Interview with Ioan Allen

The Dolby Film Sound Revolution:

Looking Back and Looking Forward
with Audio Pioneer Ioan Allen

First of Two Parts

by Leslie Shatz

I originally became aware of Dolby as a teenager, through my passion for music. At that time, the word was synonymous with the highest standards of quality in music recording and reproduction, and Dolby was a company steeped in the culture of the music business. The



Allen at A.E.S. New York, 1979.

original process, invented by founder Ray Dolby, takes a sound signal and splits it into 4 frequency bands. During recording, the volume of low level signals in each band is increased while loud signals are left untouched. Then during playback, low level signals are reduced to their original levels, and the unwanted noise is reduced, as well. The Dolby process became essential for the rock and roll musicians I was listening to, who were experimenting with new 8-track, 16-track, and seemingly impossible, 24-track tape recorders, where the noise from each track, when added together, became a real detriment to quality. Dolby later found wide acceptance in consumer cassette decks, which because of the size of the tape and the slow speed at which it traveled over the heads had inherent noise problems, as well. Having conquered the music world, Dolby set their sights on film sound. Enter Ioan Allen.

A native of the UK, Ioan was educated at Dartmouth Royal Naval College and spent several years in artist management and record production before joining Dolby in 1969. As you’ll see in our interview, he was largely responsible for the origination and development of the Dolby Stereo film program, participated in the development of Dolby’s multitrack digital film sound technology and is now working on their efforts in the area of digital cinema and digital projection. He is a SMPTE Fellow and member of the Audio Engineering Society and the British Kinematographic Sound and Television Society. Past chairman of two key SMPTE committees covering audio recording and projection technology, he now serves as SMPTE’s Engineering Director for Motion Pictures. In that capacity, he is responsible for U.S. digital cinema standardization activities. He holds several patents, is the author of many journal articles and is an Adjunct Professor at USC. He received Scientific and Engineering Awards from the Motion Picture Academy in 1979 and 1987 and an Oscar in 1989. We spoke in his office in San Francisco.

Ioan Allen: I joined the company back in 1969, and by about 1970, most of my activities were involved with music studios. That coincided with the adoption of multi-track tape recorders, which were just coming into play in London, New York and Los Angeles. We succeeded in equipping every track of every 16-track machine with noise reduction. This led me to thinking about where we would go with the technology. There was a limit to how many 16-



Allen on the set of ‘The Dolby System’, a demonstration film the company made in England in 1975.

track machines would ever be out there. It seemed that the place where there were lots of generations — not many tracks, but lots of generations —was film sound. At that time, a film mix used many analog pre-mixes, which led to a very high noise build-up, because each new pre-mix was a copy [or generation], and each copy added noise. So it occurred to me that there was a market for Dolby noise reduction in film sound.

Was there a particular film where you applied Dolby to the editing and the mixing and finally...

It came in stages actually, but the first film that used Dolby noise reduction in every pre-mix was A Clockwork Orange, in 1971. But of course there was no equalization of the loudspeakers, and the prints that went out were not noise-reduction encoded. The prints were just straight Academy mono prints. And they suffered from all the problems of the Academy mono soundtrack. So even with our noise reduction technology in the creation phase, the prints still sounded bad.

[Much as the name Dolby became synonymous with the highest quality in sound, the mysterious Academy Curve became synonymous with the worst. It is a graph that describes how different sound frequencies were actually reproduced in the cinemas of the time. Ideally this curve would be as flat as possible, meaning that the greatest range of frequencies, from low to high, would be reproduced equally. However, the curve, which reflected acoustic measurements taken by members of the Motion Picture Academy in theaters throughout the US, was anything but flat, as shown in Figure 1. In practice, there were several different curves documented by the Academy, one for each common loudspeaker system in use at the time. Mixing stages at that time would simulate how a mix would be heard in a real theater “out there,” by putting a roll-off filter in their loudspeaker system, which was the equivalent to the sum of all of the degradations down the line.]

