Nagra 4.2 Portable Analogue Audio Instructions Manual
Have a look at the manual Nagra 4.2 Portable Analogue Audio Instructions Manual online for free. It’s possible to download the document as PDF or print. UserManuals.tech offer 1 Nagra manuals and user’s guides for free. Share the user manual or guide on Facebook, Twitter or Google+.
The Nagra 4.2 fitted with a QSLI can be used to measure the playback pilot signal. 0.85 V corresponds to a meter deflection of 1 V on the scale calibrated from 0 to 2 V and normally used to measure the voltage per cell of the batteries. 1.7 V corresponds to full scale deflection. PILOT CONNECTORS On the Nagra 4.2, the pilot playback signal is to be found on the Power Pack connector (48), which is a 6 pin Tuchel socket, located on the right-hand side of the recorder. Pin No 2 is connected to the chassis and the output pilot signal is connected to pin No 3.
6.0 BASIC THEORY CONTROL OF THE INPUT SENSITIVITY (MODULATION),DYNAMIC RANGE, SIGNAL-TO- NOISE RATIO, DECIBELS The dynamic range is the ratio between the loudest and softest sound levels. The dynamic range is large for a symphony orchestra compared to that of an announcer reading a news bulletin. The signal-to-noise ratio is related to the dynamic range. It is important that the softest sound level to be recorded is still considerably stronger than the noise. Thus, sound with a large dynamic range requires a high signal-to-noise ratio. However, this ratio can be practically equal to the dynamic range in the case where the noise level is close to the threshold of audibility. The subjective perception of the sound level follows a law, which is approximately logarithmic. It is for this reason that it is customary to measure sound level as a logarithmic unit. This is the decibel (dB). Each time the sound power is multiplied by 10, the number of decibels, which that represents is increased by 10. Thus an increase of 100 times equals 20 dB, a 1000 times equals 30 dB etc. It should be remembered that the power is proportional to the square of the amplitude. The voltage, which a microphone gives, is proportional to the amplitude. In other words if the voltage increases 10 times, the power increases 100 times and corresponds to 20 dB. The decibel is a measure of power ratio and not an absolute value. In taking as a reference, a sound corresponding to a variation of pressure of 2?10-4 µbar (value considered as the threshold of audibility at 1 kHz) a scale in absolute value will be obtained. A sound of 90 dB will therefore mean 90 dB above 2?10-4 µbar. The frequency response of the human ear varies with frequency. In order to compensate for this, the sound level should be measured with filters simulating the variations of sensitivity of the ear. Thus the decibels become the phon referred to 2?10-4 µbar. The potentiometer scales of the Nagra 4.2 are graduated in decibels referred to 2?10-4 µbar. At 1 kHz, these decibels are the same as phons but as the Nagra does not have psophometric filters, it cannot be considered as a phon meter. With a potentiometer control placed on X dB, a sound of X dB, captured by a normal microphone (0.2 mV/µbar into 200 ?) and attacking a normal sensitivity preamplifier, produces a recording at nominal level. The modulometer will indicate 0 dB. COMPRESSION OF THE DYNAMIC RANGE The ideal installation for recording and reproduction should restitute exactly the sound levels, which have been recorded. The listener should hear exactly what the microphone heard. The human ear has a dynamic range of more than 120 dB. The Nagra 4.2 has a signal-to-noise ratio, which is exceptionally high. However, this ratio can only just reach 70 dB. An amateur tape recorder should, according to the DIN standard, reach 45 dB. It is clear that the ideal installation is not possible without compressing the recording and expanding it again on playback. Listening to a signal with a dynamic range of 120 dB poses some practical problems. The ambient noise of an apartment or a cinema auditorium is considerably greater than 0 phon. 120 phons becomes painful to listen to. Therefore, apart from exceptional cases, the listening dynamic range should be reduced. The choice of this dynamic range and, in consequence, the degree of compression is one of the essential tasks of the sound engineer. Classical music discs designed to be listened to on a Hi Fi chain can have a very high dynamic range. A chamber orchestra can be recorded practically without compression. A symphony orchestra should be slightly compressed, and this is done with the music score, and requires a good musical culture. A transmission designed to be listened to on a Hi Fi chain can have a very high dynamic range. Practically, everything should be at maximum level. On television, the dynamic range can be fairly high. At least in those countries where habitation in individual houses is dominant. Apartment blocks limit the maximum power. In any case, evening transmissions should have a lower dynamic range, the listening level being considerably reduced, but the pianissimo should still be audible. It is true that the ambient noise level is also reduced during the evening. In cinema work, the dynamic range depends upon the public for which the film has been made. In certain countries the cinemas are very noisy. A comedy film provokes laughter, and this should be taken into account. The dialogue following a joke should always be at a high level otherwise it will
