This article needs additional citations for verification. (December 2009) (Learn how and when to remove this template message)
Print-through (sometimes referred to as bleed-through) is a generally undesirable effect that arises in the use of magnetic tape for storing analogue information, in particular music, caused by contact transfer of signal patterns from one layer of tape to another.
Print-through is a category of noise caused by contact transfer of signal patterns from one layer of tape to another. Print-through can take two forms: 1) thermo-remanent magnetization induced by temperature, and 2) anhysteretic magnetization caused by an external magnetic field.
The former is unstable over time and can be easily erased by rewinding a tape and letting it sit so that the patterns formed by the contact of upper and lower layers begin to erase each other and form new patterns with the repositioning of upper/lower layers after rewinding. This type of contact printing begins immediately after a recording and increases over time at a rate dependent on the temperature of the storage conditions.
The audibility of print noise caused by contact printing depends on a number of factors: 1) the amount of print due to conditions of time and storage; 2) the thickness of the base film that acts as magnetic barrier (thin C-90 cassette tapes are more susceptible than studio mastering tapes that use a base film four times thicker); 3) the stability of the magnetic particle used in the tape coating; 4) the speed of the tape (the wavelengths of the prints shift so that higher speeds move printed signal closer to the range where the ear is more sensitive); the dynamics of the musical program (very quiet passages adjacent to sudden loud signals can expose the print signal transferred from the loud signal); and the wind of the tape (A-winds for cassettes with the magnetic layer facing outward have stronger print signals after a loud signal--"post-print"--than B-winds used in modern open-reel recorders that have stronger "pre-print" signals preceding a loud passage. echo.
Tape speed is a factor because of the shift in wavelengths. For example, the strongest print signal on a C-60 cassette running at 1.875 ips is about 426 Hz (605 Hz for a C-90), while an open-reel tape recorded at 7.5 ips would have its strongest signal at 630 Hz if the tape were a professional tape with a 1.5 mil base film or 852 Hz if the tape were a consumer version with a base film of 1.0 mil thickness.
The cause of print-through is due to an imbalance of magnetic and thermal energy in the magnetic particle. Once the magnetic energy is only 25 times greater than the thermal energy, the particle becomes unstable enough to be influenced by flux energy from the layer above or below the tape. The amount of magnetic energy depends on the coercivity of the particles, their shapes (long, thin particles make stronger "magnets"), the ratio of ideally shaped particles to defective particles, and their crystalline structures. Metal particles, although very small, have very high values of coercivity and are the most resistant to print-through effects because their magnetic energy is seldom challenged by thermal energy. Particles fractured by excessive milling prior to coating will increase levels of print depending on their ratio compared to their well-formed neighboring particles.
Anhysteretic print signals are almost as strong as intentionally recorded signals and are much more difficult to erase. This type of print noise is relatively rare because users are typically careful about accidentally exposing recordings to strong magnetic fields, and the magnetic influence of such fields decreases with distance.
Digital tapes can also be affected by contact print effects in a phenomenon known as "bit-shift" when upper or lower layers of tape cause a middle layer to alter the pulses recorded to represent binary information.
Since analog video is recorded by frequency-modulation of the video signal, the FM capture effect shields the signal against this noise; however, the linear audio and (depending on format) chrominance signals of a video cassette may have some print effects.
While print-through is a form of unwanted noise, contact printing was used deliberately for high-speed recording of video tape. DuPont invented a form of thermal magnetic duplication known as "TMD" by which a high-coercivity metal tape was brought into direct contact with a chromium dioxide copy tape. The recording on the metal tape was a mirror image of the video signal. Immediately before the chrome copy tape came into contact with the mirror mother master, a focused laser beam heated it to its Curie point at which its value of coercivity dropped to very low values so that it picked up a near perfect copy of the master as it cooled. Sony developed a system known as "Sprinter" that used a similar master source tape forced into close contact with any blank copy tape and run across a transfer head in which a weak AC high frequency sine wave is used to transfer the information anhysteretically to the blank tape with minimal erasure of the source tape on each pass. Sprinter master tapes did suffer enough loss that they had to be replaced after a number of passes. This form of high-speed recording was very cost effective when recording in the EP (extra long play) mode because it was three times faster than recording in SP (standard play) mode while real-time recording took the same amount of time whether in EP mode that used less tape or SP mode that used a greater amount of tape. High-speed video recording of EP video produce far more consistent results than real-time recording at the slowest VHS speed.
- Audio Engineering Society. Technical Bulletin A011194.
- C.P. Bean and J.W. Livingston, "Superparamagnetism," Journal of Applied Physics, April 1959, pages 120S-129S.
- William Manley, "Thinking about Print-Through," Audio Magazine, September, 1977, pages 55–85.
- Jay McKnight, "Tape Print-Through Reduction," Ampex Research Department Report 106, November 1957, 12 pages.
- Terence O'Kelly, "BASF Inventor's Notebook--Print-Through," Bulletin No. 9, August 1980.