The Academy Curve was caused by a combination of things. In the projector, sound is played back by passing a beam of light through the release print’s track area via a slit and onto a photocell. The height of this slit causes some high frequencies to be lost. Then the highs were further reduced by the loudspeakers that were in use at the time, which were not very efficient at reproducing high frequencies. And a final slice of high frequencies were taken out by playing the sound through the screen, much like putting a blanket over the speakers. Traditionally, people always thought of the Academy Curve as just rolling off the high frequency. But in practice, the low frequencies are rolled off as well, due to loudspeaker inefficiency. So what you finished up with was a playback response that was little better than a phone line.

In the 1930s, power was extremely expensive. An amplifier with 15 watts of power was considered big stuff, because vacuum tubes were used then, and they were difficult to get and expensive to use. So as a result, the loudspeakers were made to be very efficient. But that efficiency came at the cost of fidelity, because an efficient loudspeaker is best only in the midrange. Of course, there was a fringe benefit. Because the high frequencies are reduced, you don’t hear the optical soundtrack noise up there, either.

Figure 1. Dating back to the advent of ‘talkies,’ the Academy Curve imposed severe limits on theatre sound playback. It was replaced in the late ’70s by the gentler X Curve, which brought "high fidelity" to most movie theatres for the first time.

So as speakers improved, why didn’t theater sound improve? Why were we still limiting quality by mixing through the Academy curve?

If all you do is eliminate the Academy curve, you’ll suddenly hear all the hiss and rumble inherent in the optical track. Dolby noise reduction made it possible to extend the highs and lows without the noise becoming so obvious. We replaced the Academy curve with the “X Curve” (See Figure 1), which is a minimal curve, practically no roll-off. And we extended the high frequency response with third-octave equalization, which we apply in our cinema processor installed in each theater. [Third-octave equalizers divide the 10-octave range of audible sound frequencies into a total of 30 bands, with a separate fader for each. This allows for very precise control over the timbre of sound reproduction, which is calibrated as part of the initial set-up for the theater.]

So now you were attacking on two fronts: speaker equalization, and noise reduction on the optical track itself. But then there was the third aspect, which was the introduction of stereo...

Yeah. At that time there was very little stereo in the theaters; it was only four-track magnetic. [Big “roadshow” pictures would sometimes add to their release with a few special, 35mm



Allen in a recent publicity photograph.

magnetic “foxhole” prints. These prints were coated with four narrow stripes of magnetic oxide around both sets of perfs, and then each print was individually recorded, or “sounded” in real time.] That was extremely expensive in the early 70s. Unless you lived in a big city with a big showcase theater, you never heard a stereo film. But we believed it was possible to get good quality stereo from an optical photographic soundtrack. If you split the track in two, you get perfect compatibility for mono theaters, in the same way that a stereo FM radio station is compatible with mono radios, or an old fashioned cassette can be played in mono just by adding the two tracks together. But our experience led us to the basic requirement of four channels for the cinema, instead of two as for records. We first considered four discrete tracks. The problem there was compatibility with mono. If you went to four discrete tracks and then summed them all together in a mono theater, the center channel, where the dialog is, comes out too low. Another reason is that with four discrete tracks you need a guard band [a small blank space] separating each of the tracks on the film, which would end up in throwing away about 40 percent of the track width available on the print, resulting in a very small track size and very high noise level.

[No, not a mystical realm from a popular movie, but a clever electronic way of using phase relationships to sandwich four different tracks into two. It was a miracle when first introduced to film, but soon became a burden, as filmmakers pushed the technology to the limits.]

Actually, it dates way back to 1934. Alan Blumelein not only invented the matrix, he developed the stereo microphone and then went on to be a major contributor in the design of the first radar technology. So the matrix is not new. It was used in the late 60’s by Sansui and by CBS SQ, with their two competitive systems, as a technique to get four channels of music on a long-playing record. We adopted it in 1975. The first commercial movie release using our



Allen with Ray Dolby, demonstrating the CP50 Cinema Processor in ‘79.

system was A Star is Born in 1976. We started out using a modified version of the Sansui US Matrix, but then by 1978, we developed our own matrix, which is better tailored to film.

