Companies Seek to Save ‘Audio Heritage’ Through Use of Lasers, Fiber Optics
We are well into the age of audio recording with digital media, where MP3, MPEG and AU files are the norm. Still, the need to play back 20th-century wax cylinders, 78 RPM discs and vinyl records remains. It is estimated that approximately 30 billion phonograph records were produced and sold in the past 100 years.
Some of this material is too esoteric to make commercial transfer to CD worthwhile. Other cylinders and discs are physically broken, or so badly damaged that playback with conventional equipment is nearly impossible. Some of these are one-of-a-kind recordings with historical significance. A substantial portion of our audio heritage is at risk of being lost forever.
In response to this need, scientists and engineers are developing high-tech alternatives to disc playback with a conventional stylus. Using laser beams, fiber optics and digital imaging techniques, once-lost recordings are being brought back from the grave. The results from these devices must be heard to be believed.
The ELP Corp. of Japan has been developing laser turntables since 1989, when it acquired the rights from the Finial Co. in America.
Some of the earliest machines suffered from mediocre audio, but today’s turntables have been redesigned for ultimate audiophile quality. ELP’s machines feature analog outputs that can connect to the phono input of a conventional stereo system.
The current models use a pickup head with four lasers for contact-free tracking and audio reproduction. Two tracking laser beams are directed to the left and right shoulders of the record groove. Only the part of the beam that reaches the groove is reflected to two PSD – Position Sensitive Detector – optical semiconductors. The signals are sent to a microprocessor via analog-to-digital converters, and then to servos to maintain the pickup head’s position directly above the groove.
Two additional laser beams are directed to the left and right groove walls just below the individual tracking beams. Modulation of the individual grooves is reflected to scanner mirrors and onto left and right photo-optical sensors. Variations of the modulated light cause the audio sensors to develop and electrical representation of the mechanical modulation in the grooves. The entire signal path is analog.
A separate laser is used in the ELP turntables to maintain the distance between the record and the pickup heads, and is similar in function to the focus laser used on CD players. It enables the turntable not only to accommodate records of different thicknesses, but also recordings that are badly warped.
The sonic advantages of an optical pickup and analog signal path are apparent when playing any recording, and there is a noticeable reduction in noise when playing scratched records. Where the turntable excels, however, is in playing discs that have been physically broken. ELP’s demo CD includes a track that claims to be a recording of an LP that has been shattered, with the broken pieces reassembled in the tray of the turntable. While there are obvious clicks and pops, such a disc could not be played at all using traditional stylus-based equipment.
Despite its sophistication, there are some types of recordings that the laser turntable cannot play. Clear or colored records that are translucent, popular in the 1980s, cannot reflect the laser beam. Vertical cut records such as the early Edison “Diamond Cut” series cannot be played because the modulation is up and down rather than lateral. Records with a rounded groove shoulder will not work. A badly-worn disc with grooves that are rounded in the bottom can produce distortion. And don’t bother digging out your copy of The Who’s Quadrophenia; the ELP turntable will not play four-channel recordings.
Laser turntables don’t come cheap. ELP manufacturers three models ranging in price from $10,500 to $14,300. According to President Sanju Chiba, ELP has supplied turntables to broadcasters NHK Japan and a station in Mexico as well as several universities, museums and libraries.
Fiber optic approach
Another “optical turntable” is under development by the Laboratory of Metrology at the Federal Institute of Technology in Lausanne, Switzerland. Noting that other optical systems suffer limitations due to the condition of the grooves, the reflection coefficient of the used part of the disc, and variation of the distance between grooves, the Swiss have taken a different approach to extracting the information.
An optical fiber is used to track the grooves on the disc, in a manner similar to a conventional stylus. Light is injected into the fiber by a semiconductor laser. Light reflected from the undulating fiber is focused by a lens onto a position-sensitive X-Y detector. Signal processing of the detector output creates one signal for the X axis, and another for the Y axis. The X signal corresponds to the audio, while the Y signal is used to control the vertical position of the optical head. For cylinders which use vertical modulation, the Y signal can be used for audio output.
While the fiber system is in physical contact with the record, its tracking force is limited to 60 mg, about 40 times less than that of a traditional stylus. The low force is possible because the optical head is guided by servomotors controlled by the optical signals. The system also allows the pickup head to pass over cracks and the small areas without grooves common on older resin discs. The EPFL optical turntable can retrieve material from both cylinders and discs of all formats.
Because the fiber contacts the record grooves in areas where a stylus typically does not, the system is usually working with virgin material.
Drawing on digital imaging technology originally developed to study subatomic particles, physicists Carl Haber and Vitaliy Fadeyev of the Lawrence Berkeley National Laboratory have developed a radically new way to digitally restore and preserve audio recordings. Much of their research centered on particle physics, and required the ability to image the tracks made by elementary particles as they hit detectors, and then find these tracks in a cacophony of noise.
Haber said, “We thought these methods, which demand pattern recognition and noise suppression, could also analyze the grooved shapes in mechanical recordings.”
Haber and Fadeyev took a precision optical metrology system that had been designed to inspect silicon detectors and programmed it to map the modulating grooves cut in shellac phonograph discs. The images were digitally processed to remove surface noise, and run through a program designed to simulate how a stylus would respond to the grooves.
Finally, the virtual stylus motion was converted to a digital sound format. Before and after samples reveal astonishing results.
Haber and Fadeyev’s research led to an interagency agreement between Berkeley Labs and the Library of Congress, and will eventually provide the library’s staff with a technology to restore some of the 500,000 items that it provides preservation treatments to each year, from a collection totalling almost 128 million items in all formats.
An additional benefit to the digitization program is that it will give the public greater access to these materials than has been possible in the past. Many of these recordings have been off limits to library patrons due to their fragile condition.
As to the future of the technology, Haber said, “There are no current commercial plans. We are seeking government support to continue the efforts and to develop a prototype ‘production’ level machine, which could be a first step in developing something for wider use.”