Ampex doesn’t really need much of an introduction. It was the company that commercialized magnetic tape recorders in their present form in the United States, after World War II. They developed many good tape machines over the years and also pioneered several advances in magnetic tape formulations. By the 1970s, they were already long established in the recording industry.
They embarked on the design of a new machine, which was to go beyond the established notions of tape transport design. In 1976 they introduced the Ampex ATR-100. Designed by Robert P. Harshberger Jr, Alastair M. Heaslett and Roger R. Sleger, under the management of Frank Santucci, the ATR-100 was a radical departure from what had previously been done by Ampex or other companies in the design of tape machines.
Its most striking feature is the absence of a rubber pinch roller! In most, if not all previously-made tape machines, the pinch roller “pinches” the tape against the rotating capstan and is driven by it. It increases friction, ensuring the tape remains in firm contact with the capstan and does not slip. In doing so, it also mechanically decouples the tape path before the capstan from the tape path after the capstan. Typically, the tension was not constant throughout the duration (playing time) of a reel of tape, but varied depending on the amount of wound tape in each reel. Tension was controlled by a variety and often a combination of means, including spring loaded tensioners, mechanical supply reel braking, electronic supply reel braking, electronic torque control of the takeup reel motor, flywheel rollers, and so on.
By contrast, the ATR-100 has an unusually large diameter capstan, perhaps the largest seen on a tape machine until 1976, at just over 2-3/8 inches OD (outside diameter). Interestingly, this size was intended to be directly related to tape velocity multiples (take the circumference, which is 2πr, where r = radius), to provide 7.5 ips (inches per second) at 60 rpm, 15 ips at 120 rpm, and so on.
The tensioner arms are electromagnetically driven instead of spring-loaded. The idea is to provide constant tension throughout the tape, regardless of the amount of tape wound on each reel. It is still the capstan which controls and regulates tape speed, but the regulation of tape tension is accomplished by electronic means, by active control of the reel motor torque and electromagnetic tensioners.
Does this actually work?
It works very well, as evidenced by the widespread acceptance of the ATR-100 in the industry and the many excellent-sounding albums recorded with it.
The tape heads were made of ferrite and the bias frequency used in recording was 432 kHz, as compared with 100-150 kHz used by most other tape machines in the 1970s. The head assembly could be plugged in and out to change over between 1/4-inch and 1/2-inch tape.
The range of Ampex ATR machines was quickly expanded. The ATR-102 was the stereo version, the ATR-104 was the 4-channel version, the ATR-116 was the 16-track version and the ATR-124 was the 24-track version, the last two intended for multitrack recording.
While Ampex had previously made a preview-head tape machine (the Ampex 300, often used together with Scully disk mastering lathes in the US), which provides the control signals used by the recording pitch and groove depth automation systems of a disk (record) mastering lathe, the ATR-100 was never offered with this option. Instead, the ATR-100 came at a time when a lot of decision makers in the industry had gotten a bit too excited about digital delay lines, which were used in disk mastering installations with the primitive digital delay units available at the time, instead of the good old analog equivalent of the preview head. Since the control signals for the lathe automation systems must arrive in advance of the signal to the cutter head, the digital delay units were delaying the audio signal, not the control signal. So, the audio signal from tape had to undergo an A/D conversion and a subsequent D/A conversion, before being fed to the cutting amplifiers.
This was not a good idea for fidelity.
Doug Sax (the late, well-renowned mastering engineer and one of the founders of Sheffield Lab, along with Lincoln Mayorga, which produced some of the finest direct-to-disk recordings I have ever heard) had expounded his views on this practice in a 1980 interview for Audio magazine, conducted by Gary Stock (the interview can be found here: http://www.magneticfidelity.com/sax.shtml). He felt that the (CD-quality and often higher) sampling rates used in professional-grade digital delay systems at the time were inadequate, producing audible degradation of the sound quality. He believed that sampling rates in excess of 100 kHz would be needed in order to achieve decent fidelity in a digital audio system. It has only taken around 40 years for his views to be reflected by the leading authorities in digital audio, who have finally settled this matter by means of several peer-reviewed papers, published in the Journal of the Audio Engineering Society in a special issue dedicated to high-resolution digital audio: (“High-Resolution Audio: A History and Perspective,” Vol. 67, No. 5, 2019 May).
Some people are a bit ahead of their time. Oh well, better late than never…
The legendary Town House studio complex, one of the finest recording and mastering facilities in the UK at the time and, constructed with no expenses spared, had several ATR-100 machines in regular use.
One of these lived in their cutting room, and was the typical digital-delay-line-equipped ATR-102, with 1/4-inch and 1/2-inch head blocks, able to cut from 14-inch reels at up to 30 ips. They had a Neumann VMS 80 disk mastering lathe and their brochure boasted about having the ability to also cut from a 3M Digital Recorder (now entirely obsolete and unlike analog tape, not particularly missed by anyone) as well as direct-to-disk from a live source performing in the studio next door!
They also had four Ampex ATR-102 machines in their tape copying room, for real-time copying of master tapes. In addition, they had 24 Aiwa 6900 cassette decks for 1-7/8-inch ips cassette tape duplication, from a 1/4-inch tape master playing on one of the ATR-102 machines, or a variety of other sources.
Electrical Audio, an outstanding recording facility owned by musician/producer/engineer Steve Albini (one of the very few audio engineers who still refuses to use computers for recording and does the entire process on tape instead), has Ampex ATR-100 machines, both in 1/4-inch and 1/2-inch configurations, as their stereo mixdown machines, where the final mixes from 24-track or 16-track multitrack tape are recorded (their multitrack machines are Studer A820).
Decades after the introduction of the ATR-100, Michael Spitz converted an ATR-104 into a preview head tape machine for George Marino at Sterling Sound, then one of the major mastering facilities in New York City (now in Nashville, TN). The conversion involved an extension to the machine’s top platform, where an additional reproducing head and a few rollers could be added. The ATR-104 was used as a base, since 4 channels of audio are needed in a stereophonic preview head tape machine; two for the preview/control signals and two for the audio. This machine is still in active service, nowadays used by Ryan Smith and Joe Nino-Hernes at Sterling Sound’s Nashville facility. There was one more custom preview-head ATR-100 machine, of a different design, reportedly used by Stephen Marcussen at Precision Mastering in LA, but it still remains a mystery who put this together.
In the next episode, we will have a look at the rarest, wildest example of an Ampex ATR-100!
Header image: an Ampex ATR-102 sold by ATR Services, Inc.