It was pretty tough. The first thing I realized was that the studios were reluctant to get involved with something new. I had to go to the artistic community to get their commitment on each film. We spent a lot of time with directors and editors, teaching them the benefits of this new type of soundtrack. Once we got their commitment, we were able to introduce the concept to the studio. And then we went to the theater chains. We got a play list from the studio and then went to the different theater chains and said, “You will be playing this movie, what about equipping some more houses with Dolby?”

Change. People distrusted change, and we were the new kids on the block. Nowadays, everybody in the theater industry and the studios knows who we are, but in those days they had no idea what Dolby was. But from the mid 70’s, there was a steady escalation, as all the big film titles started coming out with Dolby-encoded stereo soundtracks. And the theaters suddenly realized that unlike four-track mag, where they paid a lot of money for the equipment but only got prints for special releases, they would be sure of getting a stereo print because all the major films that were coming out were Dolby stereo prints. That was A-type noise reduction.

Then eventually, Dolby SR [Spectral Recording] superseded A-type noise reduction.

SR was introduced to the music industry in ’86. It was used in motion pictures for the first time in ’89.

What about the greater dynamic range offered by SR? Was that something that was looked upon positively?

Absolutely. And people wanted even more — louder. That was one of the reasons why we went to the digital track in ’92.

The original Dolby analog soundtrack was considered to be compatible with Academy mono playback. It could be played in an old-fashioned, non-Dolby theater, as you already described. Now you have a new type of processing, SR, which Dolby claimed, could be played acceptably in a theater that only had A-type noise reduction on their projectors. As a mixer, I was skeptical about this, and a lot of mixers still are skeptical about this. But in the interest of advancing the technology, we all went ahead and worked with it. In retrospect, was this a little bit of sleight of hand on Dolby’s part?

A lot of discussions happened at the time. We could have made the SR track much more spectacular. We could have allowed you to raise the level by 6 db as opposed to 3 db, which would have made for much greater dynamics. The reason we stuck at 3 db was to achieve

You can take a 1935 projector and still play today’s analog soundtrack. You’d understand the words, you’d understand the music, it would play.

commercial acceptability for A-type playback. I don’t like the word compatibility. I’m conscious of the fact that we put an onus back on you guys, the mixers, to ensure that there weren’t some horrendous low-level effects that would rise up if you played it back A-type. [The effect of playing a Dolby SR-encoded soundtrack through Dolby A equipment is similar to playing a Dolby encoded audio cassette without the Dolby button pushed in.] But you’ve got to make progress, you’ve got to move onwards. Within two years, we had quite a few SR-equipped theaters. These backward-compatibility issues become less important as time goes by. Today, you can still play one of those stereo analog soundtracks in mono. But find me many mono theaters in a big urban center — there aren’t any. In the same way, I think 90 percent of the screens in this country now have SR capability. So A-type playback becomes less important.

Could I generalize by saying that as technology continues to move forward, this problem of compatibility will become worse and worse? For instance, I have all this stuff on floppy disks, but I might as well just throw them away, because I can’t find anything that will read a floppy anymore.

What we have on the analog soundtrack is really extraordinarily compatible. You can take a 1935 projector and still play today’s analog soundtrack. You’d understand the words, you’d understand the music; it would play. The dimensions of the soundtrack area are the same. We in the film industry have taken enormous care to ensure that things are backwards compatible. Also, you can take a film from here and play it in Norway or Zanzibar, and it will play just fine on their analog projector. You can’t do that in radio; you can’t take your FM radio to Japan. It won’t play the radio stations, because they have different frequencies. You can’t do it with AM in Europe because the spacing of the stations is different. When we got to Dolby Digital, we managed to achieve 100 percent perfect compatibility by putting the digital data somewhere else on the film and leaving the SR track totally untouched.

In our next issue, Ioan talks about the end of white-light optical soundtracks, the advent of Dolby Digital and the conundrum of film sound loudness.

Go To Part 2

Leslie Shatz is a sound designer and mixer. His credits include Ghost, Dracula,
Good Will Hunting, The Mummy and The Mummy Returns. He can be reached via email.

Guild Home | Magazine Home