Jesse Fox - HW#5 for Music 220a

Recreation of the THX "Deep Note"

Since its introduction to moviegoers in 1983, the THX "Deep Note" (heard here) has become well known as a sure indication of a good movie theater sound system. It is also the first sound to be trademarked. More relevant background information can be found here.

In this homework assignment, I set out to program my own deep note using the scheme programming language. The general idea behind the deep note is a sound produced by hundreds of oscillators, intially starting out at random frequencies and simultaneously moving to a new pitch while increasing in intensity until settling on a frequency consonant with its neighbors. In the original deep note, the ending chord is approximately an F Major chord (this can be verified at a piano, or through spectral peak analysis of fundamental frequencies).

The results of my attempt to synthesize a deep note can be found here, and here is my code.

While somewhat successfull, my sound is still quite different from the original. I think my synthesized version is perhaps too "sterile", and some built-in variance or de-tuning between frequencies might be useful. Furthermore, many of the parameters I used in creating this file were derived empirically, mostly from observations of sonogram data from the original sound. For example, I concluded that the original deep note has a beginning frequency range of about 0-4kHz, and its frequencies seem to ramp in a non-linear fashion. As a result, I chose a frequency envelope for my scheme code based upon a hyperbolic tangent curve, and I chose a range of 0-4kHz for my starting frequencies and 0~25kHz for my ending frequencies. I also specified an amplitude envelope and a scaling factor of 1/f, which decreases the prominence of higher frequencies. This last step is useful since our ears naturally "hear" (perceive) higher frequencies to be louder in volume. Unfortunately, I think this also has the adverse effect of making ultra-low frequencies too loud, and they end up dominating the sound. This is easily fixed in my code, and perhaps I will post a revised file in a day or two.