be drowned in the noise of the auditorium. On the other hand, a suspense scene permits the use of very low sound levels. Generally, for dialogue an effect can be obtained not by the absolute level of the sound but by the contrast. A burst of sound will be much more effective if it is preceded by a passage at a moderate level. This trick is well known amongst cinema mixers -the level is lowered before a forte. WHEN SHOULD COMPRESSION BE DONE? A) Recordings indented to be transferred on to a disc. The signal-to-noise ratio of the modern disc is excellent, but it is important that the noise level of the tape should not be transferred. If a compression is decided upon, it should be done at the time of the original recording, otherwise an increase of the pianissimo will also increase the sound level of the tape noise. It is difficult to use the complete dynamic range of the recorder without the risk of exceeding the maximum level when a fortissimo is produced. For this reason, it is prudent to work simultaneously with two or three tape recorders in parallel, but whose input sensitivities are varied by a few decibels. The tape which has been recorded at the highest level, but without the maximum level having been exceeded, will be the one used for playback. Also, it will be possible to choose, during editing, certain passages from tape No 1 and others from tape No 2 etc. B) Recordings intended for radio transmission. The compression should be done at the time of recording. For reporting ect., the use of an automatic level control can be of interest. It gives a tendency to always obtain the maximum level, that is, it compresses to a large degree. C) Recording intended for radio transmission to be reworked in a studio. In this case, two methods are possible. The signal-to-noise ratio of the Nagra 4.2 is greater than that of the radio transmission; therefore it is not essential to use the complete dynamic range of the Nagra. It is possible, to adjust the sensitivity in such a way that the fortissimo reaches 0 dB. As the maximum level of the Nagra is +4 dB, there is, therefore, a safety margin. The compression can be done according to the needs in the studio whilst working on the final recording. D) Cinema and Television, where the sound is always edited during the final mixing. The important thing is to preserve the maximum amount of sound information. The very large dynamic range of the Nagra 4.2 allows the fortissimo to be recorded at a level below the maximum, avoiding accidental distortion due to a burst of sound. In many cases, it may even be desirable to work on automatic level control, but this decision depends on the circumstances, and those who have to make the decision need no advice. The problem of microphone and preamplifier noise should be considered. Very often, the background noise of the recording is not dominated by that of the tape, but by that of the microphone. In these cases, it is useless to increase the sensitivity during recording. The general level will be greater but so will that of the noise level. Nothing is gained in information, but the risk of saturation by a loud sound will be uselessly increased. The point above which it is useless to go is around the 80 dB mark on the potentiometer scales. This can be easily verified: replace the microphone by a resistance equal to the nominal impedance of the microphone to ensure that ambient noise does not upset the measurement. Record and playback simultaneously (Line and Phones switch on Tape), listen with good headphones and increase the microphone sensitivity. Even with the potentiometer in the extreme left hand position a noise will be heard. Turn the potentiometer clockwise. Up to 90 dB on the scale, the noise level will hardly vary. As from 80 dB, the noise of the resistance replacing the microphone and of the preamplifier becomes dominant. This point varies according to the quality of the tape used. With a poor tape it can be 78 dB, whereas with an excellent tape 82 dB. It is also to be supposed that the playback will be made on a Nagra 4.2 or on a machine with similar performance. If an ordinary machine should be
