Optical replay of records


From the birth of the telephone, switchboard operators were besieged with calls asking the time.

The demand eventually became so great that the telephone companies decided that a mechanised speaking-clock would be an excellent service they could offer their subscribers — and charge for it!

A speaking-clock would require the recorded sound of a voice but clearly could not involve a medium like a gramophone (phonograph) record where the reproduction involved physical contact. The incessant, repeated playing would wear out the medium too rapidly.

The solution was to use glass discs onto which the waveforms were recorded photographically — like the soundtrack of a movie film. A bright light, shining through the discs, was intercepted by a photo-cell and converted to an audio signal. The first British speaking-clock mechanism using this technique is illustrated above.

From speaking-clocks to laughing policemen

The idea of adapting this idea to play phonograph (gramophone) discs without any physical contact using a reflected light source to sense the groove modulation is almost as old as electric recording itself.

For example, in US Patent 1,916,973 (filed 21 May 1929) a Reginald T. Friebus of the Columbia Phonograph Company, describes a system in which, ...light is shone along the grooves, and the reflection from a groove flank is detected by a (sic.) phonocell as a varying amount of light.

It is unlikely that this system provided anything but a very distorted signal, but it demonstrates the interest in the technique. And the inventors kept at it — through the age of shellac records and into the LP era.

In a US patent (US 3,138,669 filed 1961), Jacob Rabinow and Arthur Morse describe a technique in which, a beam of light (from a low-persistence CRT) is made to scan back and forth perpendicular to the LP record groove, and every time the reflection changes markedly, the direction of scan reverses. In this manner, every other reversal represents one side of the groove and those in between represent the other.

The system is thus a sampled system and time is converted to a representation of groove modulation amplitude in an integrator.

The arrival of the LASER

In 1977, William Heine presented a paper at the 57th Audio Engineering Society (AES) convention¹. It is the first modern description of a physically contact-less record player — a development which was founded on the invention of the LASER (hereafter no longer capitalised).

His paper details a method by which he employed a single 2.2 mW helium–neon laser for both tracking a record groove and reproducing the stereo audio from a phonograph record in real time.

Heine's text reads like something from the nineteenth century in its directness and lack of sanctimony. It could have been written by Edison.

After only a few short hours of playing around with the laser, a simple magnifying glass, and a phonograph record, I discovered a unique and fascinating phenomenon. By simply focusing the laser beam on a record groove, I was able to produce a brilliant arc of light, much like a red "rainbow, ' that spanned the room, across the ceiling, from wall to wall..... As I turned the record by hand, the image began to dance about in response to the groove engravings, and I recognized immediately that I had found the means to reproduce the record.

Heine had, in fact, stumbled upon the same phenomeneon which produce Buchmann Meyer patterns which have been recognised since the 1930s a direct measure of the information recorded in the grooves of the record. In the months and years that followed, Heine worked to refine his initial eureka moment into a practical record player.

Lack of physical contact deprives the optical phonograph of the simple mechanism which drags the reproducing element across the record so that it follows the groove spiral. . Like the linear tracking turntable, the sensors, motor and control system must be developed to do this; rôles which are made all the more complicated due to the lack of any physical contact as well as the irregularities (warp, eccentricity etc.) of real reccords.

Heine admitted that it took him four years to develop these systems to the point at which he could apply for a patent for first laser-scanning phonograph record player, the LASERPHONE².

Sadly, the LASERPHONE never got beyond the prototype, and it would be another 10 years before Robert Stoddart, who had studied electrical engineering at Stanford University was to file a patent³ for a laser-based turntable building on Heine's ideas.

Stoddart started a company (Finial Technology Inc.) and raised significant venture capital to develop and productionise laser record players. The Finial Technologies' design betrays a sophistication that only commercially funded R&D could achieve. A production mock-up of the Finial LT-1 (Laser Turntable-1), was completed in time for the 1986 CES.

Contemporary accounts4 report that the LT-1 design still had engineering issues but a much greater problem for the company was the disappearing market for phonograph records which had, by this time, been wholly undermined by the Sony Philips Compact Disc. What role could an - inevitably complicated and expensive - turntable play in this new digital world?

So, in 1989, Finial Technologies was liquidated without ever having put the laser turntable into production and the patents were sold, ultimately ending up in the hands of a Japanese company ELP (Edison Laser Player) Corporation Japan. ELP was formed to carry on development of the turntable on the basis that there might still be enough Japanese fans and collectors of the analogue disc to justify continued development.

