Should You Upgrade from a DX7 to a DX7II?
If you already own a Mk1 DX7, you're now faced with an agonising dilemma: do you risk everything and chop it on for the MkII model? Howard Massey weighs up the pros and cons of taking the plunge.
Yamaha's new DX synthesisers have a number of features the old ones lacked, but if you already own an "old" DX7, are the new options enough to justify upgrading? Compatibility may be the key.
FIRST OF ALL, let's discuss some of the improvements that everyone has been talking about. Probably the most important one is the fact that the DX7II, unlike its predecessor, is multi-timbral. It can make two sounds at once - sort of like the DX5 (or a DX7 with a TX7), but better, because the two sounds can now interact in many more different ways. Like the DX5 (or most bi-timbral synthesisers, for that matter), the new DX7 lets you split its keyboard or layer two sounds over its entire keyboard. Unlike the DX5, however, the new machines allow you to edit either sound while hearing both in this layered keyboard mode (called Dual mode), meaning that you can really approach the DX7II as a 12-operator digital FM synth. This is great if you think you've explored all the possibilities of creating sounds with six operators (although quite honestly, I haven't run across anyone yet - including myself - who has).
The interactivity between whichever two sounds you call up at any given time goes well beyond this, however. The DX5 (and the DX7/TX7 combination alluded to earlier) allow you to send audio outputs to separate channels, making for a pleasing stereo effect. The DX7II takes things a step beyond, by offering programmable panning of those outputs across the stereo image. This is an area of the instrument that I am personally very excited about, because it takes what is normally a directional, focused sound and allows you to turn it into an ambient sound - straight from the DX7II's front panel. True, if you normally play your instrument through a single amplifier/speaker combination, this will have little impact. But if you work in the studio with an engineer who is sharp enough to record you in stereo, it could mean a heck of a lot. You can pan by key velocity, by note number, or with an LFO - and there's even an envelope generator which is totally dedicated to this function. Furthermore, you can shift the mixed signal of your two patches, or you can alter the output level of one and not the other, or you can alter the output level of both - in a complementary fashion so that as one increases, the other decreases proportionally.
The second most striking improvement of the DX7II lies in the quality of its output. You know all of that digital fizz that accompanies so many DX7 patches? Well, it won't be gone completely in the DX7II, but it will be much, much quieter and virtually unnoticeable in 90% of your patches.
The clarity of the sound itself has also been improved, thanks to 16-bit digital-to-analogue conversion and faster processing. Not only are sounds crisper and better defined, I find the bass end to be improved a great deal as well. This, along with the many different detuning options offered by the DX7II, allows you to create those big, beefy sounds (hold the onions, please) that DX7 owners have been crying out for over the last four years. Maybe I exaggerate - but there is no question that the DX7 carried the often unjustified label of being "thin-sounding". Have no fear - that's something the DX7II cannot be accused of.
Another big improvement is that the "function controls" of the old DX7 are, happily, no more. Instead, all of these parameters (which include things like pitch-bend wheel assignment, portamento, and real-time controller routings) are voice edit parameters on the DX7II. This means that you program them individually for each voice, and that they change per patch just as surely as operator output levels do. At long last, you can customise your controller routings for the particular sound you've created. What this really means is that there is virtually no need to alter anything about a sound in live performance - you really can set everything up in advance.
Speaking of controllers, it's worth noting that the DX7II allows the use of more controllers than the old DX7 did. You can have two sustain pedals, for example, as well as two foot controllers. One of the sustain pedals can take on some pretty nifty functions, like acting as a soft pedal, or duplicating the effect of a piano sostenuto pedal.
There are also a couple of continuous sliders on the front panel of the DX7II (an idea borrowed from Yamaha's KX-series MIDI keyboard controllers). These allow you to alter voice edit parameters (things like EG settings, operator output level or frequency, or even the algorithm) directly from Play mode. Furthermore, the movement of these sliders is transmitted via MIDI, so that they can be used to alter various continuous controller parameters on slave synthesisers. You can also link these via MIDI to controllers from other master keyboards (assuming you're using the DX7II as a slave instrument), so that, for example, moving the modulation wheel of an Emax changes the LFO speed on your DX7II - or almost anything else about the DX7II patch or patches called up.
Ah, MIDI. Here's an area where DX7 owners have been deprived for years. In Yamaha's defence, it should be pointed out that the original DX7 was designed in 1983 - prehistoric days, by MIDI standards. Come the new DX7, however, and we have a machine with such a comprehensive MIDI implementation (and a better keyboard to boot), that it's quite suitable for use as a master keyboard controller - something the original DX7 could not claim.
