One day we'll wonder why we put up with reel-to-reel analogue tape for so long. Here is a selection of classic analogue problems which the ADAT doesn't have.

REEL SCRAPE

Newcomers to the recording studio find this noise intensely annoying, although many engineers seem to be able to filter it out somewhere along the connection between the ear and the brain. No matter how well aligned your machine, no matter how flat the flanges of the spool, no matter how carefully you thread the tape, that irritating rasping sound comes back again and again. If you try to correct the problem by bending the flanges of the spool you'll only make it worse, and you may be tempted to adopt the solution many continental Europeans do, which is to use spools without an upper flange. Be warned, however, that you have to use European tape if you take a screwdriver to your empty take-up spool, because the American stuff will throw its coils skywards when you fast wind.

NOISE

Think of all the trouble you take to lay a nice clean signal down onto tape, and then consider what happens when it is deposited on the rusty metal particles of the tape's magnetic coating. The result is that your lovely music plays back mixed with a random signal, which is commonly known as 'tape hiss'. Tape is a noisy medium, an order of magnitude noisier than high quality audio electronics, and tape, rather than any other component of the recording chain, has set the limit on our achievements ever since it was invented. Things have improved vastly over the years, particularly with the various forms of Dolby noise reduction, but when you hear background noise on your finished recording, you can be sure that most of it is coming from the analogue multitrack recorder. Digital recording doesn't eliminate noise; a digital multitrack recorder should really have a resolution of at least 18 bits to bring the finished mix up to full CD standard, since noise build-up is inherent in any multitrack recording process, but the 16-bit ADAT provides a very noticeable improvement on what most of us are used to.

MODULATION NOISE

Tape hiss is one kind of noise, and is due to the fact that the signal is recorded on particles of finite size. Modulation noise is something else, and is arguably more annoying — noise that rises and falls in level with the level of the signal. Modulation noise makes a recording sound 'dirty', and is particularly noticeable if you are recording without the benefit of Dolby noise reduction. If you use dbx noise reduction, this has the effect (through a different mechanism) of making the modulation noise worse. Digital recording isn't totally free of modulation noise, but at 16-bit resolution you would be quite hard pressed to hear any.

PRINT-THROUGH

You have just made a technically perfect recording of a piece of music with true beauty and meaning. You take the tape off the machine and put it carefully on the shelf ready for mixing tomorrow. Tomorrow comes and you play the tape — it isn't at all like you remember it. All the loud percussive sounds have strange echoes, not only after the initial sound but also before it. This is print-through. Particularly during the first 24 hours after a recording is made, the magnetic force of each loud sound will seep through perhaps two or three layers of tape in both directions, causing pre- and post-echoes. Other than fast winding the tape a couple of times in each direction (which doesn't make very much improvement), there is no safe cure for print-through. It has messed up your recording forever. Two methods of reducing the effects of print-through are to use noise reduction and to store the tape tail out. Although there will physically be print-through on a digital tape, it won't affect playback in the slightest. Score so far: Digital 4, Analogue 0.

HEAD ALIGNMENT

All analogue recorders are aligned so that the gaps of the record and playback heads are precisely at 90 degrees to the direction of tape travel. If this statement were true, then head alignment wouldn't be a problem. In practice recordings are made on machines with wrongly aligned heads, which then will not play correctly on a properly aligned recorder. This causes a loss of high frequencies and has a damaging effect on the stereo image.

Although the alignment of digital heads is important, a small deviation wouldn't make any difference at all to the sound, and since the head alignment isn't user adjustable, no one can get it wrong — apart from the manufacturers, who really ought to know what they are doing!

GAP SCATTER

Even if the heads of a multitrack recorder are aligned as well as it is possible to align them, there is still the manufacturing defect known as 'gap scatter'. This is when the gaps of the individual record or playback elements of the head are not all at exactly the same angle. This makes it impossible to align the head properly for all the tracks. In rotary head recorders of all types, including ADAT, there can be no gap scatter.

