genotf.h File Reference

Functions
void	genotf (ccell **potf, loc_t *loc, const dmat *amp, const dmat *opdbias, const dmat *area, real thres, real wvl, const dmat *cov, real r0, real l0, long npsfx, long npsfy, long nsa, long pttr)
dmat *	mk2dcov (loc_t *loc, const dmat *amp, real ampthres, const dmat *cov, int norm)

Detailed Description

Routines to generate short/long exposure OTFs of an aperture in present of atmosphere turbulence.

Function Documentation

genotf()

void genotf	(	ccell **	potf,
		loc_t *	loc,
		const dmat *	amp,
		const dmat *	opdbias,
		const dmat *	area,
		real	thres,
		real	wvl,
		const dmat *	cov,
		real	r0,
		real	l0,
		long	npsfx,
		long	npsfy,
		long	nsa,
		long	pttr
	)

Generate OTFs for an aperture or multiple subapertures. ALl these apertures must share the same geometry, but may come with different amplitude map and/or OPD biasas. if pttr is 1, the OTF will have tip/tilt removed. make r0 to infinity to build diffraction limited OTF. make r0 to infinity and opdbias to none null to build OTF for a static map.

2020-01-21: Compute OTF using nyquist sampling and then upsample with FFT.

Parameters

potf	The otf array for output
loc	the aperture grid (same for all apertures)
amp	The amplitude map of all the (sub)apertures
opdbias	The static OPD bias (complex part of amp).
area	normalized area of the (sub)apertures
thres	The threshold to consider a (sub)aperture as full
wvl	The wavelength. only needef if opdbias is not null
cov	The covariance. If not supplied use r0 for kolmogorov spectrum.
r0	Fried parameter
l0	Outer scale
npsfx	Size of PSF
npsfy	Size of PSF
nsa	Number of (sub)apertures
pttr	Remove piston/tip/tilt

mk2dcov()

dmat* mk2dcov	(	loc_t *	loc,
		const dmat *	amp,
		real	ampthres,
		const dmat *	cov,
		int	norm
	)

Average spatially the 4-d covariance function to create a 2-d covariance function. For OPD f defined on points x (2-d coordinate), the 4-d covariance is simply <f'f> where f is vector form of the OPD and the average is over time. The 2-d covariance is additionally averaged over all the points so that B(r)=<f(x)'f(x+r)>_x,t To compute B, we first figure out the number of overlapping pairs of points for each r and then compute the averaging. When the amplitude is less than the threshold, the point does not count.