arXiv Science⌕ Search

arXiv · 2609.37120

First laboratory demonstration of iEFC for high-contrast imaging at small angular separation on ELT scales

Abstract

Implicit electric field conjugation (iEFC) is an efficient, model-free wavefront control technique for dark hole creation. Its model-free nature, relying on calibration with real data rather than a synthetic model of the instrument, makes it uniquely versatile and highly attractive. It has been successfully demonstrated in a variety of environments, both in the laboratory and on-sky, using different experimental platforms and coronagraphs. A common characteristic of these demonstrations is that the resulting DHs exhibit moderate to large inner working angles and wide fields of view, often extending over several tens of $λ/D$. Here, we report the first demonstration of iEFC at small angular separations, achieving a DH spanning 1 to 4 $λ/D$ on a segmented pupil in H-band under laboratory conditions. This combination of aggressive IWA and reduced FoV highlights the uniqueness of the SPEED facility (segmented pupil experiment for exoplanet detection). SPEED is a high-contrast imaging testbed designed for extremely large telescopes (ELTs), developed in the context of the ELT/PCS instrument (planetary camera and spectrograph), currently in pre-phase A. Its small-IWA coronagraph, an APCMC (apodized pupil complex mask coronagraph), and its multi- and out-of-pupil-plane deformable mirror (DM) architecture are specifically optimized for DHs at very small IWAs, down to the diffraction limit. Our results demonstrate (i) that widely separated out-of-pupil DMs are well suited to generating DHs with IWAs down to 1 $λ/D$, and (ii) the effectiveness of iEFC, when combined with the SPEED wavefront control architecture, for high-contrast imaging with segmented telescopes. Ultimately, these results pave the way toward meeting the stringent high-contrast requirements of ELT-class instruments, opening new opportunities for exoplanet detection at unprecedented angular resolutions.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Charles Sallard, Patrice Martinez, Alain Spang, Kian Milani, Lyu Abe, Mathilde Beaulieu, Carole Gouvret, Aurélie Marcotto, Julien Dejonghe, Olivier Preis. 2026-09-29. First laboratory demonstration of iEFC for high-contrast imaging at small angular separation on ELT scales. https://doi.org/10.1117/12.3102891

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

Using 23 Years of ACS/SBC Data to Understand Backgrounds: Significant Reductions in Expected Background Levels

We have used 23 years of Hubble Space Telescope ACS/SBC data to study what background levels are encountered in practice and how much they vary. The backgrounds vary considerably, with F115LP, F122M, F125LP, PR110L, and PR130L all showing over an order of magnitude of variation in background between observations, apparently due to changes in airglow. The F150LP and F165LP filters, which are dominated by dark rate, not airglow, exhibit a far smaller variation in backgrounds. For the filters where the background is generally dominated by airglow, the backgrounds measured from the data are significantly lower than what the ETC predicts (as of ETC v33.2). The ETC predictions for `average' airglow are greater than the median of our measured background values by factors of 2.51, 2.64, 105, and 3.64, for F115LP, F122M, F125LP, and F140LP, respectively. A preliminary analysis suggests this could be due to certain OI airglow lines usually being fainter than expected by the ETC. With reduced reduced background levels, the shorter-wavelength SBC filters can conduct background-limited observations much more rapidly than had previously been expected. As of ETC v34.1, a new option will be included for SBC calculations, allowing users to employ empirical background percentiles to estimate required exposure times.

astro-ph.IM↗

Using 23 Years of ACS/SBC Data to Understand Backgrounds: Explaining & Predicting Background Variations

Recent analysis of 23 years of Hubble Space Telescope ACS/SBC data has shown that background levels can vary considerably between observations, with most filters showing over an order of magnitude variation. For the shorter-wavelength filters, the background is understood to be dominated by airglow; however, what precisely drives background variations is not well constrained for any filter. Here, we explore the causes of the background variation. Using over 8,000 archival SBC observations, we developed a machine learning model that can accurately predict the background for an observation, based upon a set of 23 observational parameters. This model indicates that, depending on filter, the SBC background is generally dominated by Solar elevation, Solar separation angle, Earth limb angle of observation, SBC temperature, and target Galactic latitude.

astro-ph.IM↗

Beyond the BLUE I: the advantage ceiling - how much can any estimator beat the matched filter in mm/submm survey data?

Papers claiming that convolutional neural networks outperform the matched filter are now common in astronomy. Since that filter is the best linear unbiased estimator (BLUE) for any noise of given covariance, and the minimum-variance unbiased estimator under Gaussian noise, such a claim requires that the noise covariance varies from image to image or the noise is non-Gaussian. We introduce the advantage ceiling eta >= 1, which quantifies both cases: the Fisher information for the amplitude in units of the matched filter's, computable from noise-only simulations before any network is trained, and a bound on the variance of any estimator unbiased at each true amplitude. We compute eta for a suite of millimeter/submillimeter survey noise models and for compact and extended source templates, using a variational score-matching ladder whose rungs are estimator classes of increasing statistical order, from linear to convolutional. With an instrumental white-noise floor, Gaussian noise returns eta = 0.94-1.06 (+-0.04); covariance mixtures give exact ceilings of 10.2 (spectral tilt, extended source), 2.1 (leaked components) and 1.1-1.8 otherwise; residual artifacts give eta <= 1.09. Source confusion gives the largest advantage, eta >= 3.8 (extended) and >= 2.4 (compact). Three unmodelled channels - a zeroed DC mode, a missing white floor and floating-point rounding - can manufacture spurious advantages of up to two orders of magnitude. A ResNet regressor is 6-15% less efficient than the matched filter on Gaussian noise once prior shrinkage is divided out, and realizes a factor 1.8 of the 10.2 available on the spectral-tilt mixture; a companion paper tests such regressors against these ceilings.

astro-ph.IM↗