Part 2 – 0 Audio tape basics
During the 1950s and 60s, magnetic audio recording tape came very much into its own. It had been around for a long time but it was not robust and the quality was poor. The detail is not for a blog like this.
The modern tapes of the time were 1/4″ wide single track tapes made of a plastic called cellulose acetate. It was strong and flexible and did not easily stretch. If a recorded tape is stretched out of shape and then played back, the sound will be distorted. To make recording tape, small magnetizable metal-oxide particles, suspended in a lacquer that hardens into a flexible layer when it dried, is spread on the surface of the acetate ribbon. I have heard such tape laughingly described as a strip of plastic with rust painted onto it.
Now whilst the particles in the emulsion can be magnetised to remember the music waveform that had been recorded, the magnetism does not spread on the tape as it did in the old wire recorders. This is because the particles are discreet from each other and not contiguous. Tape from the supply reel is drawn over the recording head and taken up on another reel.The head itself is a metal ring cut through at a point to form an extremely narrow gap. A coil of wire is then wound around the head ring and the audio signal that needs to be recorded is passed through the wire. The signal current flow causes a magnetic field across the head gap with a magnitude proportional to the audio signal being recorded. Louder signals cause a stronger current in the head and magnetize the tape more. As the tape is drawn across the head gap, the particles on the tape are magnetised and when that tape is again passed over the head later, it will induce a current into the coil which can be read out as the original music.
All was well in theory but in the original systems the quality was poor due to losses and distortions. For one thing, the amount of magnetism induced onto the tape was not directly proportional to the magnetising force. Small magnetizing signals magnetized the tape to a certain extent but doubling the input signal did not double the magnetization. This led to a distorted copy of the original when played back. The problem became less severe at higher tape magnetization levels and the reproduction was reasonably good in the middle range, or so called linear part of the tape’s magnetization curve. So a way had to be found to use just that linear range of the tape magnetization curve over which the recorded signal remained a fairly accurate facsimile of the original signal. This was done originally using either a fixed magnet or a small DC magnetizing current through the head-coil, to pre-magnetize the tape to the point that it started operating on the linear portion of its magnetization curve. This was called DC bias. The varying signal current was then also passed through the head to create the recording using only the linear part of the magnetization curve. The trouble with that approach is that there is a limit to how much you can magnetise the particles on the tape. At some point the tape is fully magnetized or ‘saturated’. So if the DC bias is too strong, there is not much headroom between where the DC bias leaves the tape magnetization level and and the saturation point of the tape. To fix this, the DC biasing system was replaced by a high frequency AC biasing signal. To understand how this works though we need to understand the importance the head gap and tape speed.
The audio signals we want to record are alternating electrical signals. i.e. they swing from positive to negative and back to positive rapidly. We say they are AC (alternating current) signals. If we draw a graph of a pure musical tone it will be a simple Sin wave, i.e. one that varies in intensity in a constant repetitive fashion in time, it could look like this. The sound we hear when we strike for instance a middle A note on the piano changes polarity 440 times per second. The horizontal axis of the graph indicates the point of zero magnetising current. As the current gets more strongly positive it will rise above the line and as it becomes negative it will be shown below the line. But this graph is a snapshot of a moment in time. If you look at it a moment later it will all have moved a bit to the left. So the graph depicts how positive or negative the current is at various moments in time. Click on the arrow below to hear what it sounds like. . So if you look at this graph you will see that one cycle of the sound or one wavelength if you prefer that term, is from any point on the graph through till when it gets back to the same point and starts repeating itself. If there are 440 repetitions per second as in the case of middle A on the piano, each cycle will last for 1/440th of a second. i.e as the tape moves past the record/playback head, 440 wavelengths will pass the gap each second. It all works fine unless the head gap is too large. If the gap is the size of a wavelength, both halves of the cycle, positive and negative will be in the gap at the same time and the magnetic fields being produced will cancel each other out leaving no recording on the tape. So the gap needs to be very small compared to a wavelength so that it can leave a recorded pattern on the tape which truly represents the original signal. This is what defines the quality of a tape recording head. It is much harder and more expensive to make a head with a very small gap but the smaller the gap, the shorter the wavelength it can record or play back. If the wavelength gets shorter, the frequency (pitch) rises. The sound-wave travels past our ears at a more or less fixed rate so if the wavelength is shorter, more cycle will pass our ears and that means we hear more frequent peaks and valleys. We say the frequency increases as the wavelength shortens. Young people can hear frequencies of 15 000 Hz (vibrations per second) or even higher. That’s a very short wavelength and a head gap has to be really small to reproduce it. But then again it depends how fast the tape moves past the head. if we speed the tape up, each cycle of the recorded signal stretches out on the face of the tape. Or said differently, at high speeds, a wavelength representing a certain musical note covers a longer piece of tape. This means that a given head gap can record higher frequencies if the tape speed is higher.
So then, back to bias. We said that DC bias compresses the useable range of magnetic tape and therefore we get a poor signal to noise ratio and we hear noise on playback. if we use a high frequency signal to pre-magnetize (bias) the tape we have a different result. If the bias current vibrates at several times the highest frequency we need, it will lift the recording on its back onto the linear portion of the tape’s magnetic response but the gap cannot record this high frequency and the bias disappears as the tape leaves the head. This enables the tape to eventually record a much larger range of signal levels than DC bias does.
Its complicated to understand but it actually does work and high quality tape recording uses high speed biasing to improve the quality. This is dry stuff anyway so next time I write on this topic we can move onto video tape recording which is another thing altogether.