Professor Bose ~ Early Research ~ Orchestra ~ Inspiration ~ Automobiles ~ All In One ~ Mission ~ Gallery

Tanglewood

The sound of a violin and of an orchestra

Since the 2201 ultimately failed we asked a different question. Could we, in some way, recreate the experience of the concert hall over loudspeakers? Our approach now was to concentrate on the spatial qualities of the sound that were normally compromised by playback in a living room. In effect we took away the normal rules of recording and listening by designing for playback in an anochoic chamber.

An anechoic chamger is a room without echoes, all surfaces (including the "floor") sre covered with deep wedges of fiberglass.

In retrospect it seems we should have discovered the secrets sooner, but that is often the way it is. When you are doing it, it is not so obvious and you may try several different directions at the same time.

Since violins were the problem it seemed reasonable to learn more about how they radiated sound.

But what was it about a live performance that made the problems disappear? To try to answer that question we decided to try to recreate live performaces by recording the sound from reflecting surfaces as well as the direct sound of the orchestra. Here is how we did it.

We recorded the Boston Symphony in several different ways.

1. Binaural. A dummy head has microphones in its ears. It is placed in an ideal location in the hall. Listening is over headphones and simulates natural hear. All the sound reflections in the hall contribute to the naturalness of the sound.

We often tayed after a conert and offered headphones to the conductors and others interested. The conductor of the BSO at the time, Eric Leinsdorf, stayed long after one concert listening to the binaural and being entranced by hearing the sound as a member of the audience would hear it.

I had my own pair of headphones and offered them to Aaron Copeland whose music had been played that evening. He was entranced also. However, my headphones were connected to a control box where I could combine the left and right ear sounds. This created a monaural presentation that eliminated the reproduction of the spatial characteristics of the sound. I flipped the switch. Instantly the headphones came off Mr. Copeland’s head. “What did you do?” he said. “It now sounds like hi-fi.”

2. Numerous experiments with an 8 track recorder, capturing reflected and direct sound. A diagram of one such arrangement (among many) is shown above.

3. The third diagram shows an indistinct picture of a 9 foot wheel with 8 spokes. A directional microphone at the end of each spoke recorded the sound what would be played back in the anochoic chamber over 8 small loudspeaker arranged in a 9 foot diameter circle. The size of the anochoic chamber set the limit of the diameter of the wheel.

4. The graph is a polar plot of the sound of a violin (back in the MIT anechoic chamber) in our search for clues to better microphone placement when recording.

Without describing each experiment in detail, it is safe to say that the best of the recordings created a sound that was uncanningly real, with no trace of the artifacts heard from the 2201 loudspeaker. The implication would be that it is the spatial reproduction of the sound that is at least partly responsible for the artifact.