seeing)

Welcome to the seeing simulation and deconvolution, release 23082001!

This software is copyrighted by Christian Pellegrin 2001. It's licensed under the terms of the GPL, please see the file COPYING in the root of the distribution directory or have a look at the GNU site.

Installing

In the tarball you can find the programs compiled for my platorm:

RedHat 7.1 GNU/Linux system
FFTW 2.1.3 compiled for signle precision
gsl 0.4.1
PDL 2.2.1

Here is a brief description of the process of compilation. The directory were you unpacked the distribution will be referred with $SEEING_HOME.

layerfft

Unfortunatly (I don't plane to rework the code) you have to compile with GSL v0.4.1 (very old). Anyway I included a working library and include files in case you want to recompile. You also will need FFTW compiled for single precision. The steps you have to take are:

Go to $SEEING_HOME/layer_fft/gsl_new and do make.
Go to $SEEING_HOME/layer_fft and do make.

PDL part

This shouldn't need any modifications.

Using the simulator

First of all you have to setup some environment variables by sourcing the file setup in $SEEING_HOME/layer_fft:
. $SEEING_HOME/setup $SEEING_HOME
After this various pieces of the system should found themself in the right way. In the directory $SEEING_HOME/prog you can find many driver programs that illustrate the working of the simulator and deconvolution. The libraries involved (that live in $SEEING_HOME/Lib)are:

Convolve.pm implements convolution using FFTW.
NoiseIt.pm adds shot and gaussian noise to an image once the manignude is given.
Seeing.pm the main module, uses the layer_fft external program.
Teo.pm calculation of theoretical strucure function and other similar functions.
Utils.pm miscellaneous utilities.
SI.pm implements the Cross-spectrum and Bispectrum image recovering.

The example programms

Have a look in the directory $SEEING_HOME/prog for example programs. All of them include the file common that setups the parameters for the simulation. A template file common.template fully commented is included, you should use it as a reference. The programs currently included are:

ltPSF calculates the long time and short time atmosphere PSF for validation purposes.
makeBSnonoise shotnoise-less (infinite brightness) Bi-Spectrum restoration simulation.
makeCSnonoise shotnoise-less (infinite brightness) Cross-Spectrum restoration simulation.
makeSnoise CS and BS simulation that takes in account the magnitude of the imaged object.
restoreBS final step in restoration of data outputed by makeBSnonoise .
restoreCS final step in restoration of data outputed by makeCSnonoise .
restoreComp final step in restoration of data outputed by makeSnoise with comparative output.
testD , testLT and testLT-RW , calculation of theorical functions for validation purposes.
testSee generic simulation loop.

LayerFFT options

Syntax: layer_fft [options] filename

-s int: set scale
-x double: x dimesion of layer
-y double: y dimension of layer
-r double: r0
-L double: L0
-h int: number of subharmonics
-k int: aspect ratio correction 0=none, 1=auto, 2=user supplied
-w double: x aspect ratio
-W double: y aspect ratio
-f string: outputs structure function to file
-v :be verbose
-z string: load wisdom from filename
-e int: set seed
-m int: subharmonic method (1,2)
-S int: subharmonic screen reduction ratio
-c string: read coefficents of subharmonics from file
-D int: calculate structure function in number iterations
-G double: use a Gaussian freq filter with given SIGMA
-A double: SIGMA in direct space (amplitude of initial gaussian noise) (only for -G and -H)
-a :to keep power integral kostant in gaussian mode (only for -G and -H)
-H double: top hut freq filter at given spatial frequency
-T double: non Taylor case: cuts after given value the filter

Available modules

Convolve.pm

$convolutor = Convolve->new($image1);
$convolved = $convolutor->convolve($image2);

NoiseIt.pm

$nphotons = phmag($magnitude, $diameter, $integration_time);
$shot_noise_image = Shot($image, $nphotons);
$image_plus_gaussian = AddG($image, $sigma);
$image_plus_poisson = AddP($image, $sigma);

Seeing.pm

$seenig = Seeing->new([[$fname,$scale,$x,$y,$r0,$L0,$h,$v,$rot,$nontaylor,$trasmittivity,$tA,$tCut],...],$D,$Ds,$outscale) ;

Predefined layers:

$Seeing::test_layers = [
			["layer1",32,60,30,0.1,15,4000,10,0],
			["layer2",16,60,30,0.18,15,8000,10,1]
		       ];

$Seeing::test_layers_ext = [
			["layer1",32,60,30,0.1,15,4000,10,0,0,"tra1",0.03,10],
			["layer2",16,60,30,0.18,15,8000,10,1,0,"tra2",0.03,10]
		       ];


$Seeing::test_layers1 = [
			 ["layer1",32,60,30,0.1,15,4000,10,0],
			 ["layer2",32,60,30,0.18,15,8000,10,1]
			];