After eight more years of dedication and development the laser turntable was finally put on sale in 1997 as the ELP LT-1XA Laser Turntable, with a list price of US$20,500.

The mature ELP product incorporates five lasers – one on each channel to track the shoulders of the groove and one on each channel to pick up the audio (which scan just below the tracking beams). The fifth laser tracks the surface of the record (the land) and keeps the pickup assembly at a constant height above the groove. See panel for information on contemporary ELP Corporation Japan laser players.

As light as light

It is often said (ELP themselves say it), that the great advantage of the laser turntable's ephemeral interaction with the disc is the lack of physical (friction-based) wear. But, unless the disc needs to be played, many, many times — as in the speaking clock — the advantage of contactless play is honestly relatively marginal.

Provided a tracking weight of less than 2 grams is used for playback and the other precautions are correctly observed (side-thrust compensation, stylus condition etc.), there is no real reason to worry that making a needle-drop recording with a conventional stylus-based player will subject the record to further unnecessary or irresponsible wear and tear.

In fact, it's quite likely that the record suffers more damage due to incorrect or careless storage than it will by being played occasionally with a modern, lightweight stylus.

Nevertheless, the real advantages of the optical player do derive from the massless nature of the light beam.

A massless "stylus" cannot resonate with the plastic compliance to produce a high-frequency resonance. Neither can the mass of the cartridge armature resonate with the bearing compliance. And neither resonance can be excited by groove damage or dirt.

A massless laser-beam cannot be pulled sideways in the groove to produce a side-thrust. Neither can a light beam become worn, or become clogged with lint.

A spot of laser light cannot suffer from the effects of tracing distortion or wavelength loss — its tiny dimensions may reliably follow the most convoluted curves of the groove and it may be accelerated instantaneously; without limit. It may also "run" at a different height in the groove (above the mid-point) so that it is scanning a part of the groove-wall never touched by a physical stylus the profile of which is illustrated in red in the diagram above.

A beam of light cannot transmit mechanical vibrations back into the tonearm or transmit physical sound back to the disc. And it cannot bounce the tonearm weight on the spring of its compliance.

Neither is a laser based pick-up affected by external magnetic fields or cable capacitance.

In short, a massless optical pickup has the potential to circumvent virtually all the compromises inevitable in an electrodynamic pickup; although, as in all engineering, it may introduce some of its own6.

Edison Laser Player (ELP) Japan

ELP Corporation of Japan is based in Saitama, near Tokyo. It is run by Mr. Sanju Chiba who, after a career in General Electric, set up ELP in 1989 to manufacture the laser turntable originally developed by Finial Technologies Inc. (see main article).

Initially, Mr. Chiba tried to interest US and Japanese companies to take on the product, but couldn't and eventually decided to go it alone knowing the path would not be an easy one. As he says himself5

It is my belief that Analog Music is very precious for the human. ...... Therefore without Profit and Loss calculation, I acquired the technology in 1989.

Chiba's intuition about the difficulties involved with developing the laser turntable was prophetic — Finial's prototypes still required significant engineering to make a viable product — and his self-imposed defiance of traditional profit and loss led to a financial as well as technical struggle.

It is greatly to his credit that he persevered in developing the commercial players he manufactures today. But, as he says,

Music in vinyl records is more valueable than money.

Today, ELP produce two models, the LT-basic and the LT-Master.

The LT-basic player (above) supports all record sizes (72, 10", 12") at 33⅓ RPM and 45 RPM speeds. The LT-Master has enhanced audio (analogue) electronics but otherwise similar specifications. The LT-Master is illustrated at the top of the page.

Although, to talk about standard models is a bit misleading, ELP only sell direct to their clients with whom they maintain a close and personal relationship. So the precise machine specification may be adapted for the individual customer.


Alternatve history

The ELP player remains the only optically based record player which performs in real time - and with entirely analogue signal-pickup and signal-processing.

Confined to a tiny market niche, the ELP players remain expensive — although ther are good value for extraordinary engineering produced in small quantities.

A laser based turntable has so many advantages over a stylus-based player (see main article). If the production quantities could have been increased, there are few reasons why a laser-based record-player could not have become a standard, low-cost consumer item: like the CD player became.

