The Analyzer Settings determine the frequency and time resolution and other aspects of the spectrum, spectrogram, and pitch analysis. In particular, the Time Resolution has a large influence on the kind of information that the analyzer will extract from a recording.
The Sampling Rate determines the number of measurements (samples) per second recorded from the input source. You can measure frequencies of up to half the sampling rate. For example, if the sampling rate is 11025 samples per second, then the analyzer can measure frequencies up to around 5500Hz. Standard music CDs have a sampling rate of 44100Hz. This should be a reasonable setting for most practical purposes, as it gives a good balance between frequency resolution and sound quality (however, see the note below).
The sampling rate can be changed on the Recording Settings page.
VoceVista Video uses two separate mathematical methods to analyze sound. One method is the Fast Fourier Transform (FFT), which calculates the Spectrum. This gives the intensity of the individual frequency components of a sound. The other method is the detection of the fundamental pitch, which is completely separate from the FFT. The two methods are both dependent on the sampling rate of a recording. However, while lowering the sampling rate can increase the accuracy of the FFT in same circumstances, the pitch detection works best with sampling rates of 44100Hz or higher, and the EGG works best with 48000Hz.
The frequency resolution is the smallest difference between two frequencies that the analyzer can distinguish. Internally, this setting is stored as the size of the Fast Fourier Transform (FFT), which is the number of points that are computed for each update.
A higher FFT Size gives you more accuracy and shows more detail in the spectrum and spectrogram, but it also requires more processing power and may slow down your computer. In general, you should choose the highest setting that still gives you acceptable performance when moving the range slider on the Timeline.
This setting determines whether the Analyzer should be more accurate in the frequency or in the time domain. In other words, are you more interested in measuring the exact pitch, or in measuring the melody (the variations of the pitch over time)? Fewer updates per second will increase the accuracy of the pitch display but hide the melody, while more updates per second will cause the analyzer to show the melody more clearly, but with less accuracy in the pitch.
The Spectrum on the Short Term View will be averaged over the given time period to avoid flicker and make it more stable. With a smaller duration, it will be more responsive to sudden changes, but flicker more, while a longer value will make it appear sluggish but more stable.
This is similar to the Time Resolution for the FFT, but is used by the algorithm for pitch detection, which is separate. The Time Resolution value determines the length of a recording segment that the pitch algorithm considers to find its fundamental pitch. Changing this value is for advanced users only, as the default value works best in most situations.
The current pitch will be averaged over this period of time during recording and playback to make it more stable and avoid flickering.
The pitch detection module will ignore frequencies below this value, which can avoid false detections when the recording has low-frequency noise.
Due to the way the FFT method works, its input needs to be filtered through a window function to avoid incorrect results. A discussion of this can be found at en.wikipedia.org/wiki/Window_function.
For VoceVista Video the default window function Blackman-Nuttall works very well and gives a useable dynamic range of about 100 decibels. It should not be changed unless you understand windows functions and their respective properties and trade-offs.