Got some MIDI questions? We've got some answers. Yes, you can transmit on any channel (though unfortunately, you can't transmit on more than one channel). Yes, the two patches can receive on the same or different channels. Yes, you can receive in Omni mode. Yes, Local On-Off is supported (meaning that you can disconnect the DX7II keyboard from its own sound-generating circuitry). Yes, all 127 velocity increments are transmitted. Yes, volume can be externally controlled. Yes, program changes can be prevented from being transmitted (the DX7 always sent them) and you can even re-map these program changes (so that, for example, calling up patch 22 on your DX7II causes it to send out a "change to patch 47" command to slave synths). And yes, you can even transmit the contents of the DX7II edit buffer via MIDI - meaning that you don't have to store sounds in memory before you can ship them over to a patch editor. In short, you want MIDI - you got MIDI.
Speaking of memory storage, there's a lot more of it on the DX7II than there was on the old machine. Twice as many voices, to be exact, can be stored in the internal memory, along with 32 performance setups, a couple of micro tunings (more about these later), and the MIDI setup.
The same amount of data can be stored on the DX7II's RAM4 cartridges, which are larger - both physically and in memory capacity - than the original DX7 RAMs. These can also be used to store micro tunings alone, or data for a fascinating new function called Fractional Scaling. Anyone who used the DX7's keyboard level scaling function will be interested in this. In a nutshell, keyboard level scaling allows you to increase or decrease an operator's output level (thus causing a change to the volume or timbre of the sound) as you play different areas of the keyboard.
The original DX7 used a system of preset curves to initiate these changes, and this system is still implemented on the DX7II. But the new machine takes things a step further with the introduction of this "fractional scaling" parameter. The fact of the matter is that the so-called "curves" used by DXs for level scaling are not curves at all. Instead, the output levels of various operators are altered in groups of three semitones at a time according to one of four preset tables. Fractional scaling allows you to dig right in and change this table, or even to create your own tables. This gives you a much, much finer degree of control over changing volume or timbre on areas of the keyboard, and also allows you to create true hard splits within a single voice - something that becomes, admittedly, less vital when you are working with a multi-timbral instrument that already has a split keyboard mode.
If you go for the "FD" model of the DX7II, you'll also have floppy disk storage at your disposal. Beyond holding DX7II data, this drive can also be used for generic MIDI data storage. The maximum file length of 20 kilobytes makes its use somewhat limited, but I am told that Yamaha plan to increase this maximum file capacity in a ROM update.
The much-vaunted micro tuning capabilities of the DX7II - none of which were implemented on the stock DX7 - allow you to do all kinds of weird and wonderful things to the intervals between semitones. You can create microtones, macrotones, inverted keyboards, or "dead" areas of the keyboard where the pitch remains the same no matter which note you play (great for percussive patches). You can also work with any of the 11 tuning presets (which include those created by such luminaries as Kirnberger and the unforgettable Vallotti and Young - weren't they a song-and-dance team in the '50s?), though we have become so accustomed to equal temperament in modern music, it remains to be seen if these alternative tunings will be used a whole lot. Still, hats must go off to Yamaha for at least offering the user the option of getting away from the norm.
There are many, many other improvements that are less significant, but may nonetheless be important to you. For one thing, those horrible membrane switches of the DX7 have been replaced by real switches - some of which are now equipped with LED status lights. The LCD is much larger, allowing for two lines of 40 characters (as opposed to the two lines of 15 characters in the DX7), so that you can see much more data at any given time. For example, you can view all the EG settings for a particular operator in one single display, without having to cycle through all the rates, then all the levels. Cursor switches, similar to those used on the DX5, allow you to get around from parameter to parameter within the display. The LCD is also back-lit, making it much easier to use on stage. The master tuning display now shows you numbers - so you can actually see if you're at A440 or a bit sharp or flat, though as with the DX7, there is still no reference A440 tone.
Physically, the DX7II is just a little bit smaller (though not so much smaller that it won't fit in your existing flightcase) and a little bit lighter than the original DX7, but appears to be just as roadworthy.
BUT NO MATTER how numerous and dramatic the improvements made by the new-generation DX7 over the old model may be, there is one thing that all owners of old models will want to be assured of before they part with the cash for a new machine. Namely, that all the time (and quite possibly money) they have invested in creating a library of sounds for their old DX7 will not be rendered useless as soon as the replacement arrives on the doormat.
Well, let me assure you of this. All DX7 sounds can be played by the DX7II - and will, in most instances, sound exactly the same (or slightly better, due to the higher bit resolution and processing times). This is accomplished in one of two ways. The first method involves simply shipping voice data from a DX7 to the DX7II via MIDI (the DX7II will respond to the DX7 SysEx ID code). You then, of course, have the option of storing that received data in internal, cartridge, or (on the FD model) disk memory. The second method involves the use of a special adaptor (the ADP1, available from your nearest friendly - or unfriendly, as the case may be - Yamaha dealer). This adaptor sits in the cartridge slot of the DX7II (which, you'll remember, is larger than that of the DX7), ready to receive your old-style DX7 RAM. Insert the old cartridge, and your new DX7II has direct access to all the patches stored on it. Beware, though, that this is not a two-way street. The DX7II is compatible with the DX7, but, as the Americans would put it, it's only upwardly compatible. That means you cannot send DX7II patches to a DX7 - it simply won't understand them, because the System Exclusive code is much more complex. More importantly, that DX7 RAM cartridge that's sitting in the adaptor in the DX7II cartridge slot has now become a ROM cartridge - because you cannot write DX7II data to it.
