next up previous print clean
Next: Acknowledgments Up: Vlad: Enhanced random noise Previous: Practical issues

Conclusions

The spurious coherent noise patterns passed by the two spatial prediction filtering methods are different enough that their coherence is destroyed when added together. A formal mathematical investigation on why this holds true in the vast majority of cases may yield useful insights on optimally choosing the parameters for the two methods so that the final signal-to-noise ratio is maximized. However, even on a heuristic basis, Enhanced Random Noise Attenuation works. It is simple, fast, intuitive, and does not require any coding if f-x and t-x decons are already implemented. Without being more unstable than either of these techniques, it builds on both of them in order to produce a better result.
next up previous print clean
Next: Acknowledgments Up: Vlad: Enhanced random noise Previous: Practical issues
Stanford Exploration Project
7/8/2003