+/** This Spectrogram shows the spectral power distribution of incoming audio samples
+ over time. See http://en.wikipedia.org/wiki/Spectrogram.
+
+ The Spectrogram makes use of two caches:
+ * A cached image where only the most recent line needs to be appended instead of
+ having to recalculate the whole image. A typical speedup factor is 10x.
+ * A FFT cache storing a history of previous spectral power distributions (i.e.
+ the Fourier-transformed audio signals). This is used if the user adjusts parameters
+ like the maximum frequency to display or minimum/maximum signal strength in dB.
+ All required information is preserved in the FFT history, which would not be the
+ case for an image (consider re-sizing the widget to 100x100 px and then back to
+ 800x400 px -- lost is lost).
+*/
+