Add a detector-response tutorial (MTF and MCF)

[TomaSusi]

Gap: abtem/mtf.py (MTF) and abtem/mcf.py (DiagonalMCF) have zero documentation coverage — no tutorial, walkthrough mention, or example uses them.

Why it matters: Detector response directly affects quantitative comparison between simulated and experimental images; users comparing to real cameras need to know these exist and when to apply them.

Suggested content: A tutorial notebook (e.g. user_guide/tutorials/detector_response.ipynb) that:

• explains what the modulation transfer function models and how to specify MTF(func=...) for a real camera;
• shows DiagonalMCF (parameters like semiangle_cutoff, focal_spread, source_size) and how it differs from/complements the MTF;
• applies both to a simple HRTEM or STEM simulation and compares against the ideal-detector result side by side.

---

From a docs coverage audit (July 2026) comparing the abTEM public API against documentation content.

🤖 Generated with Claude Code

[TomaSusi]

Investigated this against abTEM dev (e62e7dbe) and found both MTF and DiagonalMCF are broken, not just undocumented:

• abtem.mtf.MTF.__call__ reads measurement.calibrations, an attribute that no longer exists on any current measurement class (replaced by axes_metadata in an earlier refactor). Calling it on a real Images/DiffractionPatterns object raises AttributeError: 'Images' object has no attribute 'calibrations'. test/test_mtf.py only exercises it against a MagicMock built to have .calibrations — its own docstring says as much: "MTF.call uses the legacy calibrations attribute (not axes_metadata). We test it via a minimal mock that matches the interface the code expects."
• abtem.mcf doesn't import at all: from abtem.transfer import ArrayWaveTransform — ArrayWaveTransform was renamed/removed when transfer.py was refactored (current base classes are WavesTransform / EnsembleTransform etc.). from abtem.mcf import DiagonalMCF raises ImportError outright.

Neither is exported from abtem/__init__.py, and nothing else in the codebase references either module — this is unreferenced dead code that bitrotted after a refactor, not a documentation gap.

Writing a tutorial per the original suggestion isn't possible without either (a) presenting code that crashes when readers run it, or (b) working around the bugs in a way that misrepresents what the shipped API actually does. Filed for a fix upstream; this issue should stay open until that lands, at which point the original suggested content (MTF for camera response, DiagonalMCF for partial coherence, applied to a simple HRTEM/STEM comparison) is still the right scope.