next up previous print clean
Next: Helix low-cut filter Up: Quantitative effect of the Previous: Quantitative effect of the

Helix derivative filter

For the helix derivative filter, the response is nearly a linear function of $\vert\bold k\vert = \sqrt{k_x^2+k_y^2}$ and does not have the ``cut-off frequency''. The main feature here is the zero-frequency response. Figure 8 shows the zero-frequency response R0 as the function of filter length when $\rho=0$. na is the half size of the filter. I find that R0 decreases as na increases and $R_0 \propto n_a^{-1}$. The empirical relationship between R0 and na is  
 \begin{displaymath}
 R_0 \approx \frac{0.44}{n_a}\end{displaymath} (7)
R0 is the sum of the helix filter's coefficients, so when $\rho \neq 0$,according to Equation (3), the zero-frequency response is  
 \begin{displaymath}
 R_0 \approx \frac{0.44}{n_a}(1-\rho)\end{displaymath} (8)

 
drv-r0f-ar0
drv-r0f-ar0
Figure 8
The zero-frequency response of helix derivative filter when $\rho=0$. The approximate zero-frequency response is $\frac{0.44}{n_a}$.
view burn build edit restore

The zero-frequency response of the enhanced helix derivative filter is controlled by both na and $\rho$, I can compute the approximate value of R0 using Equation (8).


next up previous print clean
Next: Helix low-cut filter Up: Quantitative effect of the Previous: Quantitative effect of the
Stanford Exploration Project
4/20/1999