Our enjoyment of synthesized or recorded sound is greatly enhanced when we perceive the actual locations of the various musicians on stage. In a high quality stereo recording of a jazz trio, for example, you can tell that the drummer is perhaps located slightly to the left of center, the saxophonist is on stage right and the bass player is on stage left. If you have ever flipped on the "mono" (monaural) switch on your stereo amplifier, then you know that the resulting sound is comparatively unpleasant, especially when wearing headphones.
We live in a three-dimensional soundfield, and our ears continually experience slightly different versions of any given sound. These variations allow us to place (or localize) the sound source, even when we cannot see it.
In this module, learn about two localization cues called interaural intensity difference and interaural timing difference, abbreviated IID and ITD, respectively. Each cue relies on presenting a slightly different version of a sound to each ear.
Interaural Intensity Difference (IID)
The interaural intensity difference (IID) localization cue relies on the fact that an off-center source produces a higher intensity sound in one ear compared to the other. This technique is also called intensity panning, and is most effective when the listener is wearing headphones. When a multi-speaker arrangement is used in a room, a given sound propagates to both ears. Lower frequencies have longer acoustic wavelengths and are therefore less directional than higher frequency sounds. Consequently, the IID effect works better at higher frequencies above 1400 Hz (Dodge and Jerse, 1997).
The Figure 1 screencast video continues the discussion by deriving the equations needed to implement the IID effect using a two-speaker arrangement.
Interaural Timing Difference (IID)
The interaural timing difference (ITD) localization cue relies on the fact that sound waves from an off-center source arrives at one ear slightly after the other ear. We can perceive this slight difference in time down to about 20 microseconds (Dodge and Jerse, 1997), and this difference helps us to place the sound source either to the left or right. The ITD cue works best in the lower frequency range of 270 to 500 Hz (Dodge and Jerse, 1997).
The Figure 2 screencast video continues the discussion by deriving the equation needed to implement the ITD effect using a two-speaker arrangement.
Project: Implement the IID and ITD Localization Cues in LabVIEW
You can easily experiment with both the IID and ITD localization cues in LabVIEW. The cues are probably easier to perceive when you choose a speech signal for your source.
The ITD cue requires a delay or time shift between the two stereo channels. The delay line can be constructed as a digital filter with the transfer function shown in (Reference):
Stated as a difference equation, the filter is y(n)=x(n−N), which states that the output is the same as the input but delayed by N samples. The forward (or "b") coefficients are therefore zero for all but bN=1, and the reverse (or "a") coefficients are zero for all but a0=1.
The Figure 3 screencast video provides LabVIEW coding tips to implement the equations and to generate a stereo audio signal.
- Dodge, C., and T.A. Jerse, "Computer Music: Synthesis, Composition, and Performance," 2nd ed., Schirmer Books, 1997, ISBN 0-02-864682-7