used, not having a sufficiently quiet playback chain, it may be desirable to increase the input sensitivity above 80 dB. On the other hand, the problem is completely different if the tape produced should be used without reworking. In this case, it is necessary to compress according to the needs even if the noise level of the microphone considerably exceeds that of the tape. For these applications, there is a special range of high gain amplifiers available (see section 5). 5.3. INTERFERENCE To obtain a good signal-to-noise ratio, it is not sufficient to place the microphone well, it is equally important that no interference can be introduced into the system. An explanation showing how this interference occurs will be given and also the means to eliminate it. ELECTROSTATIC INDUCTION The microphone and the cable, which connects it to the Nagra and the plugs, should all be well shielded. If these conditions are respected, no electrostatic induction can occur. However, certain parts of some microphones are grounded by a simple contact, which is not protected against corrosion. Often the plug shielding is similar, and certain cables have only a symbolic shielding. In these cases any electrostatic field can induce interference voltages into the system, in particular if the Nagra is not grounded but is connected to a camera, which is not interference suppressed and the entire system is at a voltage above ground. Hence, the ground itself, as seen by a cable, is a potential source of interference. In other words, interference is introduced at the slightest defect of the shielding. The induced interference can be at an audio frequency or a high frequency, which can be detected within the recorder. ACTION TO BE TAKEN AGAINST ELECTROSTATIC INDUCTION 1. Good shielding. Above all, check the plugs. 2. Avoid the conditions where the Nagra is floating above ground with an interference voltage. One possibility is to use a photoelectric coupler between the camera and the Nagra. Obviously, no trouble can arise if quartz crystal synchronization is used. 3. If, for any reason, it is not possible to follow the advice given, it is possible to reduce the interference level by: a) Using shielded input transformers, which attenuate the passage of indirect interference. b) Using symmetrical input (with the middle point grounded, which will attenuate the passage of direct interference. c) In the most hopeless cases (such as with a radio transmitter close by) add external filters. It should be noted that the microphone preamplifiers of the Nagra 4.2 are already fitted with filters, but their action only begins at around 500 kHz, because they are, above all, designed to reduce very high frequency interference for which the usual shielding is ineffective.
MAGNETIC INDUCTION Certain microphones are sensitive to magnetic fields and they should never be placed close to any motor, or transformer etc. The cables being double and twisted, means that voltages induced should cancel one another, whether the input is symmetrical or not. It is essential, of course, that the microphone should be floating, that is to say, that neither of its output wires should be grounded (except in the case where the shielding should be connected to the shielding of the cable). The only case where a magnetic induction can be dangerous is where a microphone cable runs along-side a power cable. The latter radiates a magnetic field, which is not homogeneous, and any irregularities in the twisting can suffice to induce interference voltages. CABLE PREAMPLIFIERS It would seem somewhat illogical to have to take great precautions to transmit a signal as weak as that given by a microphone when it would be easy to amplify the signal close to the microphone, and thereby transmit a higher voltage. This can be done with a cable preamplifier. This accessory is placed close to a dynamic microphone and gives an output voltage similar to that of a condenser microphone. The power supply requirements are also similar, so that it is possible to interchange a condenser microphone with a dynamic microphone fitted with a cable preamplifier. Under these conditions the Nagra should also be fitted with a plug-in preamplifier designed for a condenser microphone. As the power supply requirements vary according to the type of condenser microphone, so there exists a corresponding range of cable preamplifiers, to be fed from each type of preamplifier. 5.4. VOLTAGE OR CURRENT FEED By altering the negative feedback, it is possible to adjust the input impedance of any preamplifier to practically any value. If the input impedance is high, the microphone will not supply any current, and only the voltage will be used to transmit the signal. This is called a voltage feed. If the impedance seen by the microphone is very low, the voltage at the microphone terminals will remain negligible, but the microphone will supply a current, which will transmit the signal. This is called a current feed. A dynamic microphone whose impedance is constant as a function of frequency can be used indifferently for current of voltage feed. Current feed offers certain advantages: the performance of the input transformer has much less influence on the overall result, the noise level is minimum when the input is open etc. This last point ensures that a recording will not be spoilt if an unused channel is left with the level control open. Therefore, when there is a possible choice, current feed would seem to be preferable. Unfortunately, microphones with a cardioid characteristic possess an internal impedance which varies greatly with frequency and they can be used only for voltage feed. Thus the standard pre-amplifiers