Technical Reference Documentation

2.5 Recording Length and Sample Rate

Recording Length Criteria

It should be selected considering the two-way time (TWT) of the deepest horizon of interest and the migration aperture time to collapse diffractions generated by the deepest formation. Old interpreted seismic sections could also help to define this parameter.

Frequency Bandwidth & Sampling

In seismic acquisition data for oil and gas exploration, the frequency range recovered from the subsurface is typically between 10 and 100 Hz. In some cases, it could recover a wider frequency range (e.g., 6-120 Hz) but not much more than that, especially if it is used accelerometer seismic sensors (such as Sercel-DSU).

A sample rate of 2 ms can get frequencies up to 250 Hz without aliasing (Nyquist Theorem). Actually, all anti-alias filters in the recording instruments are 0.8 Fnyq. That means, at 2ms sample rate, up to 200 Hz frequencies could be recorded. More than enough!

Constraints on 1 ms Sampling

Avoid using lower sample rates (e.g., 1 ms – up to 400 HZ) for several reasons. Three of them include:

  • Unnecessary size of seismic data will be generated because you will duplicate the number of samples per seismic trace.
  • Internal noise of the Analog to Digital Converter (ADC) will be increased, reducing its dynamic range.
  • Spurious frequency noises from geophone sensors could be recorded, such as Sercel SG-10 geophone Spurious frequencies beyond 240Hz.

Geological Requirements Integration

The recording length and the sample rate are related to each other according to the geological requirements to be imaged. A long recording length (e.g., 8000 ms) and a too-small sample rate (e.g., 1 ms) make no sense and could affect the performance of a recording instrument.