Analyze frequencies and generate audio

Plot a spectrum

Select audio and choose Analyze → Plot Spectrum…. Choose an FFT size from 256 to 65,536 samples, a Hann, Hamming, Blackman, or rectangular window, and a logarithmic or linear frequency axis. Point at the graph or type a frequency to read its level, jump between detected peaks, or export every bin as CSV.

Larger FFT sizes separate closer frequencies; smaller ones follow changes over time. Analysis reads the edited audio, including gain and fades, never changes the project, and can be cancelled on long selections. Levels are relative to full scale, not loudness measurements.

Generate at the cursor or replace a selection

Use Generate → Tone… for sine, square, sawtooth, or triangle waves at a chosen frequency and amplitude; Noise… for white, pink, or brown noise; and Silence… for quiet samples. Each dialog asks for a duration, which starts at the selection length or the last value used.

At a cursor, generated audio is inserted as a new region and later audio moves along. With a time selection, it replaces that span. Generation follows the project’s sample rate and channel count, remembers successful settings, and supports cancellation and one-step undo.

Use generated audio in an edit

Add a short tone as a cue, build a test signal, or insert silence to make room for an echo or reverb tail. Begin playback at a comfortable volume. To find rhythmic edit positions in existing audio, use Analyze → Detect Beats.