This work belongs to a series of quantum metrology schemes that exploit the interference of indistinguishable photons. Neighboring studies have already shown that frequency resolution can bypass the limitations of standard Hong-Ou-Mandel interferometry, but so far this has been demonstrated either with nonclassical sources or in theoretical proposals. This paper is the first to implement the scheme with classical light and an SPAD array, making the method practical and scalable. It fills the gap between theory and experiment for realistic detectors with finite resolution.
The setup is based on a Mach-Zehnder interferometer with a pulsed laser (531.5 nm, 40 MHz). One arm introduces a tunable delay, while a piezoelectric mirror in the other arm randomizes the phase, eliminating first-order interference. After the second beam splitter, the photons are directed onto a diffraction grating (1200 lines/mm) and a lens that focuses the spectrum onto 8 pixels of a 32-channel SPAD array with a temporal resolution of 1.5 ps. Coincidences between pixels are recorded in two time windows: an 'early' window (bunching) and a 'late' window (antibunching), each 2 ns wide. The frequency resolution δω ≈ 1.8 THz is determined by the pixel spacing and the 50 μm diameter of the sensitive area.
The experimental data show quantum beats in coincidences between pixels with different frequencies, with oscillations visible even at delays several times the coherence time τ ≈ 0.44 ps. For the unresolved scheme, the uncertainty grows rapidly with delay, as theory predicts. In the frequency-resolved version, the uncertainty per coincidence is about 10 ps for delays up to 4 ps, orders of magnitude better than the unresolved case. The obtained values lie near the Cramér-Rao bound for finite frequency resolution and confirm the prediction that frequency resolution erases the temporal distinguishability of photons.
Unlike neighboring works that use tunable filters or dispersive fibers, this work employs a diffraction grating with an SPAD array, enabling parallel spectral readout and high data acquisition speed. This is the first experimental demonstration of delay estimation with frequency-resolved Hong-Ou-Mandel interferometry using classical light, lifting the brightness and bandwidth limitations inherent to single-photon sources. The work confirms that quantum advantage in metrology is achievable without entanglement.
The method can be extended to estimate transverse displacements and other parameters where frequency resolution also lifts degeneracy. Increasing the number of pixels and improving their temporal resolution will bring the scheme closer to the quantum limit of precision. Integration with fast electronics and machine learning algorithms could make the setup suitable for field measurements.
The results will find applications in quantum lidar, optical coherence tomography, and synchronization of remote clocks, where precise measurement of time delays is required.
Plans include adapting the scheme to measure transverse displacements and investigating behavior at higher photon fluxes to test the influence of multiphoton components.
This work is directly related to a fundamental problem in quantum metrology: achieving the Heisenberg limit with limited resources. It shows how to use classical light for tasks previously considered exclusively quantum, and clarifies the role of frequency resolution in erasing which-path information.
This section is addressed to the author of the paper and their colleagues. Neighbouring works were found by meaning-based search across our archive, not by word overlap. We do not review: we show the field around the work.
What the work does
The study demonstrates that frequency-resolved Hong-Ou-Mandel interferometry can measure delays exceeding the coherence time of photons. Using weak coherent states and a diffraction grating coupled to a SPAD array, the authors achieved an uncertainty of about 10 ps for delays up to 4 ps, which agrees with quantum estimation theory.
Where its strength is
The strength of the work lies in the practical implementation of frequency-resolved Hong-Ou-Mandel interferometry with classical light and parallel spectral readout, which lifts brightness limitations and makes the method scalable.
Where it could go next
Worth trying
Why this may matter
The method could serve as a foundation for compact and inexpensive time measurement devices that do not require non-classical light sources, which is important for practical quantum metrology.
Nearby works in our archive
The closest work in the archive sits at 0.65 by meaning — the usual level for our corpus.
The Hong-Ou-Mandel effect was first observed in 1987, measuring time intervals between two photons with sub-picosecond precision. Now the same physics works for delays ten times longer than the pulses themselves.