Plots

All images are spectrograms of the sound file, with formant frequencies indicated using points. Filled points indicate observed formant frequencies, hollow points indicate predicted formant frequencies using the linear models. When only a single track is visible that means the model was able to perfectly predict formants. Single image files show one analysis, usually the winning or final analysis. Comparison image files compare all analyses ofr a single sound file in a single image. Analyses are ordered top to bottom, left to right in number and increasing cutoff frequency.

For example, when Fast Track is run on the sound above we can have Fast Track select the best analysis for us:

 

We can automatically generate an image comparing the alternate possible analyses: