The Technology

Reveal Audio Services uses specialized software that was developed just for the purposes of noise removal - reducing the noises that interfere with and mask speech signals.   This is not the kind of software you find in a music recording studio.   Music editing and audio restoration deals with the fine points of noise reduction using gentle curves and subtle processes to preserve as much of the musical tones as possible.  The standard music tools just don’t go far enough for forensic audio work.

The software I use is adaptive by design, meaning that as the interfering noises change, the software algorithm “learns” them and adapts in a matter of milliseconds. Interfering beeps and pulses can be diminished greatly.   60-cycle hum can be nearly eliminated.   Traffic “swooshes” can be minimized.

The range of settings available is more powerful as well.  For instance, a typical music equalizer may have 6 bands to work with, with a range of +/- 18 decibels.  For forensic audio enhancement work, I am using an equalizer with 4,096 bands and a range of +24 to -96 decibels.

I can even go further with Enhanced Spectral Detail Editing.  This is a syllable-by-syllable method of processing that allows me to not only hear the speech, but to visually see it and manually separate it from the surrounding noise, or identify the noise and reduce or remove it to better hear the masked speech.  I am not removing speech or adding speech. I am identifying speech so that it can be separated more accurately from the noise.

There are limits to the effectiveness of speech clarification and enhancement. The noise reduction process works exponentially better as the signal-to-noise relationship increases.  The closer the speech is to the noise floor, the less effective the process.  Removing too much noise will also remove the speech.  If the noise reduction process is too aggressive it will create artifacts that sound like speech themselves.  If the speech is below the noise floor, it just becomes more noise - there is nothing to recover.

This is not Hollywood CSI magic that appears at the touch of a button.  This is the real world and the process is not that simple.  It requires a different listening mind-set, a tremendous amount of audio knowledge and expert proficiency with the software to achieve results that range from an acceptable improvement to jaw-dropping clarity.