Bear in mind, also, that in many instances you may need to tweak a couple of parameters in order to get a DX7 voice to sound exactly the same once it is in the DX7II. There are four areas of slight incompatibility, so let's take them one at a time.
Function controls. As we've seen, these no longer exist on the DX7II. Instead, they have grown somewhat in complexity and evolved into standard voice edit parameters. When you transfer a DX7 voice over to the DX7II (either via MIDI or via the ADPI cartridge adaptor), no function control parameters are transmitted, simply because they are stored in a separate, non-MIDI area of the DX7 memory. Therefore, these particular parameters all reset to initialisation default values in the DX7II, which simply means that everything is off except the pitch-bend wheel, which defaults to a range of two semitones. So, if you created a DX7 sound with a specific controller in mind to, let's say, route an LFO signal, you'll need to put the DX7II into edit mode and set up that routing again from scratch. The big difference is that this time, when you store the sound, the controller routing is stored with it as well. The rule of thumb here is that all DX7 function controls that you may have used will need to be re-entered into the DX7II - a tedious if not particularly difficult process. Besides the old function controls, the DX7II also contains many new edit parameters, such as the new Pitch EG parameters discussed below. All those parameters which don't exist in the DX7 simply assume initialisation defaults when a DX7 voice is sent to the DX7II.
Amplitude modulation sensitivity. On the DX7, this parameter has a range of 0-3, with 3 representing maximum sensitivity to both LFO and EG bias modulation. On the DX7II, the range has been increased to 0-7. So, when transferring DX7 patches that use either EG bias or LFO amplitude modulation, you'll have to re-tweak this parameter. Operators which had a sensitivity of 3, for example, will still have a sensitivity of 3 - but "3" will no longer mean "maximum".
Pitch EG. The DX7II offers several additions in the pitch EG department, including rate scaling, a keyboard velocity control, and a new variable Range parameter. On the DX7, the (non-adjustable) range of the pitch EG is always eight octaves (four sharp and four flat). When you transfer DX7 voices over to the DX7II, the Range parameter automatically sets itself to eight octaves, so it seems like everything should work perfectly, right? Wrong. The problem is that the absolute speeds of movement within the pitch EG change (even though the rate values remain the same) when the Range is altered - just as operator EG speeds change as output level is altered (as the envelope "squeezes down", it takes less time to get from level to level). Unfortunately, a DX7 pitch EG rate of, say, 50, is the same speed as a DX7II pitch EG rate of 50 only when the Range is one octave, not eight. Therefore, you'll find all of your DX7 pitch EG rates are considerably higher when the sound is exported to the DX7II. This will have little effect on pitch EG rates which are quite high, but can be drastic for those which are relatively low. For example, a favourite trick of mine when synthesising "acoustic-type" sounds on the DX7 is to put in a very slight pitch EG movement at an extremely low rate (say, travelling from an L2 of 50 to an L3 of 45, at an R3 of 0). This adds a barely perceptible instability to the sound - and slight instability is a hallmark of virtually all acoustic sounds. The first time I shipped some of these patches to the DX7II, I was shocked to hear what sounded like the instrument becoming violently ill. The R3 of 0, which was really too slow to hear clearly on the DX7, was now a very obvious slow drop in pitch. The solution was simply to change the Range to one octave, and to alter the levels accordingly. The bottom line, then, is that you may well have to make similar kinds of alterations to patches which use the pitch EG.
Delay effects incurred by operator envelope generators. One of the more cunning tricks employed by DX7 programmers is to build in "delays" by taking advantage of a deliberate anomaly in the spec of DX7 operator EGs. As is shown in my first book, The Complete DX7 (plug, plug), when travelling from a level of below 20 to another level of below 20 (in other words, when travelling at inaudible EG levels), the rates of movement act as digital delay lines. This works on the DX7 even if you are travelling between two inaudible levels that have the same values. On the DX7II, however, the EG logic has been changed somewhat (possibly because editing changes made to DX7II EG levels are assimilated by the microprocessor in real time). Here (as is shown in The Complete DX7II, plug, plug, plug), no delays whatsoever are incurred when travelling between two levels that are exactly the same. Try playing a DX7 sound (like the 'Watergarden' ROM preset) that uses this trick on the DX7II, and you won't hear the delays. To fix this, simply alter one of the two levels so that it's slightly different - even a single increment will work - from the other.
SO - TO DX7IID or not to DX7IID? That is the question. With luck, I've helped to provide you with an answer. The deciding factor will simply be how many of these new features are important to you in your own work. It's a tough decision - but one the designers have made a little easier by deciding to make the new machine "upwardly compatible" with the old one. My opinion? I love the new instrument - and I still love the old one.
Gear in this article:
Feature by Howard Massey
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!