Nonuniqueness in relating velocity to porosity in core and well-log data complicates interpretation of sonic and seismic measurements. One reason for this nonuniqueness in sandstones is clay (e.g., Han, 1986). Another reason is textural variability among samples. Dvorkin and Nur (1996) examine two relatively clay-free sandstone groups in the same porosity range, but whose velocities significantly differed (Figure 1a). By comparing the data with effective-medium theories, they interpret this velocity difference as resulting from the difference in the position of diagenetic cement. The explanation is that in the "fast" (Oseberg) rocks (contact) cement is located predominantly at the grain contacts, whereas in the "slow" (Troll) rocks (noncontact) cement is located predominantly away from these contacts.