Agenda - 11-22-2004-c1 - Laserfiche WebLink

Browse Search

19 Attachment 3 Environmental Sound Page 4 is often desired The octaves aze often broken into one-third-octave bands, A finer resolution is provided by aFast-Fourier-Transform analysis of the sound, T1vs can provide a resolution ofone Hertz or less that can be related to a specific characteristic of a source. When measuring typical fluctuating environmental sounds over long periods, it is common to use average and statistical measures. One cannot simply write down severalmeasured sound levels and azithmetically average the values. The energy average of 60 dB and 40 dB will be approximately 57 dB, half the higher 60 dB level. The energy equivalent average ]eve] over a given time is called the "time-average sound level." It is also often called the "equivalent level," abbreviated LEQ, The averaging of sound levels can be complicated unless one has a special instrument that does it automatically. Average levels over a period due to individual events can be calculated by combining the "sound exposure level" (SEL) of the events, The SEL is a measure of the sound energy of the event, It is the level in decibels of a sound of the same energy lasting exactly one second. The sound level during a given period also can be character'~zed bythe levelexceeded a givenpercentage of the measurement period. For example, the levelexceeded 90% ofthe time is defined as the L90 value. It is a good indicator of the background sound level. The L50 indicates the level exceeded half the time, usually a little less than the time-average sound level. The higher sound levels are indicated by the levels exceeded a small percentage of the time, Many localnoise ordinances developed before the widespread availability of averaging meters were based on the L10, The L10 is usually a few decibels more than the time-average sound level for varying sounds, The L10 can be less than the time-average sound level if a very loud sound occurs less than 10% of the time. LOl indicates the highest levels likely to occur. The difference between L01 and L90 or L10 and L90 is a good indicator of variation in sound levels, A measurement period always must be specified for any statistical valnes such as LO1, L10 or L90 to have meaning, The number of samples measured should be at least 10 times the difference in decibels between the highest and lowest level.. Otherwise, different people measuring the same sound could get widely varying results. Ideally, with automated instrvrrients there should be at least two samples per second with slow meter response, and at least 16 samples per second with fast response. This gives a complete description of the sound variation. A special type of average sound level is used to describe the long-term acoustical climate of a location.. The Day-Night Average L.eve] is typically calculated over a yeaz, and has a 10 dBA night-time penalty added to all sound between 10:00 p.m. and 7:00 a,m. This is abbreviated DNL, or sometimes LDN.