Abstract

Use of volume visualization has gone hand in hand with improvements in computer hardware. Desktop visualization workstations are now typically configured with 2, 4 or even 8 GB of RAM. The larger visualization systems are configured with up to 16GB of RAM. Today we are seeing 100GB (8-bit) sized data sets equivalent to areas greater than 10,000 km2.

However, effective volume visualization requires that the 3D data set be completely resident in computer RAM. This leaves us with the problem of how to visualize large 3D datasets? One solution has been to reduce volume size through decimation. Through decimation not all seismic information is used. Decimation results in a loss of resolution, which might result in key visualization targets being missed or misinterpreted.

New visualization methods using volume roaming allows navigation through entire un-decimated 80 to 100 GB 3D datasets by accessing bricks or blocks of data directly from disk and placing the data in cache memory for temporary viewing. Data need to be stored in an efficient format for rapid access by the volume roaming application. Volume roaming speed is comparable to volume visualization speed in terms of panning through and accessing seismic data sets but with the added benefit of immediate access to the entire un-decimated 3D data set. Areas of interest are identified during volume roaming and are subsequently sub-setted from the main volume so they can fit into computer RAM and be analyzed using un-decimated volume-based visualization techniques.