After a fashion, the ELP players offer a glimpse of an alternative history — what record players would have become had the CD and digital audio not swept all before it 40 years ago.

Chiba-san describes the technology and the business in this video. Check out the endorsement at the end of the film!


ELP's Turntable Specifications

Frequency response: 20 - 20,000Hz +-3dB
Channel Separation >25 dB (DIN 45 543 Test Record)
Line output: (600mV (5cm/sec 1kHz)
Optional phono output: 4.8mVrms (5cm/sec 1 kHz)
Distortion: <0.5% DIN45 543 1kHz Ref. Level
S/N Ratio: >55dB (Weighted) Ref. Level
Wow & Flutter <0.07% WRMS


ELP and Stereo Lab


Provided the ELP player has the optional PHONO output, all the functionalities of Stereo Lab may still be exploited.

A flat-response preamplifier is required. The ELP players have an analogue click supression circuit which may be selected/ de-selected. This should be disengaged and the click supression in Stereo Lab used instead.



Optical techniques & archiving

None of the systems described below have any immedate, close application with Stereo Lab. They are reported here because of their interest to any record collector or archivist.

IRENE Project


Other applications of optical technology have emerged to play phonograph discs. One such is the IRENE (Image Reconstruct, Erase Noise Etcetera) System.

Developed by physicist Carl Haber, IRENE was a joint project by Physics Division of the Lawrence Berkeley National Laboratory and The Library of Congress. Using methods derived from work on instrumentation for particle physics Haber and his team investigated the problem of audio reconstruction from mechanical recordings. The idea being to acquire digital maps of the surface of the media, without contact, and then apply image analysis methods to recover the audio data and reduce noise. This work is well described in two papers 7,8.

The IRENE process creates ultra-high resolution images of the audio groove structures in either 2D or 3D, as required. The resulting image files are then processed through software that translates them into an audio file. If properly cared for, the image files serve as a digital surrogate of the physical object, “virtually” preserving the object’s condition at the time of scanning while the object continues to physically degrade over time. Haber is at pains to distance the technology from the laser-based player. He says,

Reconstruction of mechanically recorded sound by image processing should be distinguished from the use of laser or light beam based turntables...... The methods always rely on the analysis of digitally imaged frames of the recording 7.

From a technology point of view, the IRENE project is therefore more about metrology (the science of measurement) and image processing than it is about developing a light-based pickup system. However, its motivation was absolutely for the re-formatting and archiving of phonograph (and later cylinder) recordings; hence the involvement of the Library of Congress.

The early work in the IRENE project used 2D images — all that is required for laterally modulated discs (see above). Later 3D maps were developed for hill-and-dale records and Edison cylinders (right).

San Francisco PBS affiliate KQED made a short documentary about IRENE with interviews with Carl Haber himself. The video is very informative.. even inspiring!

Visual Audio & Saphir

National and other archives may have many thousands of mass-produced discs from the shellac and vinyl period and perhaps even cylinder recordings too. They also frequently possess a significant number of one-off recordings known a transcription discs or lacquer discs. Copeland11 and Pickett12 both document that lacquer discs (transcription discs) contribute a considerable fraction of the production of audio recordings between 1930 and 1960. The main users for such recordings were radio broadcasters who used transcription discs for all the studio and transmission duties replaced by audio tape from the 1950s onwards. But these discs were also in use for recording the minutes of meetings, trials, and other events of historical importance. All these recordings participate to the cultural heritage of the Twentieth Century9.

The French Institut National de l’Audiovisuel (INA) estimate that they have more than 20,000 such discs awaiting digitisation in their archive alone. And a large fraction is critically endangered. Lacquer discs, such as those illustrated (right) are much more fragile than vinyl or even older shellac records. This is due to the contrast between the high stability of the core (aluminium, zinc, or glass…), and the chemically evolving composition of the lacquer that is used as the recording layer.

The lacquer contains plasticisers which evaporate or migrate to the surface with time. The layer often tends to shrink and cracks appear, in a typical radial/tangential pattern 9. Discs like this are now impossible to play using a stylus.

Optical techniques, because if their lack of physical contact, offer the promise of the recovery of information from discs which have become damaged to the point that playback with a stylus is simply impossible and this has provoked interest amongst archives and organisations all over the world.

