Why do the DSL-IO targets for speech and pure tone differ on the Unity system?
The targets for speech and pure tones are different due to the differences in the way that broad band and narrow band signals are detected/processed. Pure tone targets will tend to relate to the peaks of speech and hence some quieter components may not be audible if these targets are used. In general, speech targets should be used and a speech like (preferably modulated) signal should be used for verification.