LINE-UP

Analogue recorders are very moody. Their performance changes with the state of wear, type of tape, the weather, and any one of innumerable variables. An essential task involved in operating an analogue recorder is line-up. At the small studio level, line-up will be performed rarely but (hopefully) regularly. Major studios consider line-up as important as tuning a piano. You wouldn't be happy if you were paying several hundred pounds for a day in the studio and found that the piano was out of tune. You also wouldn't be happy if the recorders were not working at the peak of their performance. Digital recorders do need lining up, but this is a job for specialist engineers and, as long as they are running within specifications, the tape will record and play back perfectly.

WOW & FLUTTER

There is no such thing as a perfectly flat surface, a perfectly round shaft, or a perfectly spherical ball bearing. It's also impossible to align two cylinders so that they are perfectly concentric. The result of this is that the tape machine that runs at an absolutely steady speed will never exist. In the analogue world, this matters because tape speed has a direct bearing on pitch. Slow variations in speed cause wow, faster ones cause flutter. A digital machine will not run at a perfectly constant speed either, but since the data can be read into a buffer memory and then be clocked out with the accuracy of a crystal oscillator, it doesn't matter at all if the mechanics are slightly wobbly (within specification, of course).

TIMECODE PROBLEMS

Timecode is to recording engineers what spots (zits) are to teenagers, and sometimes we forget that, unlike spots, timecode does indeed have a few benefits. But when your synchroniser won't synchronise, or your sequencer drops out of record at the wrong moment, you will be snarling and cursing and wondering why you didn't take up a career as an interior decorator. If you have ever looked at raw timecode on an oscilloscope, straight from the tape machine, then you won't be at all surprised that you get the occasional troublesome moment. In fact, if you look at a steady 1 kHz sine wave replayed from tape, you will be amazed that recordings of music sound anything like the real thing! Even on a good tape recorder the sine wave will be bobbing up and down like a dinghy crossing the wake of the Isle Of Wight ferry — and make that the QE2 when the head gets worn. But if your timecode is recorded onto digital tape, as described in the main text, then it will be absolutely rock solid and almost as dean as when it was fresh out of the generator. The low-cost timecode readers usually found associated with MIDI sequencers will particularly appreciate this.

EDGE TRACKS

What did your mother tell you when you were small? Never record important parts on the edge tracks! Of course we all know this, but the problem is that we don't always know how a piece of music is going to develop. Sooner or later, when all the tracks are full, the producer is bound to say, "Let's ditch the castanets (which are on track 1) and put down this important vocal counterpoint I've just thought of." Or words to that effect. If you use an analogue multitrack I'll ask you the question, "How good are your edge tracks?" Try recording stereo music on tracks 1 and 2 and play back on headphones. If your heads are at all worn, you will hear image shift and dropouts on track 1. Since ADAT is a rotary head recorder, it can't be said to have an edge track in the normal sense, so you can be sure that each track will have identical performance within infinitesimally small margins.

DROP-OUT GAP

So your guitarist needs 100 takes to get his solo correct. It shouldn't be a problem as long as you can drop him in at the right points to fix dodgy notes. Dropping in is easy enough, and analogue recorders have carefully-timed erase current ramping so that the changeover will be almost perfect. Dropping out is another matter, however. It is possible to drop in almost at any point, as long as your timing is good, but dropping out always produces a glitch in the recording, so you have to drop out on a gap in the music. If there isn't a gap, then you're in trouble. On the ADAT, dropping in and dropping out are both gap-free. If you try it on a 1 kHz sine wave (which is the ultimate test), you'll hear a short crossfade on entry and exit. On music, you are very unlikely to hear it at all.

CROSSTALK

In any analogue equipment there will always be crosstalk, which is defined as a signal leaking from a path where it is wanted to a path where it is unwanted. This is a terrible nuisance, especially in multitrack recording when, let's say, a loud snare drum beat leaks onto the vocal track. The crosstalk will be an inconvenience until you decide that the song doesn't need that snare drum after ail — even when you erase it, it will still be clearly audible on the vocal track, and it becomes a major problem. Digital multitracks cannot be totally free of crosstalk either, because there are still analogue signals within the equipment, but there is no crossover from one track to another, which is a major benefit (except that you will soon become dissatisfied with the crosstalk performance of your mixer!).

