On Friday Sept 13, 2013, Ray Dolby passed away. With the passing of Dr. Bose several months ago, the year 2013 was not among the best for the audio industry.
I met Ray Dolby at an Audio Engineering Conference (AES) in Los Angeles about 20 years ago and it was my least-cool life moment.
After being introduced, I proceeded to explain to him exactly how his system worked with such vigor and excitement that people gradually started to move away from us. Fortunately for me, Ray was trapped in a corner and could not withdraw any further. This was perfect; I had time to explain in detail every aspect of his system. I don’t know if Ray Dolby had “groupies” but I was one of his more ferocious, at least on that day.
For years, in a 3rd year course, I had taught (and still do) how the Dolby system reduced tape hiss and improved the signal to noise ratio (SNR) in the higher frequency region in acoustic phonetics at the University of Toronto. Meeting Ray Dolby was like meeting Gunnar Fant or Peter Ladefoged – both giants in the field of acoustics and speech production.
For those who really know me, as a Canadian, I tend to be understated. I have made ear monitors for the Stones, U2, Rod Stewart, and I even see Elvis Presley from time to time in my Musicians’ Clinic. I never lose my cool with these folks, but Ray Dolby was a notch above Blue Suede Shoes.
Now that I have you hooked, here is how the Dolby system worked.
- Tape media have a noise floor that is frequency independent. The noise level is as high for the lower frequencies as it is for the more treble notes.
- When speech or music is recorded onto a tape, since most of the energy is in the lower frequency region, there is a good SNR for the lower frequencies, but alas a poorer one for the higher frequency region. Because of this, the higher frequency tape noise is audible- the sibilants are not sufficient to block it out. That is, despite the tape noise being the same for all frequency regions, it is only audible in the higher frequency ranges.
- Now here is Ray Dolby’s genius – knowing that there would be a poor high frequency SNR, Mr. Dolby pre-emphasized the higher frequency region of the music or speech BEFORE placing it on the tape. This was his real genius!
- The pre-emphasized music or speech was then recorded onto the tape but now with a much better SNR.
- Before the playback, this pre-emphasized high frequency speech or music was undone (i.e., a slight high cut filter). This brought back the speech or music spectrum to the initial state but the noise floor in the higher frequency region (being filtered out) was now inaudible.
- The result is that the initial spectrum was identical to the final spectrum so it was ideal for music and for speech analysis. One could make a recording onto tape with Dolby and would not alter the elements of the speech at all- just get rid of some audible his.
I have had some phonetics professors over the years who have sworn by the hiss. When eliciting speech samples from various languages they always felt more at home, and more comfortable, doing a transcription task if there was that constant background hiss. But, other than these rogue phonetic professors, Dolby noise reduction had no downside.
As the recording media became more advanced – CD, MP3 files, and .wav files- this initial innovation was no longer necessary. It was a direct response to the inherent noise floor found on tapes. Yet Dolby as a company continued to innovate.
After more than 20 years I am sure that Mr. Dolby’s colleagues forgot about that incident and I am sure that the restraining order against me will have been forgotten as well…