This tool uses a Discrete Fast Fourier Transform (DFFT) to separate the audio at the current selected position of the waveform into its frequency components. To use it, set the waveform cursor to the point in the audio you want to analyze and select Tools -> Frequency Analysis. In the window that opens, you should see one or two graphs displayed, known as FFT graphs. If the audio file you are analyzing is of mono format, there will be one blue graph shown. If the file is stereo, there will be one blue graph for the left channel and one pink graph for the right channel.
In the top right-hand hand corner are the frequency and decibel values of the point in the graph where the mouse cursor is currently located. The decibel values range from 0dB (loudest) at the top, down to -127dB (softest). The frequency range depends on the sample rate of the audio file, ranging from 0Hz on the left to half the sample rate of the audio at the right.
Two window types, Hanning and Hamming, are provided to apply FFT.
To see the FFT graph in more detail, click the Zoom In buttons on either the bottom of the window or along the right-side (this will zoom the graph horizontally or vertically respectively). While zoomed in, you can use the scroll bars along the bottom and right-hand sides to scroll horizontally or vertically.
To zoom the graph out, either click on the respective Zoom Out buttons, or click the button in the bottom-right corner to set the view back to full-scale.
An alternative to using the Zoom In and Zoom Out buttons is simply resizing the FFT graph window. To do this, move the mouse cursor to any edge or corner of the window, and hold the left-mouse button down and move your mouse as appropriate.
This tool calculates an FFT analysis over time (TFFT), and uses color to display the intensities of the spectral information. To use it, select an area of the audio waveform you would like to analyze and select Tools -> Temporal Frequency Analysis. In the window that opens, you should see a graph displayed, known as the TFFT graph. Time is represented along the horizontal axis and has a range the same as the region of the audio waveform you have selected. Frequency goes along the vertical axis, and goes from zero to half the sample rate of the audio waveform. The colors represent the decibel levels for a specific frequency at a specific point in time, with brighter colors meaning stronger intensities. The decibel values range from 0 (loudest) down to -127dB (softest). The values of time, frequency and decibels can be viewed in the status bar at the bottom of the TFFT window, and will depend on where your mouse cursor is currently located in the graph.
If you are performing an analysis on a stereo waveform, you will see the effect of both channels combined into the one graph.
To view spectral information in the 0 - 4000Hz range, click on the zoom button in the top right-hand corner of the graph. To view the graph at normal zoom, click the button directly below it.
The slider bar lets you change the brightness levels of the graph to either dim or highlight the lower intensity areas. Move the slider up to increase the brightness, and move it down to decrease the brightness.
If you feel the gridlines of the graph are obstructing your view of the analysis, then you can turn them off by toggling the button in the lower-right corner of the window ("Toggle the gridlines on or off").
You can also play the selected area of the audio waveform and watch the cursor move along both the audio and the TFFT graph. This will help you to relate what's going on in the audio to what is going on in the graph. You can also left-click your mouse on any point in either the audio or the TFFT graph to set the cursor at that point.
If you have perchance lost the selection area being analyzed in the audio waveform, you can get it back by clicking the appropriate button in the lower-right corner of the TFFT window ("Reselect this analysis region in the audio waveform"). Note that you cannot select regions in the TFFT graph - to get the graph to analyze a different part of the audio, you must select that part in the audio waveform itself and re-run the TFFT analysis.