The BPM tap toy is a tiny application to estimate the speed of a song by tapping a button in sync with the rhythm.

In contrast to many existing applications, this application uses a weighted moving average in estimation, and also gives a confidence interval so the user can see the accuracy of his tapping.

A fully automatic computation based on audio input is not done, since my experience with such algorithms is that they can't distinguish the main rhythm from sub- or super-rhythms and are just not reliable.

• Any key (except special keys) and mouse clicks will be counted.
• After a short pause, computation restarts automatically!
• The two symbols in the upper right corner continue to flash in the current speed, and are thus useful for validating the computation result.
• The think green line in the diagram is the curren estimation, the blue line are the raw values, and the thin green lines give the confidence interval.

Installation is not required, the application can be started right away (the .py file).

Requirements:

• Python (usually included with Linux, OS X)
• PyGTK (often included with Linux)

These need to be installed separately. This should also be easily possible for both Windows and OS X, but I can not offer you help with installation. Linux users most likely have everything they need already installed (Debian: apt-get install python-gtk2)