for the Nagra 4.2 use the voltage feed method, but current feed preamplifiers are available for special cases, where desired. FILTERING PREAMPLIFIERS In a large number of cases, it is desirable to attenuate very low frequency signals captured by a microphone. As the human voice contains practically nothing below 50 Hz, a flat frequency response down to 20 Hz is not only useless, but can be objectionable because low frequency noises can perturb the recording chain. The filtering is usually done at the final mixing, but if it is known that there are very low frequencies, which should be eliminated, it is better to do it within the preamplifier. Several models of plug-in preamplifier are available for doing this correction. In the preamplifier code, the figure following the letter Y indicates the attenuation of a signal at 50 Hz in decibels.
MAGNETIC HEADS There are four tape heads on the NAGRA 4.2, these being ERASE, RECORD, PILOT and PLAYBACK (see CONTROLS at the front of this manual for their location). Contact between the magnetic head and the tape must be perfect. Some tapes leave deposits on the heads. Fortunately, the deposits are clearly visible. A dirty playback head gives a muffled sound, lacking high frequencies. If the high frequencies come and go rapidly (one to ten times per second), the azimuth needs adjusting. A dirty recording head will record at a low level, and the sound will be distorted. Under the same conditions an erase head will give poor erasing. To remove the deposit, it is necessary to soften it with the aid of a solvent. The simplest method is to take some absorbent cotton or a rag, soaked in alcohol, or even water and to lightly rub. Chlorinated solvents, such as trichlorethylene, should be used with care, as it is possible they may lightly attack the resins used in the construction of the heads. ADJUSTEMENT OF HEAD AZIMUTH Theory The recording and the playback of a magnetic tape is a function of gaps in the recording and playback heads. These gaps should make a certain angle, arbitrary in itself, with the tape, but which should be the same for recording and for playback. Any angular difference between the recording and playback heads will produce a loss of level. This phenomenon becomes more important as the wave length (that is to say the ratio between the tape speed and the recorded signal frequency) becomes short. Practically, a faulty azimuth gives muffled recordings without high frequencies. To ensure the interchangeability of tapes, the azimuth angle has been standardized: the angle between the gap and the tape should be 90?. Special recorders, whose heads have been optically aligned, have been constructed in order to produce Standard Tapes whose purpose is to permit the adjustment of the azimuth on ordinary tape recorders. It is to be noted that if a recorder has been used for recording with a badly adjusted azimuth, it is perfectly possible to save the recordings: it is sufficient to adjust the playback head in consequence. This can be done by means of the ear, by orienting the playback head to obtain the sound, which is richest in high frequencies. This method is evidently also applicable to tapes, which have been deformed by a faulty spooling, or by climatic conditions. If the tape is sabred, or curved the notion of azimuth becomes delicate, and depends upon the relative positions of the heads and the guides. Parenthetically, it is these problems, which limit the use of very low speeds, below 7½ ips, because it is difficult to ensure a sufficiently precise azimuth unless only a narrow track is used. In effect, the error tolerance of the azimuth increases as the width of the track decreases. In reducing the width of the track, so is the signal-to-noise ratio reduced. Thus the solution of low speed and narrow track is, above all, used nowadays for amateur machines. Variation of the High Frequency Level with Azimuth Error When the azimuth is very slowly adjusted past the optimum position, the high frequencies are reduced at first very slowly, then their attenuation is accelerated the further the adjustment is made from the correct point. A curve, which represents this attenuation as a function of the angle of error, has a rounded summit, and sides more and more steep. This is important, because if the azimuth is adjusted simply in looking for a maximum, it is quite possible that instead of being at the summit of the curve-that is to say at the optimum point-it could be at one side or the other. If the errors of the playback head and the recording head are additive, the tape could be recorded out of tolerance. The recording head is adjusted by referring to the playback head. Thus the azimuth of the recording head has the sum of the errors: that which is produced in adjusting the recording head, and the playback head. It is therefore important that the azimuths should be adjusted as closely as possible to the top of the curve. This is possible by looking for two points about the maximum, which correspond to a certain attenuation of the high frequencies, and then adjusting to the midpoint.
