Online Spectrum Analyzer

Understanding Spectrum Analysis for Audio

Spectrum analysis is the foundational measurement in professional audio. Every equalization decision, feedback identification, and room acoustic assessment begins with understanding the frequency content of the signal. The Fast Fourier Transform (FFT) algorithm, first described by Cooley and Tukey in 1965, converts time-domain audio samples into frequency-domain magnitude and phase components in O(N log N) operations.

FFT Resolution and the Uncertainty Principle

Frequency resolution equals the sample rate divided by the FFT size. At 48 kHz with an FFT size of 4096, each bin spans 11.7 Hz. This means you cannot distinguish two tones closer than 11.7 Hz apart. Increasing the FFT to 16384 improves resolution to 2.93 Hz but increases the time window to 341 ms, making the display update slower. This trade-off between frequency and time resolution is fundamental to all spectral analysis per the Gabor limit.

Octave Band Analysis per IEC 61260-1

IEC 61260-1 defines the center frequencies, bandwidths, and filter response requirements for fractional octave band analysis. One-third octave bands divide the audible spectrum into 31 bands with a bandwidth factor of 1.2599 (the cube root of 2). SonaVyx implements octave analysis by energy-summing FFT bins within each band, achieving Class 1 accuracy when paired with a calibrated measurement microphone. The standard specifies minimum attenuation of 0.3 dB at band edges and 2.3 dB at adjacent band center frequencies.

Window Functions and Spectral Leakage

Applying a window function before the FFT reduces spectral leakage, the artificial spreading of energy from one bin into adjacent bins. SonaVyx supports Hann (good general purpose), Hamming (slightly better sidelobe rejection), Blackman-Harris (excellent dynamic range at the cost of frequency resolution), and flat-top (accurate amplitude measurement per ISO 18431-2). For most audio analysis work, Blackman-Harris provides the best combination of dynamic range and frequency selectivity.

Practical Applications in Live Sound

Sound engineers use spectrum analyzers at every stage of system setup. During line check, the RTA confirms that each input channel has the expected frequency content. During system equalization, pink noise through the PA reveals room modes, boundary effects, and coverage irregularities. During the show, real-time spectrum monitoring helps the mix engineer maintain consistent tonal balance as the room fills with audience and ambient temperature changes affect air absorption at high frequencies.