice fit, and the reference basis; the map beside it updates from the same fit. Basis consensus, residual, and indexed fraction are reported alongside every result rather than left implicit. Open full size ↗

Capabilities

  • Data exploration

    Real and reciprocal space presented together, with calibrated scale bars, contrast control, and GPU-rendered CBED display.

  • Virtual imaging

    Bright-field, annular dark-field, and HAADF geometries, plus arbitrary annular, rectangular, and point detectors positioned interactively on the diffraction plane.

  • Calibration

    Probe and origin fitting across the scan, elliptical distortion correction, real–reciprocal rotation, and reciprocal-space sampling against a known crystal.

  • Bragg disk detection

    GPU cross-correlation with parabolic and DFT-upsampled subpixel refinement, and live acceptance diagnostics.

  • Strain mapping

    Robust local lattice fitting against a whole-scan or region reference, with basis consensus, residual, and indexed fraction reported alongside every map.

  • Orientation mapping

    Polar-correlation template matching against a validated crystal catalogue or an imported CIF, with reliability and inverse-pole-figure output.

  • Phase reconstruction

    DPC, iDPC, parallax, and single-slice ptychography.

  • Interoperability

    EMD export of Bragg vectors, calibrated datacubes, and preprocessed products, readable directly in py4DSTEM.

Provenance and validation

The analysis algorithms are ported from py4DSTEM and measured against it on experimental datasets, so results are traceable to the established reference implementation in the field. Work published using mac4DSTEM should cite py4DSTEM for the underlying methods.

Every quantity carries its calibration and provenance through display, export, and reopening. Where a measurement is not quantitatively supported, the result is labelled accordingly rather than presented as though it were.

That verification is numerical — the application is tested against known-correct values, not against its rendered output.

Reference
py4DSTEM 0.14.19
Test suites
30 automated
Parity harnesses
Experimental data
Export
EMD (py4DSTEM-readable)
Runtime
Self-contained
Licence
GPL-3.0

Availability

  1. Version 1.0 complete

    The analysis workflow is finished and covered by automated tests, including parity harnesses measured against py4DSTEM on experimental data.

  2. Download

    Version 1.0 is available as a signed, notarized disk image. Open it and drag mac4DSTEM to your Applications folder — it launches without a security warning, and needs nothing else installed.

  3. Source release

    The source is published on GitHub under GPL-3.0.

  4. Distribution

    Direct download only. The Mac App Store is not a route for this application: its terms are incompatible with the GPL-3.0 licence mac4DSTEM inherits from py4DSTEM. It stays free either way.

Requires macOS 14 or later on Apple silicon. No external runtime, toolchain, or library installation is needed — all dependencies are contained within the application. The Merlin MIB and EMPAD RAW readers are preview-grade pending further vendor data.

Enquiries, issue reports, and collaboration

If you work with 4D-STEM data and something here is useful — or missing — I would like to hear about it.

mail@mac4dstem.com