One such, is the Swiss Visual Audio process — a collaboration between Département d’Informatique de l’Université de Fribourg and Ecole d’Ingénieurs et d’Architectes de Fribourg12. The Visual Audio system uses an intermediate, analogue photographic stage so that very high quality photograph is first obtained of the disc. This is the primary archive. Scanning and post-processing is then used to derive the audio.

Another project in a similar vein is the experimental Saphir system developed by the Institut National de l’Audiovisuel in France 9.

The Saphir process uses a specifically designed colour illuminator which exploits the reflective properties of the disc material to highlight subtle changes in orientation of the groove walls. A standard colour camera is used to collect pictures of small sections of the disc and an Elementary Shortest Path Solver (think GPS!) is used extract the audio signal from the collected pictures. None of these systems is anywhere approaching real-time. These processes is slow — several hours per disc. But this is largely irrelevant in a context such as this.

Saphir, the most recent of the systems reported here, has proved to work well on manufactured discs, from earliest Berliner recordings to recent vinyl records. But, the designers sound a word of warning.

When mechanical playback is possible, signal quality is usually audibly better than using our optical process. .... Frequency response is good.... On the outer track .... the frequency response is -8dB at 20kHz....

Using the Slope decoding scheme, however, THD+N stays over 18% (-15dB SNR) on the same section; this is in the same range as found by VisualAudio: -19dB SNR on a 78 rpm between 500 and 10,000Hz.....

From those findings, we estimate that in most cases signal quality will be, using our system, poorer than when using a well-calibrated mechanical playback process with the most appropriate stylus. For this reason, we are currently focusing on discs that cannot be played using a stylus.

This is clearly the forte of these systems: decoding extremely damaged (broken, de-laminated, oxidised) discs which are physically unplayable.



Notes and references

1. A LASER SCANNING PHONOGRAPH RECORD PLAYER Heine, W. K. Presented at the 57th AES Convention May 10-13, 1977.

2. "Patent US3992593 – Disc phonograph record playback by laser generated diffraction pattern.

3. US Patent 4,870,631 OPTICAL TURNTABLE SYSTEM WITH REFLECTED SPOT POSITON DETECTION (filing date 30 May 1986, granting date 26 September 1989) to Robert E. Stoddard (Finial Technology and later ELP Corporation).

4. https://en.wikipedia.org/wiki/Laser_turntable

5. http://www.elpj.com/why-is-the-laser-turntable-lt-only-from-elp/

6. We know of two already: 1) Deprived of its physical contact with the groove, the optical system does not brush dust and dirt out of the way so that records must be very clean to get good results from the laser turntable. And 2) If the optical situation is different, the system won't work - for example for coloured and picture discs won't play on the ELP machines.

7. Reconstruction of Mechanically Recorded Sound by Image Processing. Fadeyev, V. and Haber, C. J. Audio Eng. Soc., vol. 51, no. 12, pp.1172-1185 (2003 Dec.)

8. Reconstruction of Recorded Sound from an Edison Cylinder using Three-Dimensional Non-Contact Optical Surface Metrology. Fadeyev, V. Haber, C. et al. J. Audio Eng. Soc., vol. 53, no.6, pp.485-508 (2005 June).

9. Saphir: Optical Playback of Damaged and Delaminated Analogue Audio Disc Records. Jean-Hugues Chenot, Louis Laborelli, Jean-Étienne Noiré. Journal on Computing and Cultural Heritage, Association for Computing Machinery, 2018, 11 (3), pp.14.1-29. ?10.1145/3183505?. ?hal-01885324?

10. Peter Copeland. 1991. Sound Recordings. British Library

11. Preservation And Storage Of Sound Recordings. Library of Congress. Andrew G. Pickett and Mike M. Lemcoe. 1959.

12. Optical Retrieval and Storage of Analog Sound Recordings. Stefano Cavaglieri, Ottar Johnsen, and Frédéric Bapst. Proceedings of Audio Engineering Society 20th International Conference (2001).


Links

Pspatial Audio Home page

For all support issues, go here.

For Pspatial Audio sales, email: sales@pspatialaudio.com



© Pspatial Audio 2020. All rights reserved. Apple Certified Developer. Stereo Lab, Aria 51, Aria 20, Head Space, Groove Sleuth, iLOOP and FRANCINSTIEN T-Sym are trademarks of Pspatial Audio. FRANCINSTIEN and Bride of FRANCINSTIEN (BoF) are trademarks of Phaedrus Audio. Macintosh and the Mac logo are trademarks of Apple Computer, Inc.