Comment on "Quantum Limits to Incoherent Imaging are Achieved by Linear Interferometry"

Based on the provided text, here is a detailed technical summary of the paper "Comment on 'Quantum Limits to Incoherent Imaging are Achieved by Linear Interferometry'".

1. Problem Statement

The paper addresses a critical flaw in a recent study by Lupo et al. [1], which claimed that linear interferometry followed by photon counting always saturates the Quantum Fisher Information (QFI) limit for imaging $N$ weak incoherent emitters.

While the authors (Brumpton et al.) agree that the central claim—that linear interferometry can achieve quantum limits—is likely correct in principle, they demonstrate that the specific construction of the optimal interferometer proposed in the Supplemental Material of Ref. [1] is mathematically flawed. Consequently, the method proposed by Lupo et al. yields a suboptimal solution for general source configurations, failing to achieve the true quantum limit.

2. Methodology and Technical Analysis

The authors employ a rigorous mathematical analysis of the purification matrices and fidelity bounds to identify the error and propose a corrected construction.

Identification of the Error in Ref. [1]

• The Flawed Transition: The error occurs in the transition from Supplemental Eq. (31) to (32) in Ref. [1].
  • Ref. [1] defines the classical fidelity $f^c_{r,r'}$ as the sum of row norms: $\sum_v \|a'(v)\| \|b'(v)\|$.
  • They then apply a QR decomposition to choose a unitary transformation $R$ such that the transformed matrices $A'$ and $B'$ become upper and lower triangular, respectively.
  • They incorrectly assume this simplifies the fidelity to the sum of the products of diagonal elements: $\sum_v |a'(v,v)| |b'(v,v)|$.
• The Mathematical Contradiction: For general triangular matrices, the row norm $\|a'(v)\|$ is strictly greater than or equal to the absolute value of the diagonal element $|a'(v,v)|$, with equality holding only if the matrices are diagonal.
• Consequence: Since a QR decomposition of an arbitrary matrix does not generally yield a diagonal form, the construction in Ref. [1] results in a classical fidelity ($f^c$) that is strictly larger than the quantum fidelity ($f$). This creates a gap between the Classical Fisher Information (CFI) and the QFI, meaning the proposed interferometer is not optimal for general cases.
• Why it worked in Ref. [1] examples: The authors note that the specific examples in Ref. [1] involved two point sources with inversion symmetry. This symmetry forced the singular vectors to diagonalize the Gram matrices, making the QR decomposition effectively diagonal and masking the error.

Proposed Correct Construction

The authors derive the correct method to construct an optimal linear interferometer $R$ that ensures $f^c_{r,r'} = f_{r,r'}$ (saturation of the quantum limit).

1. Saturation Condition: The quantum fidelity is given by $f_{r,r'} = \text{Tr}(D) = \text{Tr}(B^\dagger A)$. Saturation occurs when, for each pair of rows in the transformed matrices $A'$ and $B'$, the vectors are either zero or proportional ($b'(v) = \lambda_v a'(v)$ with $\lambda_v > 0$).
2. Defining the Transformation:
   • Assuming the support of the source matrices is locally consistent (sharing the same column space), the authors define a matrix $P = B A^+$, where $A^+$ is the pseudoinverse of $A$.
   • This matrix $P$ is shown to be Hermitian and positive semidefinite.
3. Optimal Interferometer: The optimal unitary transformation $R$ is obtained by diagonalizing $P$:
   $$P = R^\dagger \Lambda R$$
   where $\Lambda$ is a diagonal matrix of non-negative eigenvalues. This ensures $RB = \Lambda RA$, satisfying the proportionality condition required for saturation.

Proof of QFI Saturation

The authors prove that in the limit of small parameter separation ($\delta\vartheta \to 0$), this construction diagonalizes the Symmetric Logarithmic Derivative (SLD).

• By expanding the density matrix $\rho$ and the matrix $P$ to first order in $\delta\vartheta$, they identify the SLD ($L$) as being proportional to the Hermitian part of the expansion of $P$.
• Since $R$ diagonalizes $P$, it simultaneously diagonalizes the SLD.
• Diagonalizing the SLD is the necessary and sufficient condition for a measurement to achieve the Quantum Fisher Information.

3. Key Contributions

• Correction of Literature: The paper identifies and mathematically disproves the validity of the specific QR-decomposition-based construction proposed in Ref. [1], showing it leads to suboptimal imaging performance for asymmetric or general source configurations.
• Counterexample Demonstration: Using a numerical example (Figure 1) with asymmetric source placement, the authors demonstrate that the QR-based method yields lower Fisher information compared to the true quantum limit, whereas their proposed method achieves the limit.
• General Solution: The authors provide a rigorous, general algorithm for constructing the optimal linear interferometer $R$ for any configuration of weak incoherent emitters by diagonalizing the matrix $P = B A^+$.

4. Results

• Suboptimality of QR Method: The QR-based approach fails to saturate the QFI for general source configurations, creating a gap between classical and quantum estimation precision.
• Optimality of Proposed Method: The new construction, based on the diagonalization of $P = B A^+$, successfully saturates the QFI.
• Validation: The results confirm that while linear interferometry can achieve quantum limits (validating the title of Ref. [1]), the specific implementation matters. The corrected interferometer $R$ ensures the title's claim holds true.

5. Significance

This comment is significant for the field of quantum metrology and imaging because:

1. Practical Implementation: It prevents researchers from implementing a suboptimal interferometer design that would fail to reach the theoretical maximum precision in real-world imaging scenarios (especially those lacking high symmetry).
2. Theoretical Rigor: It clarifies the mathematical conditions required for saturating the quantum limit in incoherent imaging, distinguishing between specific symmetric cases and the general case.
3. Future Research: It provides the correct mathematical framework (diagonalizing $P = B A^+$) for designing next-generation quantum imaging systems that utilize linear interferometry to resolve weak incoherent sources with ultimate precision.