next up previous contents
Next: Function: corr_coef() Up: Binary Inspiral Search on Previous: Details of Normalization   Contents


Function: strain_spec()

void strain_spec(float flo, float srate, int n, float *adc_spec,
		 float *response, float *mean_pow_spec, float *twice_inv_noise)

This routine computes the strain spectrum $\alpha S_h(f)$ and twice the inverse noise $2/(\alpha S_h(f))$ from the ADC spectrum $S_{\scriptstyle\mathrm{ADC}}(f)$ and the strain response function $R(f)/\ell$. The arguments are:

flo: Input. The low frequency cutoff (Hz) for computing the strain spectrum and twice inverse noise. These are set to zero for frequencies below flo.
srate: Input. The sampling rate in Hertz.
n: Input. Sets the size of the following arrays.
adc_spec: Input. The vector adc_spec[0..n/2] of the ADC power spectrum $S_{\scriptstyle\mathrm{ADC}}(f)$.
response: Input. The vector response[0..n+1] of the detector strain response function $R(f)/\ell$. Here, $\ell$ is the armlength of the detector, so that response[] has dimensions of $\mbox{(ADC counts)}^{-1}$.
mean_pow_spec: Output. The vector mean_pow_spec[0..n/2] containing the strain power spectrum $\alpha S_h(f)$.
twice_inv_noise: Output. The vector twice_inv_noise[0..n/2] containing twice the inverse strain noise power spectrum $2/(\alpha S_h(f))$.
The strain power spectrum is $S_h(f)=\vert R(f)/\ell\vert^2\times S_{\scriptstyle\mathrm{ADC}}(f)$, and the normalization factor $\alpha=\texttt{n}\times\texttt{srate}$ is present for agreement with the output of the routine avg_inv_spec().

Author: Jolien Creighton, jolien@tapir.caltech.edu
Comments: Note that the strain power spectrum differs from $S_h(f)$ by the low frequency cutoff, and by the factor $\alpha$.


next up previous contents
Next: Function: corr_coef() Up: Binary Inspiral Search on Previous: Details of Normalization   Contents
Bruce Allen 2000-11-19