Secondary Maxima If the misalignment of the azimuth is continued at the same time as observing the playback of a high frequency signal, it can be seen that after having passed by a minimum, the signal will increase again to pass by another secondary maximum. If the principal maximum corresponding to the correct angle applies equally to all frequencies, the secondary maximum applies only to one frequency, which is fixed for the particular conditions. If this frequency changes, the position of the secondary maximum is displaced. It is clear that the secondary maximum corresponds to an adjustment which can not be used and which should be avoided. If it is necessary to adjust the azimuth, it should be done by very small degrees, so that no risk exists of reaching a secondary maximum. On the other hand, if it is necessary to adjust the azimuth from scratch, it is better to make the adjustment with a fairly low frequency (1, then 3 kHz) to obtain an approximate azimuth position. At these frequencies, the secondary maxima are outside the range for adjustment of the heads. Orientation of the Heads on the Nagra 4.2 The heads of the Nagra 4.2 are pressed down on a cam. On turning this cam, the azimuth is varied. The outside of the cam is in the form of a gear wheel meshed with a pinion, which is visible in front of each head. The pinion can be turned by means of 2½ mm Allen key. The Allen key should be demagnetized before using it as a magnetized tool can induce a very low frequency into the playback head, which could upset the adjustment. Bias To record, that is to say, to magnetize a magnetic tape, it is necessary to submit it to a magnetic field, which passes a certain threshold value. Below this value, no permanent magnetization will be produced. To reach the threshold, and to pass into the linear part of the magnetization curves, the audio frequency signal is superimposed upon it. The peaks of the high frequency signal always make an excursion into the linear region. The low frequency signal determines, in effect, to what point the excursion will be made. This is called high frequency bias. Its amplitude influences notably the quality of the recording obtained, and the determination of its level should be made precisely. Effect of Bias Signal Amplitude on the Recording If a low frequency signal (400 Hz) is applied to a recording head whose bias level is varied, several effects can be observed. A low level bias signal will give a weak distorted signal on playback. As the bias level increases, so the level of the signal increases, rapidly. A maximum will then be reached, after which the signal level will be very slowly reduced. The maximum can be called the Point of Maximum Efficiency. It corresponds also to the point where the distortion is the minimum. The fact that the signal becomes too great, renders the determination of the optimum point rather difficult. A high frequency signal (e.g. 10 kHz) will give its maximum level for a bias level noticeably lower and which corresponds to a point where a low frequency signal would become distorted. This is due to the fact that the magnetic layer of the tape is not infinitely thin. The point of maximum efficiency for a low frequency corresponds to an optimum recording throughout the whole of the magnetic layer. The outside part will in fact be over-biased and, to a certain extent, even partially erased. The middle of the layer is further away from the heads, hence the loss of high frequencies. It is essential to remember that in over-biasing, not only is the efficiency of the recording of the high frequencies diminished, which can be compensated for by an increase in the recording current, but the playback signal is attenuated, thereby showing saturation of the magnetic tape. On the other hand, an over- bias will lower the noise level of the tape.
