Ask any interpreter and she or he would tell you that their ideal way of working would be to parameterise the algorithms and get immediate response, which would guide their analysis.
What's the value?
- Interactive, GPU accelerated attributes allow users to let the data guide them
- Interpreters don't need to wait for what might possibly be the wrong result and start over
Applications that rely on HueSpace for GPU acceleration allow their users to do exactly that : Spend time on their analysis, not on creating outputs in batches. Spectral decomposition has many uses. One of these is to infer geology - such as channels or fans. Another is to detect thinner beds than the available seismic resolution. Regardless of use, physics dictate that frequency response change with depth, so instead of guessing, let the visual feedback drive the process. Users can change frequency bands on the fly, and immediately see the response on time slices or stratal slabs, to name two examples.