$Seeing::test_layers2 = [
			 ["layer1",32,60,30,0.1,10,4000,10,0],
			 ["layer2",32,60,30,0.18,10,8000,10,1]
			];

# max is circa 13s for greenwood

$Seeing::test_greenwood = [
			["gl1",32,100,50,0.084,15,200,5,0],
			["gl2",16,100,200,0.310,15,2000,10,1],
			["gl3",16,200,100,0.536,15,10000,15,0],
			["gl4",8,200,400,4.588,15,18000,20,1]
		       ];

$Seeing::test_greenwood_tra = [
			["gl1",32,100,50,0.084,15,200,5,0,0,"gl-tra1",0.03,10],
			["gl2",16,100,200,0.310,15,2000,10,1,0,"gl-tra2",0.03,10],
			["gl3",16,200,100,0.536,15,10000,15,0,0,"gl-tra3",0.03,10],
			["gl4",8,200,400,4.588,15,18000,20,1,0,"gl-tra4",0.03,10]
		       ];

$Seeing::test_greenwood30 = [
			["gl1",32,100,50,0.084,30,200,5,0],
			["gl2",16,100,200,0.310,30,2000,10,1],
			["gl3",16,200,100,0.536,30,10000,15,0],
			["gl4",8,200,400,4.588,30,18000,20,1]
		       ];

$Seeing::test_greenwood_tra30 = [
			["gl1",32,100,50,0.084,30,200,5,0,0,"gl-tra1",0.03,10],
			["gl2",16,100,200,0.310,30,2000,10,1,0,"gl-tra2",0.03,10],
			["gl3",16,200,100,0.536,30,10000,15,0,0,"gl-tra3",0.03,10],
			["gl4",8,200,400,4.588,30,18000,20,1,0,"gl-tra4",0.03,10]
		       ];

# max is circa 9s for slcn

$Seeing::test_slcn = [
		      ["sl1",32,100,50,0.157,15,200,5,0],
		      ["sl2",32,50,100,0.164,15,2000,10,1],
		      ["sl3",16,200,100,0.846,15,10000,15,0],
		      ["sl4",16,100,200,1.517,15,18000,20,1]
		     ];


$Seeing::test_slcn_tra = [
			  ["sl1",32,100,50,0.157,15,200,5,0,0,"sl-tra1",0.03,10],
			  ["sl2",32,50,100,0.164,15,2000,10,1,0,"sl-tra2",0.03,10],
			  ["sl3",16,200,100,0.846,15,10000,15,0,0,"sl-tra3",0.03,10],
			  ["sl4",16,100,200,1.517,15,18000,20,1,0,"sl-tra4",0.03,10]
			 ];

$Seeing::test_slcn30 = [
		      ["sl1",32,100,50,0.157,30,200,5,0],
		      ["sl2",32,50,100,0.164,30,2000,10,1],
		      ["sl3",16,200,100,0.846,30,10000,15,0],
		      ["sl4",16,100,200,1.517,30,18000,20,1]
		     ];


$Seeing::test_slcn_tra30 = [
			  ["sl1",32,100,50,0.157,30,200,5,0,0,"sl-tra1",0.03,10],
			  ["sl2",32,50,100,0.164,30,2000,10,1,0,"sl-tra2",0.03,10],
			  ["sl3",16,200,100,0.846,30,10000,15,0,0,"sl-tra3",0.03,10],
			  ["sl4",16,100,200,1.517,30,18000,20,1,0,"sl-tra4",0.03,10]
			 ];

$seeing->redoLayers($force)
$gpf_image = $seeing->getGPF($time, $arcsec_aniso, $ground_displacement)
$psf_image = $seeing->getPSF($time, $arcsec_aniso, $ground_displacement)
$psf_no_seeing = $seeing->getCleanPSF()
$image_scale_arcsec = $seeing->getArcSec
$image_scale_rad = $seeing->getRad

SI.pm

$si = SI->new($support,$total_additive_noise);
$si->initKT([[$dx,$dy],...]);
$si->initBS([[$dx,$dy],...]);
$si->add($image, $optional_fft);
$si->addKT($image);
$si->addBS($image);
$spectrum = $si->get();
$variance = $si->getVar();
$snr = $si->getSNR();
$crosspectrum = $si->getKT(i);
$bispectrum = $si->getBS(i);
($image, $phase, $modul, $snr) = restoreKT($mod, $reference_mod, $snr, $dxrp, $dxrn, $dyrp, $dyrn, $max_frequency, $alpha, $threshold);
($image, $phase, $modul, $snr) = restoreKT($mod, $reference_mod, $snr, $dxrp, $dxrn, $dyrp, $dyrn, $max_frequency, $alpha, $threshold);

Teo.pm

Miscellaneous theoretical functions.

Utils.pm

Miscellaneous utility functions.