High Frequency Pre-emphasis The signal-to-noise ratio of the magnetic tape is perhaps the least satisfactory of its characteristics. Great efforts have been made to improve this defect. It is possible to imagine, for example, a tape recorder, which sends to the recording head a current, which is proportional to the input signal, independent of the frequency (recording at constant current). Experience shows that the tape becomes saturated for a given current in the recording head irrespective of the frequency. At high frequencies the saturation takes on special characteristics. The harmonics, which the saturation should produce, go out of the range of the spectrum, which the playback head can reproduce. Therefore, a tape saturated in the high frequencies does not give a distorted signal. Simply, an increase of the recording current does not produce an increase of the recorded signal. In effect, the tape becomes a limiter, which, in addition, alters the sonority of the recording. A tape recorded under these conditions (constant current) should be played back on a head followed by an amplifier fitted with frequency response correctors so that the ensemble will be linear. It can be seen that with the sounds, which are normally recorded, the level of the high frequencies is noticeably lower than that of the middle frequencies. To be exact, the peaks of the high frequencies can have a large amplitude but their duration is very short, and a limiting will pass unnoticed. From the idea of emphasizing the high frequencies during recording and to attenuate them during playback, the noise level of the tape, which is annoying above all in the high frequencies, is effectively reduced. This is known as pre-emphasis. It is used universally in disc recording and frequency modulation radio transmission as well as in magnetic recording. This universality is very important, for if there is a pre-emphasis in any link of a chain, it is useless not to have it in the other links, because, in any case, the high frequency peaks will be limited in the link which has the strongest pre-emphasis. On the other hand, the gain in the signal-to-noise ratio is preserved in each link. As a summary, the recourse to pre-emphasis is universal, as it has been found that the possible limiting of high frequency peaks is less annoying than the high noise level without pre- emphasis. How much pre-emphasis can be accepted? The question is complex, for it depends upon the type of sound to record. The sound spectrum varies with different languages, and it is for this reason that the standards for pre-emphasis vary from one country to another. For practical reasons, it is not the pre-emphasis, which has been standardized in the case of magnetic recording, but the playback chain. The recorder should be adjusted so that a tape produced on it and played back on a standard playback chain should have a linear response, this is the same as standardizing the pre- emphasis for a given type of tape. There are tapes whose capacity for recording high frequencies is noticeably higher than that of standard tapes. To record on these tapes, according to the standard, it is necessary to have a lower pre-emphasis. Relation between Pre-emphasis and Bias The American NAB standard at 7½ ips requires a greater pre-emphasis than the European CCIR standard. In Europe, it is normal to slightly over-bias the tape. This gives a slightly better signal-to- noise ratio, but reduces the recording level of the high frequencies. The final result is practically identical to that obtained with the NAB standard without over-bias. The stronger pre-emphasis of the NAB standard gives approximately the same improvement in signal-to-noise ratio and the tapes become practically saturated at the same high frequency signal level. The NAB standard relies upon a heavier pre-emphasis and the CCIR on the higher bias level.
Practical Conclusions The result is: a) It is possible to modify the pre-emphasis, within certain limits, by adjusting the bias level but still remaining within the limits of the standards. b) To record sounds particularly rich in high frequencies, it is possible that the use of tapes, which permit high recording level of the high frequencies, could give better results. c) It is necessary to determine which link in the chain gives the greatest pre-emphasis. If all links pre-emphasize to the same degree, this will produce the most rational chain. However, if one link becomes saturated, it is better for this to be the magnetic tape, for the saturation of the high frequencies does not lead to audible distortion, which is not the case with a frequency modulation transmitter (or rather, the corresponding receiver). DETERMINATION OF THE BIAS LEVEL Tape Characteristics The tapes on the American market are very similar to one another from the point of view of the optimum bias level. This permits the adjustments to be made very close to the point of maximum efficiency. It would be undesirable to adjust directly on to this point, as an over-biasing is