Studio Sound Techniques (Part 6)
Tape Playback Alignment
This month we continue our look at how to line up tape machines correctly to give optimum performance during tape playback.
The objective in this part of the procedure is to obtain flat frequency response at the playback output at a known pre-recorded reference level and standard equalisation (EQ). The chosen reference level will depend on the type of compatibility you want or need in order to continue working. In general the music recording industry uses NAB EQ somewhere between 185nW/m and 250nW/m. The broadcasting industry uses IEC or CCIR EQ somewhere between 200 nW/m and 320 nW/m. Don't let nW/m (nano-Webers per metre) frighten you, think of them as being analogous to RMS volts, ie. the signal strength or level. Most disc cutting suites have facilities for either of the above standards at 15 ips. Whatever standard you choose to follow, a full track alignment tape will invariably be more useful than a pre-formatted one since it all head format possibilities eg. full, half or quarter track.
The equalisation in a tape recorder reduces the overall amount of distortion in the recording/playback system. It is important to realise that the signal on the tape is very top heavy, there being 25 dB more at 10kHz than 100Hz. This imbalance is corrected by an opposite curve in the playback electronics to give a flat overall response. Because parameters vary from one type of EQ to another most machines are provided with adjustments to trim the response. Since all the signals are referenced to the playback electronics, it is these that are treated first. So far we have cleaned and demagnetized all parts in the tape path.
Before threading up the calibrated alignment tape switch any RECORD selectors to OFF, SAFE, REPRO or anything that guarantees that you won't be able to erase it in the event that you press PLAY and RECORD simultaneously. Some people even go as far as unplugging the bias/erase oscillator card where this is practical.
Figure 12 shows a generalised arrangement inside the average machine. If you have a manual with your machine follow the manufacturers procedure rigorously otherwise proceed as follows:
1. Playback the section of the alignment tape with 500 or 700Hz on it.
2. a) Adjust the REPRO (Playback.) pre-amplifier gain for 0VU (with the metering switch in the appropriate position), or your own reference level.
b) Or if you wish to have a higher recorded level than that on the alignment tape, reduce this setting by the elevated amount; minus 2, 3 or 4dB for example.
3. Set the line output level to 0VU or as in 2b.
4. Playback the high frequency section 10kHz - 15kHz to set the azimuth. The playback head will have a screw adjustment for this either on the top of the head cover plate or on the head mounting base plate. Adjust for maximum output (more on this later).
5. Set the high frequency playback (for that speed) to the previous mid-frequency (500 - 700Hz) reference level.
6. This completes the playback alignment procedure since the low frequency playback is best done in the record mode. This is because of an effect known as fringing which means that unwanted pick up from an area adjacent to the head track width gives an inflated reading. Although this can be avoided by using a formatted alignment tape, the cost advantage of only having to buy one full track tape for all formats out weighs this slight inconvenience.
So that you can set this correctly it is important to understand why you are adjusting it and what effect this has. One cycle of 15kHz occupies one thousandth of one inch at a tape speed of 15 ips. Figure 13 shows that the opposite extremes of magnetisation occur each half cycle or half of one thousandth of one inch. It follows that maximum output will occur when all of the north poles pass the gap in the head simultaneously. This can only happen when the gap is parallel to the magnetised area or to put it another way; at right angles to the tape flow. Figure 14a shows the correct alignment and 14b an exaggerated incorrect alignment.
In practice you only have to be out by a tiny amount (½ a thou') to have very poor HF response. Such a small amount cannot be checked visually, hence the necessity for the alignment tape. Also there is the danger that if the head is out by a whole cycle a false peak will be observed. This peak will be lower than the true peak and give rise to uneven HF response where different frequencies add and subtract by different amounts. When adjusting azimuth therefore, it is worth covering most of the adjustment range so that you know you have adjusted for the maximum peak. See Figure 15.
A further complication in this area arises in the form of a 'herring bone gap'. This problem is quite difficult to detect without a 'scope or phase meter. If the gap in the head deviates from a straight line, part of the magnetic information will be not-quite-in-phase with the rest. Consequently there will be a plateau in the response rather than a peak. With full or half track mono this is not a big problem, but with stereo it can give a muddy stereo picture since the average pair of ears can discriminate between sounds of differing phase quite well. In the case of stereo (or multitrack) you may observe two (or several) distinctly separate peaks. If replacing the head (or machine) is out of the question for the time being, the response may be optimised in the following manner, using the 15kHz section of the alignment tape.
1. Adjust the azimuth so that a majority of the tracks are at a peak, making a note of which tracks are in phase.
2. Adjust the high frequency playback to give an equal reading on all tracks.
3. Mix equal amounts of all tracks onto a group output meter being careful not to overload the mixing bus amplifier by turning down the line input gain control on each channel.
4. Adjust the azimuth over a wide range to ensure you have found the true peak.
5. Optimise the individual readings on the machine with the combined output on the mixing console.
6. Let the alignment tape play through to the end and store the tape 'tail out'.
When recording vocals, drums (cymbals), percussion or anything with a lot of high frequency content try and keep to the tracks on the machine with the best response.
The objective in this part of the procedure is to get back from the machine signals at the same level as those sent. In this sense an ideal machine would be 'transparent'. In practice we have to optimise the response.
1. Place a fresh reel of tape from the batch you intend to use on the machine.
2. Apply 500Hz to the input at a level of 0VU (+4 dBm). Maintain this level throughout.
3. Switch the output meter to 'record input' and set the record gain for 0VU.
4. Switch the output meter back to the reproduce output position and adjust the reading to -2 VU temporarily. (Note the delay in seeing any change you make due to the distance between the record and playback heads).
5. Change the input frequency to 15kHz and adjust the record head azimuth for maximum output, following the previously mentioned guide lines.
6. Change the input frequency to 10kHz and turn the bias controls to a minimum (usually counter clockwise).
7. Increase the bias until a peak is reached. Continue increasing the bias until the reading has dropped 3 to 4 dB from the peak value. The machine is now properly biased.
8. Change the input back to 500Hz and set the record level so the output reads 0VU.
9. Change the input to 12kHz and adjust the record high frequency for 0VU at the output.
10. Change the input to 50Hz and adjust the 'playback' low frequency for 0VU.
11. Record 1 minute of 50Hz, 500Hz and leave a few minutes of blank tape after these tones. That way if you take your tape to another studio the maintenance man will have something to bias his machine with and set his playback electronics.
12. Switch the track selectors to SYNC and adjust the 'sync level' on the previously recorded 500Hz tone, the 'sync high frequency' on the 10kHz tone and the 'sync low frequency' on the 50Hz tone. These should all be set for 0VU.
13. As a final check put all the tracks into the record mode and sweep the entire audio range (30Hz to 16kHz) to see that the overall response is flat whilst monitoring the output audibly.
The erase setting differs from machine to machine, however, they all aim to reduce a pre-recorded signal to a minimum. If you are in any doubt about your erase settings use the previously recorded 30Hz - 16kHz. By erasing these frequencies and listening to the erased tape you will be able to spot any weak erasure or erase noise indicating a faulty bias/erase oscillator.
Feature by P.A. Becque
Previous article in this issue:
Next article in this issue: