Common pulse retrieval algorithm: a fast and universal method to retrieve ultrashort pulses

TL;DR

COPRA is a versatile, fast, and accurate algorithm for retrieving ultrashort laser pulses across various measurement techniques, outperforming existing methods especially in noisy conditions and enabling new measurement capabilities.

Contribution

The paper introduces COPRA, a universal pulse retrieval algorithm that improves speed, accuracy, and applicability over existing methods for a wide range of ultrashort pulse measurement schemes.

Findings

COPRA outperforms generalized projections and ptychography algorithms.

It is robust against Gaussian noise in pulse measurements.

Enables pulse retrieval in previously unsupported measurement techniques like MIIPS.

Abstract

We present a common pulse retrieval algorithm (COPRA) that can be used for a broad category of ultrashort laser pulse measurement schemes including frequency-resolved optical gating (FROG), interferometric FROG, dispersion scan, time domain ptychography, and pulse shaper assisted techniques such as multiphoton intrapulse interference phase scan (MIIPS). We demonstrate its properties in comprehensive numerical tests and show that it is fast, reliable and accurate in the presence of Gaussian noise. For FROG it outperforms retrieval algorithms based on generalized projections and ptychography. Furthermore, we discuss the pulse retrieval problem as a nonlinear least-squares problem and demonstrate the importance of obtaining a least-squares solution for noisy data. These results improve and extend the possibilities of numerical pulse retrieval. COPRA is faster and provides more accurate results in comparison to existing retrieval algorithms. Furthermore, it enables full pulse retrieval from measurements for which no retrieval algorithm was known before, e.g., MIIPS measurements.