Tape backen & Kopiereffekt

[Peter Ruhrberg]

Hallo Kai,

Einige weitere Überlegungen zur Verringerung des Kopiereffekts habe ich in der englischsprachigen Literatur gefunden.


Edward P. Koeppe, RCA Victor Record Division, The Measurement and Minimization of Print-Through of Magnetic Sound Recording Tape, AES Preprint 10/1964

A method used by many recording engineers to reduce print is to store the recorded tape wound oxide-in with the start of the recording on the inside of the take-up reel. This will minimize pre-print, and permit the post-print to be masked by the printing signal. If the printing signal is a sharp crescendo followed by a quiet passage the post-print will be heard, but in many instances will sound like a type of reverberation. Storing the tape in this manner requires rewinding before playback. Rewinding the tape will permit time to elapse, and reduction of the print to take place.

The use of high frequency bias during playback of a recorded tape will effectively reduce the printed signal without serious degradation to some recordings. Proper application of bias makes it possible to salvage recordings that have excessive print. This method has received consideration before, but requires further discussion. If a recorded tape is played back with high frequency bias in the record head equal to 60% of the value of peak bias normally used in recording, a reduction of 10 dB in print is made on most tapes. Some loss in high frequency response results, but this is not drastic. The output at 10 kHz is reduced by 1,0…1,5 dB, and 15 kHz is reduced 1,5…9 dB at a tape speed of 15 ips. This reduction of high frequency response will be the same regardless of signal level. The use of bias has no effect on the low or mid-range of frequencies, or on the noise characteristics of the tape. Additional passes reduce the high frequency response by 0,5 dB per pass. The addition of bias into the record head increases the distortion of a 5 kHz signal to 5% that initially measured 3%. From 1…5 kHz there is less than a 2% increase in distortion while there is no increase in the distortion of the lower frequencies.

Another method that may be used to reduce the level of the printed signal is to increase the time that the printed signal is separated from the printing signal. … Increasing the length of the tape path between the supply reel and the playback heads will reduce the print by several dB. When playing back a recorded tape at 7,5 ips, an increase of 30 inches in tape path will reduce the printed signal approximately 2 dB.

Peter Copeland, Manual of Analogue Sound Restoration Techniques, 2009 S. 169f.

… there are several factors which influence the effect, including the following:

1. Print-through increases with time in storage (roughly logarithmically), and with the absolute temperature.

2. Excessively high peak volumes can overcome the initial step in the B-H curve, and cause disproportionate print-through.

3. Physical vibration (e.g. dropping the reel) can increase print-through, as can strong magnetic fields (don’t store tapes near a lightning-conductor or a loudspeaker).

4. Print-through is inversely proportional to the thickness of the tape base, and its efficiency is raised when the wavelength of the sound is the same. For practical professional open-reel tapes, human vowel sounds are worst affected.

5. Rewinding the tape before use helps the print-through to self-demagnetise by a few decibels.

6. Storing the tape tail-out when oxide-in (and head-out when oxide-out) takes advantage of the curved shape of the magnetic lines of force to reduce the effect slightly, so that pre-echo is quieter than post-echo (which is more natural).

7. When the tape was recorded, some reciprocal noise reduction processes could have helped, not to mention suitable choice of tape. Unfortunately, tapes with the best inherent power-bandwidth product tend to have worse print-through.

“What can we do to minimise print-through on playback?” For many years, the standard answer has been to power the erase-head of the tape reproducer at a very low level, because the print-through tends to be below the “knee” of the B-H curve and is easier to erase. However, this may also erase some of the wanted signal, especially at high frequencies. Since it can only be applied to the original (not a copy), I consider it important to transfer the tape twice, once before the treatment and once after, to be certain we do not lose anything. I am told the optimum usually occurs when the erasing current cuts the highest wanted frequency (about 15kHz on 38cm/sec tape) by about one decibel.

... some experiments done by the BBC Research Department in 1953 showed that a better cure was to run the tape past a weak erasing field applied through the back of the tape. This does not affect the high frequencies so much, since they are confined to the outer layers on the oxide side of the tape.

Grüße, Peter