I am working on a graduate research thesis that used various sound bites and I have to visually show their frequency analysis (frequency on x-axis and dB on y-axis). I did this by going to analyze > plot spectrum. The problem is- the frequency on the x-axis needs to be consistent for each sound to be able to compare them. Some sounds go to around 5000 Hz, others to 4000 Hz, and so on.
Is there a way to specify the frequency window so each sound will have the same x-axis, for the purpose of visual comparison? I see when you move the size of the window you can change the axis slightly, but not to the extent I need to make them all the same. If not, does anyone know of a program with this capability?
That’s a common complaint about Analyze > Plot Spectrum. It keeps changing to display the most useful information possible (whether you want it to or not). This makes comparisons a nightmare.
A while back, I suggested putting a single, known, loud tone in your show in addition to the work. Spectrum will try to fit that single tone into a rational display leaving the rest of the display for your work, usually the same time after time. Another possibility is two tones one very low and one very high with the work in the middle.
I don’t know a good, graceful way out of this.
You can change Audacity to display Frequency versus Time on the timeline. Drop-down menu on the left > Spectrogram.
The X-axis (frequency) is determined by the “sample rate” of the selected track.
To change the sample rate of a track, select the track, then “Tracks menu > Resample” and select the required sample rate.
Note that the highest frequency possible for any given sample rate, is half the sample rate. Thus, if you have a track with a sample rate of 22050 Hz, and it contains frequencies up to 10000 Hz, if you resample it to 8000 Hz, then audio frequencies above 4000 Hz will be lost.
For serious analysis work, you might do better to export the frequency data from Plot Spectrum, and use a better graphing application to draw the graph (such as “GnuPlot” or even “Microsoft Excel”)