RECORD CROSSTALK

This is a different manifestation of the crosstalk phenomenon. During overdubbing on a multitrack recorder, you are asking one element of the head to play back while another is recording. The recording element will be carrying a large current, while the playback element will be producing only a very tiny one. You will notice the effects of record crosstalk when you record on the track adjacent to the hi-hat. As you record and monitor the signals from the multitrack, you will notice that the hi-hat has suddenly become much louder and much brighter. If you solo the track you are recording on, you will be able to hear it quite clearly, as the result of record crosstalk. It won't be there when you play back (apart from the normal amount of track-to-track crosstalk you would expect), but it's a pain when you're recording. You may also notice that other tracks seem to jump up and down in level, corresponding to the rhythm of the track, while you are recording. This is a result of the record crosstalk sending false signals to the noise reduction system. Once again, it won't happen on playback — but on a digital multitrack it doesn't happen at all.

HEAD FEEDBACK

Related to and caused by the above, this happens when you try to bounce a signal onto the adjacent track, as you might if you were compiling vocal harmonies onto a single track. Unless you are careful with the level, you will obtain a high pitched whistle (potentially speaker blowing) which is caused by feedback within the head. Once again, it doesn't happen with a digital multitrack.

NOISE REDUCTION PROBLEMS

Noise reduction systems are great, but they are not perfect, and even if they can fool the ears of most of the people for most of the time, experienced engineers will still be able to hear their artifacts. Sixteen-bit linear digital recording uses 'brute force' techniques to give a dean, quiet recording without any psychoacoustic trickery.

HIGH FREQUENCY SQUASH

Analogue tape isn't very keen on recording high frequencies. As you try to put more level onto tape, high frequencies become distorted before mid and low frequencies, which most people would say is not a good thing — although it does contribute to the analogue 'sound', which is liked by some. Digital recording treats all frequencies equally, so your lows, mids and highs will ail be equally clean.

THREADING & RUN-OFF

Yes, there are reel-to-reel digital recorders but, eventually, all recorders will use cassettes of some type. We thread tape now because we have to. In a few years time, we'll think of it as something from the dark ages as we feed another cassette into the slot. And how much session time is wasted in re-threading tape that has accidentally spun off the reel? There are ways and means of avoiding this, but they also take up time. It'll never happen on an ADAT.

COUNTER SLIPPAGE

"Let's go back to the beginning of the first verse", says the producer; you press the appropriate locate button, which you had thoughtfully pre-programmed. Unfortunately, this was some time ago, and with all the shuttling of tape that has gone on, the counter has shifted and the tape comes in halfway through the first line. This may seem a minor point, but many minor errors in location add up to a lot of wasted time. If the tape has a time reference recorded on it, then location can be bar-accurate every time. You can use timecode on an analogue tape for this purpose, with the correct equipment, but a digital format can have this built in. ADAT has.

CLEANING?

Yes, cleaning analogue recorders is a chore but it has to be done, otherwise your recordings suffer. Cassette-based digital recorders are a fairly new invention, so there isn't a lot of data available, but it seems that we are going to have to consider what cleaning routines are necessary. I used my DAT machine for around three years, with regular applications of a cleaning cassette, before it developed a transport problem. I opened it up and oxide was caked around the pinch roller and on several of the guides, so obviously the cleaning cassette is only a partial solution. There is no advice in the ADAT manual other than to clean the external surfaces with a damp cloth, but I suspect that this is one area where digital multitracks will not have an advantage over analogue — we'll still have to keep them clean.

PHASE DISTORTION

If you throw a bundle of different frequencies (which could be music) at a piece of equipment, you hope that it will throw them back at you with the lows, mids, and highs all in the same proportion. If it does, then we say it has a good frequency response. One might also hope that they would all come back with the same relative timings, but in practice the lows or highs may be advanced or delayed with respect to the mid frequencies. If this happens more than the minimum amount that corresponds with the frequency response of the equipment (frequency and phase response are closely linked) then there is phase distortion. Analogue tape recorders are the worst offenders in this respect. Try recording a square wave and view it on an oscilloscope — it will come back anything but square. It used to be thought that this doesn't matter, since it still sounds pretty much, in fact almost exactly, like a square wave. But almost isn't good enough, and the phase response of digital recorders is much, much better.