much less dangerous than an under-biasing. This working point close to the point of maximum efficiency is very convenient for the NAB standards, as shall be shown. In Europe there is a greater range of tape characteristics. In over-biasing the ordinary tapes, the working point remains correct for the tapes, the working point remains correct for the tapes at a high bias level. This ties in with the CCIR standards and gives good results. General Procedure It is necessary to use a reference tape whose characteristics are well known, above all in relation to other tapes on the market. The normal reference tape is CCIR PER 368 and NAB 3M 808. It is necessary to determine the bias level, which gives the greatest efficiency. The signal used will be of fairly low frequency (400 Hz). To locate the peak of the curve more easily, two points, E1 and E2, should be looked for. Point E1 has an underbias level which gives lowering of the playback level by 1 dB; E2 is an over-bias level, which gives a lowering of the playback signal by 0.5 dB. The asymmetry of the form of the curve justifies the difference in the playback levels of E1 and E2. It is evidently necessary to use a sufficiently regular tape so that variations of the sensitivity should not be confused with the loss of level due to under and over-biasing. The maximum efficiency bias level Em will be the geometric mean of E1 and E2. Multiplying E1 by E2 and taking the square root of the product find this. The working point will be Em?k where k is the coefficient of over-biasing. Variation of k The preceding rules take into account the dispersion of the characteristics of available tapes. If a recorder is only used with one specific type of tape it is possible to use a value of k, which will be optimum for the conditions. Values of k from 1 to 1.3 are possible. A small value of k can be used if the sounds to be recorded are rich in high frequencies, or if the tape is of a low quality for high frequencies. A value of k above average can be used in the opposite case. The value 1.2 is acceptable for the CCIR standard, and 1 to 1.1 for the NAB standard (50 ìms at 7.5/second).
7.0 CALIBRATION AND CARE HEIGHT OF THE NEOPILOT HEAD The middle head of the Nagra 4.2 is used to record and playback the pilot signal. Its azimuth is not critical, but its height should be correct. The cam of this head does not vary the angle, but only its height. Before adjusting the azimuth, check and adjust (if need be) this head. Adjust the height so that the tape is exactly in the middle of the head. The eye is sufficiently accurate for this adjustment. AZIMUTH ADJUSTMENT OF THE PLAYBACK HEAD. Playback a Standard Tape at 7½ ips. Place the Line and Phones switch in the Direct position. The modulometer will thus indicate the playback level. Check that the microphone potentiometers are in their extreme anti-clockwise position, and adjust the Line and Playback potentiometer to give a convenient playback level. Standard Tapes are generally recorded at between -10 and -20 dB, so that, in general, it will be necessary to put this potentiometer in the maximum position, or nearly so, to have a playback level of about -10 dB. Introduce the key into the pinion of the playback head and find the maximum playback level. After this, look to left and right for the points where the signal level is lowered by 1 to 2 dB, and place the pinion midway between these points. A correct adjustment corresponds equally to a stable playback. Errors due to sabring of the tape are hardly perceptible at the top of the curve, and only become important on the sides. Once the playback head has been correctly adjusted for azimuth it should not be touched again. AZIMUTH ADJUSTMENT OF THE RECORDING HEAD. Two methods are possible: 1. Standard Method To use this method, it is necessary to have an audio frequency generator, which can give 1, 3, 10 and 15 kHz and an AF voltmeter or an oscilloscope. It is possible to use a second Nagra as a voltmeter, by introducing the output signal of the recorder to be adjusted into the line input of the second one, and using its modulometer as a voltmeter. Procedure: a.) Introduce a signal from the generator into the line input of the Nagra, and adjust the level to -15 dB on the modulometer. b.) Connect the voltmeter to the line output. The signal to be measured will be approximately 0.8 V R.M.S. The Line and Phone switch should be put into the position Tape. c.) Thread up a tape and record. The played back signal should be observed on the voltmeter. d.) Start with 1 kHz and increase the frequency. When the playback signal starts to be reduced in level by several dBs, adjust the record head. It should be possible to arrive at 15 kHz, always looking for the midpoint between two of equal attenuation. 2. The Rectangular 1000 method Whilst the classic method necessitates bulky instruments, the Rectangular 1000 method, equally precise, can be carried out with the aid of the reference generator and good quality headphones. The ear serves in this case as detector.