1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
|
function [spd_out] = SplineSpd(wls_in, spd_in, wls_out, extend)
% [spd_out] = SplineSpd(wls_in, spd_in, wls_out, [extend])
%
% Convert the wavelength representation of a spectral power distribution.
% Takes change of deltaLambda into account to keep matrix computations
% consistent across wavelength samplings.
%
% That is, the splining handles conversion of power units according to to
% the wavelength sampling deltaWl. Power is assumed to be in units of power
% per wlband on input, and this convention is maintained on output. This
% convention, used widely in PTB, means that you can omit the
% multiplication by deltaWl when numerically integrating over wavelength,
% but is a little counterintuitive.
%
% Note that splining to 1 nm spacing puts the units into power per nm, as
% at 1 nm spacing the two conventions are the same.
%
% Handling of out of range values:
% extend == 0: Cubic spline, extends with zeros [default]
% extend == 1: Cubic spline, extends with last value in that direction
% extend == 2: Linear interpolation, linear extrapolation
%
% spd_in may have multiple columns, in which case spd_out does as well.
%
% wls_in and wls_out may be specified as a column vector of
% wavelengths or as a [start delta n] description.
%
% If wls_out is passed as a vector of wavelengths with just one sample, we don't
% know what the wavelength sampling is, and we can't do the conversion of
% power per wavelength band. This condition is checked for and an error is
% thrown. The fix is to pass the wavelengths as an S vector
% S = [theWavelength wavelengthBandWidth 1].
% This forces an explicit value for the wavelength band width.
%
% 5/6/98 dhb Change normalization method so that sum is constant.
% This is a little closer to the desired result for
% functions with big derivatives.
% 12/7/98 dhb Remove 5/6/98 change, as it produces the wrong power
% when you spline across different wavelength regions.
% 7/26/03 dhb Add extend argument and pass to SplineRaw.
% 8/13/11 dhb Update comment to reflect changes in SplineRaw.
% 5/10/12 dhb Small comment fix
% 1/15/18 dhb Can't believe I wrote this more than 20 years ago!
% dhb Put in error check for single wavelength value pass.
if (nargin < 4)
extend = [];
end
spd_raw = SplineRaw(wls_in,spd_in,wls_out,extend);
% Now take change in deltaLambda into account in power measure
if (length(wls_in(:)) == 1)
fprintf('Cannot determine delta lambda when only a single input wavelength is specified.\n');
fprintf('Call with S = [theWavelength wavelengthBandWidth 1] rather than just by\n');
fprintf('passing a single wavelength, where wavelengthBandWidth is the width of\n');
fprintf('the wavelength band for which power is specified.\n\n');
error('Change single passed input wavelength to S format');
end
S_in = MakeItS(wls_in);
if (length(wls_out(:)) == 1)
fprintf('Cannot determine delta lambda when only a single output wavelength is specified.\n');
fprintf('Call with S = [theWavelength wavelengthBandWidth 1] rather than just by\n');
fprintf('passing a single wavelength, where wavelengthBandWidth is the width of\n');
fprintf('the wavelength band for which power is specified.\n\n');
error('Change single passed output wavelength to S format');
end
S_out= MakeItS(wls_out);
convertPower = S_out(2)/S_in(2);
spd_out = convertPower*spd_raw;
|