DROPOUTS

Analogue tape and digital tape can both suffer from dropouts due to faults in the tape. The difference is that with analogue the effects can be unnoticeable, irritating, annoying or disastrous, depending on the dropout's length and depth. A digital system can be designed so that its error correction system can cover up faultlessly the worst dropouts that can be expected in everyday use. However, really bad dropouts will cause the system to mute — even if the muting can be switched off, there may be a loud digital 'splat'. Some would argue that the advantage here is in analogue's favour, since you will get at least something from the tape and dropouts never cause additional noises.

ANALOGUE ADVANTAGES

It wouldn't be fair not to mention a couple of things that analogue can do that digital can't. Spot erasing a click in a digital recording can be impossible, difficult, or time-consuming, depending on the system. On an analogue recorder with a specific spot erase facility, it's straightforward enough. Even if the recorder lacks such a facility, it may be possible to do it by rethreading the tape so that it doesn't pass between the capstan and pinch roller and then moving the tape by hand. Another analogue trick which used to be very popular before we had such an array of effects units is to turn the tape over and record backwards — even now, it's the only way to create true reverse reverb.

It may seem like a giant leap backwards to have to sacrifice an audio track for timecode when ADAT offers inherent synchronising capability, but until the extended (BRC) remote control appears, which I suspect will remedy this, we'll still have to do just this. Fortunately, it's nowhere near as problematic as using timecode on an analogue multitrack. So, on which track should the timecode be recorded on a 16-track system with two ADATs?

Suppose your master ADAT is connected to tape outputs and inputs 1-8 and the slave is connected to 9-16. If you record timecode on track 16, then the slave will take three seconds or more to start after the master enters play or record and only then can your sequencer attempt to lock up, so timecode really has to be recorded on the master. It doesn't matter which track you use, but you are now going to have to work around the 'hole' on your mixing console, which is irritating. The alternative would be to connect the slave to 1-8 and the master to 9-16, but I can't help feeling that this is thoroughly illogical. As the price of progress, I'm sure we'll get used to it.

Alesis plan to make a bigger version of the remote control which will supply the facilities the ADAT lacks. If you wanted to criticise ADAT as it stands, then you would be saying things such as:

- You lose an audio track for timecode.

- There's no provision for digital synchronisation to other equipment (which one day soon will be very important, mark my words).

- There are no record-ready switches on the remote control.

- Only three locate points are offered.

- No editing is possible...

- Its buttons are not very positive.

Alesis are not saying very much about their plans at the moment, but I suspect that, at the very least, the first five points will be answered by the BRC.

It is usually thought that any piece of equipment, in any field, that has the word "professional" printed on it is definitely for amateur use only. But the Alesis ADAT will undoubtedly find its way into many professional applications, simply because the sound quality is good and it works with the minimum of fuss. As well as jack inputs, Alesis have very sensibly provided a multi-way input and output connector that operates at balanced professional levels. If the Fostex E16, G16 and G24, and also the Tascam MSR16 and MSR24 can be used to make money out of recording, then there is absolutely no doubt that the Alesis ADAT can too.

Tape Wear

Multitrack recording usually involves going over the same sections of a piece of music again and again; by the end of the session, the tape may have passed the heads hundreds of times, so there is the distinct possibility of error rates creeping up. The ADAT won't cycle automatically for more than half an hour, so my intended overnight cycling test didn't work. I thought the next best thing would be to take a five year-old standard VHS tape, which has been re-used many times in my video, and record audio on it. It certainly isn't recommended to use anything less than the best tape, but I have to say it sounded absolutely perfect!

RMB REMOTE METER BRIDGE
Designed to mount above the BRC remote control. Provides 32 channels of LED meters.

AI-1 INTERFACE
The ADAT has only a proprietary 8-channel optical digital input and output. The AI-1 will be able to translate this to AES/EBU (professional) and S/PDIF (domestic) digital formats. Apparently, this will also include a sample rate convertor.

AI-2 ESBUS INTERFACE
One of the unfulfilled dreams of professional audio is that all equipment should have a standard synchroniser interface. ESbus is such a standard, defined by the EBU and SMPTE, but it hasn't made as much progress with manufacturers as it should. Full marks to Alesis for supporting it.