Sign up for ProSoundWeb newsletters
Subscribe today!

Speech Intelligibility: Gathering Additional Pieces Of The Audio Puzzle
+- Print Email Share RSS RSS

live sound
Figure 1 - For speech to be intelligible, the direct sound field level must exceed the level of the reverberation and/or noise at the listener location.
Related Tags
  Acoustic Measurement, Pat Brown

Additional Skills
In rooms that are reverberant, or excessively “live,” achieving acceptable SI requires a special skill set.

These include:

1. An understanding of the variables that determine the direct-to-reverberant energy ratio at a listener position. These can be understood (in an ideal sense) through the Hopkins-Stryker equation.

2. An understanding of what constitutes a “reverberant” sound field, and how a reverberant space differs in principle from a “live” space.

3. An understanding of the role of loudspeaker directivity and placement with regard to the establishment of the direct-to-reverberant sound energy ratio. I had always thought that the loudspeaker’s sensitivity and power handling were its most important specs. Regarding SI it’s more important to consider its directivity index (DI) and directivity factor (Q)—ratings I had always ignored on the spec sheet.

4. An understanding of what matters to the human hearing system with regard to SI. It’s not just about “sound quality.” Ironically, a sound system can be intelligible and not sound good. The cell phone is a notable example.

5. An understanding of how to quantify SI through measurement. While we don’t have meters that measure “music quality,” we do have meters that measure SI.

6. An understanding of how to estimate SI at the drawing board, including what room acoustics computer models can and can’t tell you about SI.

7.  Your customers cannot purchase SI at a big box store or on-line. They need YOU, not just the gear that you sell. Costco will never sell SI.

It should be noted that one can have years of experience “doing sound” and not develop competence in these areas. I was proof of that.

Figure 2 - The Hopkins-Stryker relationships allow the direct and reverberant sound field levels to be estimated at the drawing board. Credit: SynAudCon (click to enlarge)

Expanding your skill set can give you a bright future as an SI professional. Why bother?

1. There is a LOT of competition for music playback systems, resulting in low margins and the risk of being lost in the crowd. It is not likely that your competitors understand how to design systems that are intelligible, and fix systems that are not.

2. Music is subjective by nature, with no absolute reference or performance benchmark. There exists no objective test or measurement to prove that a system “sounds good.” This can make it difficult to resolve customer complaints. I was once told that a system I designed lacked “warmth.” I was tempted to set the loudspeakers on fire.

3. SI can be assessed by listening tests, PC-based measurements, and even hand-held measurements. SI can be scored to quantify how one loudspeaker system compares to another.

4. There is a basis for declaring a speech system “finished,” where it seems that music systems are constantly needing another tweak, effect or the latest wiz-bang processor.

5. There are codes and (pending) laws requiring that emergency evacuation systems be intelligible – not so for music systems. The architect of a space may have to capitulate to the sound system designer if adequate SI cannot be accomplished due to the room’s background noise or acoustics.

6. The principles and practices that govern SI are timeless. They have not changed for a century, and will not change in the future. Signal processing can only remedy some of the causes of poor intelligibility.

Music systems are like art galleries. Speech systems are like road signs. One evokes an emotion, and the other conveys information. It is important to make the distinction.


Commenting is not available in this weblog entry.





Sponsored Links