Magnetic Tape Deterioration, Part 1:  Magnetization & Base Damage  (October 2022, updated December 2023)

Since launching Play It Again, Paul, I’ve encountered several reel-to-reel tapes in less-than-optimal condition.  While most were playable, though occasionally with minor distortion, others had deteriorated to the point of major distortion, physical damage to the tape, and loss of audio.  In this pair of articles, I’ll be addressing primarily reel-to-reel tapes, since problems are encountered most frequently with this medium.  Cassettes are not immune, but incidence is less frequent, and issues have received less attention.  This article describes problems associated with magnetization and with damage to the base (or substrate—the flexible strip that gives the tape body and upon which other layers are applied) and concludes with recommendations for slowing deterioration.  Part 2 addresses tapes’ binder (a coating that holds the magnetizable and lubricating particles in place) and back coating (a layer opposite the binder that reduces static electricity and provides additional lubrication).

Magnetization

Print-Through

Tapes store audio content in magnetizable oxide or metal particles called domains.  Domains are not only magnetized during recording, but they can also be unintentionally altered by other magnetic sources.  An unavoidable source, known formally as thermo-remanent magnetization, is the magnetization that tape will absorb from adjacent layers of tape on a reel.  The result, print-through, is that one can hear not only the tape’s original audio content but also content from a few inches away at the same time.  Unfortunately, this begins taking effect immediately after a recording is made.  When it reaches audible levels depends on the amount of time that has elapsed, exposure to high temperatures during storage, thickness of the base, stability of the magnetic particles, the speed at which the tape was recorded, and the difference in volume between the original and print-through content.  There’s no way to remove print-through content, though it can be minimized in places where the original content is silent.

Several authorities suggest that print-through can be countered by winding the tape onto another reel and storing it “tail up” for an extended period.  Each layer of tape will be adjacent to different sections with different content, and as print-through inevitably begins again, the effect is to counteract the original print-through content.

My perspective on this is that the net result would be thermo-remanent magnetization acquired from two separate layers of tape resulting in accumulated noise—not an improvement.  Personally, I dislike this practice since the tape ends up on a reel likely to have different brand markings than the original, thereby removing a clue of the tape’s true brand (and possibly a label on the reel indicating the tape’s content).

Cassettes are also subject to thermo-remanent magnetization.  A significant portion of a collection of commercially produced cassettes I converted in a recent project suffered from print-through, which substantially increased the effort to achieve satisfactory noise reduction.

External Magnetization

A second type of damage is anhysteretic magnetization, which occurs when a tape is placed within an external magnetic field.  Magnetization picked up from external sources can be as strong as that of the original signal, and removal of the resulting audio noise may not be possible.  This problem is somewhat infrequent since most people using tapes are typically careful about accidental exposure.  I store all magnetic media—especially my customers’—several feet away from any electronic devices and their cables, electrical wires including those in my walls, electric motors, computers, and CRT monitors and televisions (when I still had some).  Though I haven’t seen any specific warnings regarding flat-screen monitors and televisions or wireless phone chargers, I’m cautious in avoiding these devices as well.

It is difficult to ascribe a tape’s noise specifically to the recording environment, the recorder’s transport mechanism, deficiencies in the recorder’s conversion of the electronic signal into magnetization, or anhysteretic magnetization since they can sound very similar.  A recent project involved a particularly noisy tape where the speaking tempo slowed and pitch decreased as the recording progressed, indicating it was made on a recorder that used the take-up reel as the transport mechanism rather than a capstan.  After applying a progressive speed change to the audio file, I obtained a recording with constant tempo and pitch, but it resisted my typical broadband noise reduction approaches.

Inspecting the recording’s spectrograms before and after making the speed change (Figures 1 and 2, respectively), I discovered a band of high-frequency noise that was uniform before the speed change but exhibited a gradual increase in frequency after the change.  If that noise had been caused by any of the factors occurring when the recording was made, the noise should have had the opposite pattern:  gradually decreasing frequency before the speed correction and uniform frequency after the correction.  This was a clear indication (the only one I’ve seen, actually) that the noise was a result of anhysteretic magnetization.  Broadband noise reduction assumes that the noise to be removed is consistent throughout the recording, which is why it had failed.  Applying broadband noise reduction to the original recording and then applying the progressive speed change facilitated my efforts at noise reduction.