Background Noise – How Much is OK?

A signal to noise ratio of 10dB. At least 15dB is required for good intelligibility. DSW to reverse field

A moderate level for a teacher’s voice is around 67dB at 1 meter (3feet).  However, as the sound waves spread, the level drops with distance. A well designed room that provides useful early sound reflections will in turn provide support to an instructors voice to a level of about 50dB at the farthest areas, which can cover more than half the room.  Without support, 50dB is not realistic at these farthest points.

To insure good speech intelligibility, we need to maintain an average differential between speech level and background level (called signal to noise ratio) of about 15dB, which means the A-weighted background noise level should not exceed about 35dB (this is the accepted background noise level per the ANSI Standard for typical classrooms, (volumes less 20,000 ft³)).  That is, 50dB (signal)  minus 35dB (noise) equals a signal to noise ratio (a relative sound level), of 15 dB.

Rather than boost the instructor’s voice through amplification (more on that later), we have to work to keep the background noise level inside the unoccupied classroom from exceeding 35dB. This is where systems detailing and construction techniques come into play.

A closer look at dBA levels: The background noise is reported as a single number (dBA), but acousticians understand  that the single number is actually a logarithmic  sum of the weighted raw decibel levels for each frequency band (tell your kids that at the breakfast table!). Humans are less sensitive to lower frequencies and the A-Weighting reflects this.

So what are C-Weighted levels? C-Weighted levels are a measure of the sound which does not attenuate low frequency sounds as the A-weighting does, and they are limited to values not more than 20 dB above those for the A-weighted levels. This limitation is shown in Table X above.

To summarize weightings, all people are very sensitive to the speech range frequencies (mid range and high frequencies), and as indicated above A-weighting emphasizes them over low frequencies while C-weighting does not.

Example: a signal to noise ratio of 10dB. At least 15dB is necessary for good intelligibility, but not sufficient of itself to guarantee it.

I think it is a good figure, could be revised, and with the caption it reinforces the concept.-DW

Fig. 3 An example of a uniform sounding 35 dBA signal.

Fig 3 correction: 27dBA, speech utilizes consists of predominately…

Fig. 4 Another example of a 35 dBA signal.[Consider replacing these two figure with just one which shows the typical spectrum of the human voice – this can include the very useful horizontal bars in Fig. 3

Advertisement