Hi David The analysis is based on Least-squares. To this purpose, the audio is resamples to 1000 Hz ... 10000 Hz. The higher the precision, the longer it takes. Nevertheless, there are some input sounds that do not "fit" into a rougher grid like 1000 Hz. In other words, the original audio could simply be "overlooked" during analysis. What we could do: - Round the result to 1 ms - Remove the accuracy slider - Start the analysis at a low sample rate and increase it until two or three times the same result is found. - otherwise an error is raised. We are having quite a lengthy discussion on the forum regarding this tool. Gale Andrews is particularly interested in how the tool can be made "dumb-proof". For instance in the case that an odd number of tracks is selected. However, Nyquist hasn't control over the actual number of selected tracks, thus not much can be done against such a case. Ps. #Steve: You should probably start a separate post regarding an accessible recording device, perhaps a topic in the Audacity Forum as well. I doubt that the Olympus LS100 is accessible, I've at least nothing read about voice-guidance. That's usually only available in voice recorders such as the DM-4. I personally use the milestone 512. It isn't a full qualified audio recorder though. The audacity4blind web site is at //www.freelists.org/webpage/audacity4blind Subscribe and unsubscribe information, message archives, Audacity keyboard commands, and more... To unsubscribe from audacity4blind, send an email to audacity4blind-request@xxxxxxxxxxxxx with subject line unsubscribe