While the problems seen above prevent the use of the singular value decomposition technique as a general-purpose compression method, there is still a place for it in the seismic industry. Some derived seismic images, such as velocity fields and velocity analyses, do not contain the high frequency detail seen in the reflection data and therefore will compress well with this technique. Also, for low resolution seismic displays, small details are unimportant and may be ignored. Although these displays are typically already reduced from the original sizes, they might be further compressed using the singular value decomposition.
Reflection seismic data cannot be reconstructed accurately from compressed forms using the singular value decomposition because dipping events and high frequency detail map into the small singular values, violating the assumption that these small singular values can be ignored in reconstructing the compressed image. While other applications of singular value decomposition compression exist in the seismic industry, this technique fails to compress the largest amount of data stored, the seismic reflection data.