Optical design, analysis, and calibration using ∂Lux

Desdoigts, Louis; Pope, Benjamin; Tuthill, P. G.

doi:10.1117/12.2629774

Cited by 3 publications

(3 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Preliminary experiments on GPUs have shown a decrease in runtime by a factor of ≈16×. Several investigators have been exploring the Jax Python package for its ability to perform automatic differentiation and parallel computing in support of physical optics simulations 70 – 72 Automatic differentiation could be extremely useful for future beamlet decomposition algorithms to improve the accuracy of the ABCD matrix computation.…”

Section: Discussionmentioning

confidence: 99%

“…Several investigators have been exploring the Jax Python package for its ability to perform automatic differentiation and parallel computing in support of physical optics simulations. [70][71][72] Automatic differentiation could be extremely useful for future beamlet decomposition algorithms to improve the accuracy of the ABCD matrix computation. Parallelization and vectorization is the most natural path forward for our beamlet decomposition algorithm because of the independence of the beamlet operations.…”

Section: Future Workmentioning

confidence: 99%

See 1 more Smart Citation

Hybrid propagation physics for the design and modeling of astronomical observatories: a coronagraphic example

Ashcraft,

Douglas,

Kim

et al. 2023

J. Astron. Telesc. Instrum. Syst.

View full text Add to dashboard Cite

.For diffraction-limited optical systems, an accurate physical optics model is necessary to properly evaluate instrument performance. Astronomical observatories outfitted with coronagraphs for direct exoplanet imaging require physical optics models to simulate the effects of misalignment and diffraction. Accurate knowledge of the observatory’s point-spread function (PSF) is integral for the design of high-contrast imaging instruments and simulation of astrophysical observations. The state of the art is to model the misalignment, ray aberration, and diffraction across multiple software packages, which complicates the design process. Gaussian beamlet decomposition (GBD) is a ray-based method of diffraction calculation that has been widely implemented in commercial optical design software. By performing the coherent calculation with data from the ray model of the observatory, the ray aberration errors can be fed directly into the physical optics model of the coronagraph, enabling a more integrated model of the observatory. We develop a formal algorithm for the transfer-matrix method of GBD and evaluate it against analytical results and a traditional physical optics model to assess the suitability of GBD for high-contrast imaging simulations. Our GBD simulations of the observatory PSF, when compared to the analytical Airy function, have a sum-normalized RMS difference of ≈10 − 6. These fields are then propagated through a Fraunhofer model of an exoplanet imaging coronagraph where the mean residual numerical contrast is 4 × 10 − 11, with a maximum near the inner working angle at 5 × 10 − 9. These results show considerable promise for the future development of GBD as a viable propagation technique in high-contrast imaging. We developed this algorithm in an open-source software package and outlined a path for its continued development to increase the accuracy and flexibility of diffraction simulations using GBD.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Future Workmentioning

confidence: 99%

Hybrid propagation physics for the design and modeling of astronomical observatories: a coronagraphic example

Ashcraft,

Douglas,

Kim

et al. 2023

J. Astron. Telesc. Instrum. Syst.

View full text Add to dashboard Cite

show abstract

“… – 19 Phase retrieval was memorably performed to infer and correct the serious aberration on the Hubble Space Telescope mirror at launch 20 . Earlier work in the vein of this paper has shown that the phase retrieval problem can be efficiently solved even in the case of detector nonlinearity by taking a forward model of physical optics, obtaining partial derivatives with autodiff, and optimizing its parameters by gradient descent, 21 , 22 including a conference presentation of an early version of the present work 23 …”

Section: Introductionmentioning

confidence: 99%

Differentiable optics with ∂Lux: I—deep calibration of flat field and phase retrieval with automatic differentiation

Desdoigts

Pope

Dennis

et al. 2023

J. Astron. Telesc. Instrum. Syst.

View full text Add to dashboard Cite

.The sensitivity limits of space telescopes are imposed by uncalibrated errors in the point spread function, photon-noise, background light, and detector sensitivity. These are typically calibrated with specialized wavefront sensor hardware and with flat fields obtained on the ground or with calibration sources, but these leave vulnerabilities to residual time-varying or non-common path aberrations and variations in the detector conditions. It is, therefore, desirable to infer these from science data alone, facing the prohibitively high dimensional problems of phase retrieval and pixel-level calibration. We introduce a new Python package for physical optics simulation, ∂ Lux, which uses the machine learning framework Jax to achieve graphics processing unit acceleration and automatic differentiation (autodiff), and apply this to simulating astronomical imaging. In this first of a series of papers, we show that gradient descent enabled by autodiff can be used to simultaneously perform phase retrieval and calibration of detector sensitivity, scaling efficiently to inferring millions of parameters. This new framework enables high dimensional optimization and inference in data analysis and hardware design in astronomy and beyond, which we explore in subsequent papers in this series.

show abstract

An Interpretable Machine-learning Framework for Modeling High-resolution Spectroscopic Data*

Gully-Santiago

Morley

2022

ApJ

View full text Add to dashboard Cite

Comparison of échelle spectra to synthetic models has become a computational statistics challenge, with over 10,000 individual spectral lines affecting a typical cool star échelle spectrum. Telluric artifacts, imperfect line lists, inexact continuum placement, and inflexible models frustrate the scientific promise of these information-rich data sets. Here we debut an interpretable machine-learning framework blasé that addresses these and other challenges. The semiempirical approach can be viewed as “transfer learning”—first pretraining models on noise-free precomputed synthetic spectral models, then learning the corrections to line depths and widths from whole-spectrum fitting to an observed spectrum. The auto-differentiable model employs back-propagation, the fundamental algorithm empowering modern deep learning and neural networks. Here, however, the 40,000+ parameters symbolize physically interpretable line profile properties such as amplitude, width, location, and shape, plus radial velocity and rotational broadening. This hybrid data-/model-driven framework allows joint modeling of stellar and telluric lines simultaneously, a potentially transformative step forward for mitigating the deleterious telluric contamination in the near-infrared. The blasé approach acts as both a deconvolution tool and semiempirical model. The general-purpose scaffolding may be extensible to many scientific applications, including precision radial velocities, Doppler imaging, chemical abundances for Galactic archeology, line veiling, magnetic fields, and remote sensing. Its sparse-matrix architecture and GPU acceleration make blasé fast. The open-source PyTorch-based code blase includes tutorials, Application Programming Interface documentation, and more. We show how the tool fits into the existing Python spectroscopy ecosystem, demonstrate a range of astrophysical applications, and discuss limitations and future extensions.

show abstract

Optical design, analysis, and calibration using ∂Lux

Cited by 3 publications

References 0 publications

Hybrid propagation physics for the design and modeling of astronomical observatories: a coronagraphic example

Hybrid propagation physics for the design and modeling of astronomical observatories: a coronagraphic example

Differentiable optics with ∂Lux: I—deep calibration of flat field and phase retrieval with automatic differentiation

An Interpretable Machine-learning Framework for Modeling High-resolution Spectroscopic Data*

Contact Info

Product

Resources

About