Deformable Mirror¶

Overview¶

The DeformableMirror (DM) models a continuous face-sheet mirror whose surface is described by a set of actuator influence functions. By default, influence functions are 2-D Gaussians in a Fried geometry (one actuator per subaperture, aligned with the WFS grid), but fully user-defined influence functions or modal bases can also be supplied.

Important

The DM is modelled as transmissive, not reflective, to avoid factor-of-2 confusion in OPD. The OPD produced is the physical surface deformation, not twice that value.

The DM is conjugated to the ground layer by default. An optional altitude parameter places it at a conjugate height (e.g. for MCAO simulations).

Key concepts¶

Fried geometry — actuators are placed at subaperture corners, with nSubap + 1 actuators across the diameter.
Mechanical coupling — Gaussian sigma is set so that the deformation at distance pitch from a pushed actuator equals mechCoupling (default 0.35).
coefs — the actuator command vector. Setting dm.coefs = vec immediately recomputes OPD.
Mis-registration — a MisRegistration object applies rotation, shift, and scaling to the influence functions at construction time.

Quick start¶

from OOPAO.DeformableMirror import DeformableMirror

dm = DeformableMirror(telescope=tel, nSubap=20, mechCoupling=0.35)

# Apply a flat mirror (zero commands)
dm.coefs = 0

# Push the first actuator by 100 nm
import numpy as np
cmd = np.zeros(dm.nValidAct)
cmd[0] = 100e-9
dm.coefs = cmd

# Propagate through the DM
tel * dm   # tel.OPD += dm.OPD

API reference¶

class OOPAO.DeformableMirror.DeformableMirror(telescope, nSubap, mechCoupling=0.35, coordinates=None, pitch=None, modes=None, misReg=None, M4_param=None, nJobs=30, nThreads=20, print_dm_properties=True, floating_precision=64, altitude=None, flip=False, flip_lr=False, sign=1, actuator_selection=None)[source]¶

Deformable mirror with Gaussian or user-defined influence functions.

Parameters:

telescope (Telescope) – Associated telescope object. Provides pupil, pixel scale, and obstruction geometry for actuator selection.
nSubap (float) – Number of WFS subapertures across the pupil diameter. Defines the Fried geometry actuator pitch when coordinates and pitch are None.
mechCoupling (float) – Mechanical coupling coefficient — deformation at distance pitch from a unit-pushed actuator. Ignored when modes is specified. Default 0.35.
coordinates (numpy.ndarray or None) – User-defined actuator coordinates as an (N, 2) array in metres. When supplied, all actuators are treated as valid regardless of pupil geometry. Default None.
pitch (float or None) – Inter-actuator distance in metres for the Gaussian influence function width. Defaults to diameter / nSubap when None.
modes (numpy.ndarray or None) – User-defined influence functions as a 2-D matrix of shape (resolution², n_modes), where each column is a flattened 2-D mode. Default None.
misReg (MisRegistration or None) – MisRegistration object specifying geometric transformations to apply. Ignored when modes is provided. Default None.
M4_param (object or None) – Parameter file for ELT M4 mirror model. Default None.
nJobs (int) – Number of parallel jobs for influence function computation. Default 30.
nThreads (int) – Number of threads per job. Default 20.
print_dm_properties (bool) – Print DM properties at initialisation. Default True.
floating_precision (int) – Array precision: 64 (float64, default) or 32 (float32, lower memory).
altitude (float or None) – Conjugation altitude in metres. None = ground conjugation (default).
flip (bool) – Flip influence functions vertically. Default False.
flip_lr (bool) – Flip influence functions horizontally. Default False.
sign (int) – Sign applied to DM commands (+1 or -1). Default 1.
actuator_selection (None, float, or list[float]) –
Controls which actuators are selected as valid.
- None — selected by pupil geometry (default).
- Scalar — selected by standard deviation of influence function within the pupil (threshold = scalar).
- [r_inner, r_outer] — selected by radial distance r_inner < r < r_outer.

Key properties

coefs: numpy.ndarray or float¶: Actuator command vector in metres (units of the influence functions). Setting this property immediately recomputes OPD. Accepts a scalar (broadcasts to all valid actuators) or a 1-D array of length nValidAct.

OPD: numpy.ndarray¶: 2-D optical path difference map in metres produced by the current commands.

nValidAct: int¶: Number of valid (selected) actuators.

modes: numpy.ndarray¶: 2-D influence function matrix of shape (resolution², nValidAct).

pitch: float¶: Inter-actuator pitch in metres.

misReg: MisRegistration¶: Current mis-registration state of the DM.

Methods

set_coefs_value(actuator_index, actuator_coefs)[source]¶

Update the commands of one or more individual actuators and recompute the DM surface accordingly. Use this when you need to set specific actuators without rebuilding the full command vector.

Parameters:

actuator_index (list[int]) – List of actuator indices to update.
actuator_coefs (list[float]) – List of command values in metres, one per index.

# Push actuator 42 by 50 nm
dm.set_coefs_value([42], [50e-9])

# Push two actuators simultaneously
dm.set_coefs_value([10, 42], [50e-9, -50e-9])

Note

This is distinct from dm.coefs = value, which replaces the entire command vector and expects a 1-D array of length nValidAct. Use set_coefs_value() when updating individual actuators; use dm.coefs when applying a full modal correction.

display()¶: Show the current OPD map and actuator grid in a matplotlib figure.

Operator summary

Expression	Effect
`tel * dm`	Add `dm.OPD` to `tel.OPD`
`src * tel * dm`	Full propagation: source → telescope → DM