Taylor Sappe Music Production & Instruction

Audio Spectrum Analyzer

One of the tools in a recording studio that helps to get that professional sound is a visual aid called the “Audio Spectrum Analyzer”, sometimes referred to as a “Realtime Analyzer.” The analyzer displays frequencies of the audio spectrum in bar graph form, showing the volume level of each frequency band in a musical recording.

 

 Like the graphic equalizer, the spectrum is divided into frequency bands, and the more bands the spectrum is divided into, the more expensive the analyzer. The most common and inexpensive analyzers are hand held 10 band type. However, a 31 band or more is recommended for most applications, but costs about three times as much as a 10 band.

 

 The spectrum analyzer has 3 basic applications, all of which are used in combination with some type of equalizer.

 

In live sound reinforcement, it is used to detect the feedback frequencies. The graph shows the exact frequency that is feeding back, allowing you to use your equalizer to cut back on that frequency and remove the feedback. The feedback frequency is usually removed with a graphic equalizer, parametric equalizer or notch filter.

 

 

In both live sound reinforcement and in the studio, it is used in combination with a pink noise generator and a graphic equalizer (preferred 31 band or more) to tune the speaker system flat to the room.

 

 

A pink noise generator sends pink noise through the system. Pink noise is a swishing sound that maintains all of the frequencies of the audio spectrum with equal amounts of energy in each frequency. Peaks and valleys will show up on the display of the analyzer when first running this noise through your system because different objects in the room, and different densities in walls vibrate at different frequencies, so you speaker system is not producing a flat sound that would make it accurate. The graphic equalizer is used to smooth out these peaks and valleys by cutting or boosting the jagged frequencies to create a totally flat response for accuracy. Your analyzer and graphic equalizer should be of the same number of bands for the best possible set up. The more bands you have, the more accurate your setup. 31 bands is the minimum recommended for this application.

 

 

In the studio, the analyzer is used for equalizing instruments and for mixing. Both applications are for obtaining the best possible signal to noise ratio. The signal to noise ratio is the difference in volume levels between the noise floor (produced by your system or tape if you are using a tape recorder) and the average recorded signal. If the average recorded signal is low, you will hear some of the noise floor. The higher you can record the average signal, without distorting, the less of the noise floor you will hear. If your average is low and there are occasional peaks in one frequency band or more that are much higher than the average, the peaks will push the signal into distortion while the average signal remains low and the noise floor is still audible. The realtime analyzer will show you which frequencies have hotter peaks than they should, and you can then control those peaks with either an equalizer or a multi-band limiter (also called a dynamic equalizer).

 

 

For equalizing individual instruments before recording them, you can isolate the signal coming from that instrument by soloing it on your mixer or having the instrument play by itself. Having the musician play through the entire song will give you a reading across the entire range of frequencies that instrument will play in that song. You can then check to see if any frequencies are too strong or not strong enough at any point in the song, and use an equalizer to adjust the frequencies for a more equal peak reading.

 

 

For mixing, treat the entire mix as you would one instrument. Play it from beginning to end, checking the peaks and adjusting them accordingly. When all of your peaks are adjusted correctly you can increase the output level to just below the threshold of distortion for the hottest possible average signal and the best signal to noise ratio. In a good mix, every note of every instrument should be able to be heard clearly. This is one of several techniques used in combination with each other to accomplish this.

 

For complete private instruction, email me.