Noise is an inherent problem in seismic. It is often necessary to attenuate noise in order to proceed to more advanced stages of interpretation. The dilemma is, however, what type of noise attenuation filter to use at what point in the data stream, and is not an easy one to solve. Generally, interpreters are forced to apply noise filters with certain parameters, but cannot assess the impact this may have further down the processing chain. The loss of time can be substantial, in addition to generating large amounts of data over and over again.
The functionality to analyze which type of filter is the most suitable according to the type of data is very extensive in Cegal Blueback Avary (Avary). The user has rich visualization capabilities at hand, both before and after each operation. This allows the user for example to evaluate whether the filter applied is too strong or not strong enough to reduce noise.
Fig. 1: Right: Original input with computed Timedip. Center: Denoised section. Left: Difference to estimate the amount of filtered noise.
Another major problem is being able to decide when enough is enough, as there is a risk of filtering out not only noise but also valuable and subtle information. This is where Avary enables a highly transparent workflow, offering not only state-of-the-art filters but also the possibility to chain the different processes. Working in an almost virtual way, the interpreter can optimize filter parameters in the denoise process and evaluate the impact of these operations on processes much further down the line, almost immediately.
Fig. 2: Changing a parameter early in the process and estimating its impact further down.