Stepp DG1 Digital Guitar
Few instruments have caused as great a stir in the marketplace as the futuristic-looking DG1. Stepp describe it as "a radically new instrument that plays like a guitar, and which can produce some of the most phenomenal sounds you're ever likely to hear." Ian Gilby visited Stepp to find out for himself and talk with its creator, Stephen Randall.
The Stepp DG1 is the end result of four years' intensive research and development, and the fulfillment of one man's dream. That man is Stephen Randall, a successful young songwriter and guitar player who, like many of us, turned to synthesizers to obtain new sounds for his music but felt frustrated from having to use keyboards. Eventually, that frustration led him to put together a UK design team whose objective was to produce - "a guitar that played like a guitar, responded like a guitar, and gave you a guitarist's expression and musical freedom. But one which also offered you the kind of scope and creative opportunities which have never been possible before."
Enter the Stepp DG1, the world's first digital guitar. And one of the most revolutionary and sophisticated instruments ever made.
In order to appreciate what the DG1 can do, it is necessary to understand the basics of how it functions.
Any guitar synthesizer system that is pitch-to-voltage based - the DG1 is one of few that is not - has to rely on the physical properties of a vibrating guitar string. The computer, whether it analyses the pitch by the way the string displaces light or magnetism, must 'see' a part of the vibration cycle, at least, before it can calculate what note has been played and pass on that information to the sound-generating circuitry. That all takes time - and that's before MIDI transmission has been considered or the delays inherent in the target synth that the guitar is controlling. Already we have introduced three delays into the system, which combine to produce a recognisable time-lag between a string being plucked and the sound being heard. And that is also ignoring everything to do with the expression of the guitar.
So the first criterion of the Stepp DG1 design brief was that there must be no delay in detecting pitch. Or, more accurately, no delay that the guitarist can perceive. There is obviously a delay whilst the three onboard computers process the note, but it is miniscule and indiscernible to the ear. I can vouch for that having played it. In fact, the DG1 is faster from plucking a string to producing a sound than a traditional electric guitar with a pickup.
The reason that there is no perceptible delay is that plucking a string on the DG1 is just like turning on a switch. The fretboard is 'active' and the string forms part of an electrical contact.
Unclipping the central cover plate on the DG1 reveals that the strings are not in fact joined in the middle. This is an area that will potentially give guitarists problems if they are not prepared for it, because where they normally have a lot of tactile reference between their left and right hands, on the DG1 (as on the SynthAxe) that has been cut. It is very similar to what synthesizer keyboards first felt like to pianists; they hated them because they didn't feel like a piano. Stepp's founder, Stephen Randall, readily admits that they are bound to go through a similar period of guitarists saying that the DG1 doesn't feel exactly like a guitar and accepts that as fair criticism.
The reason for the strings being physically split is that the DG1 doesn't employ traditional pitch-to-voltage conversion and thus has no need to analyse vibrating strings. Instead, its guitar-like neck has what Stepp call 'semi-conductive intelligent' (SCI) frets - twenty of them. These are made of a hard, semi-conductive material and not metal because unless each metal fret was split (like on the old Vox organ guitar and present day SynthAxe), they would conduct and short out.
Stepp considered this 'split fret' approach but felt that it was not ideal - tactile feedback is very important to the guitarist and the fingers might feel a resistance from grooves in the frets. It's also possible that as strings moved over the grooves they would leave tiny deposits of metal which could build up and make the fret detection circuitry unreliable. In other words, a problem area.
Mind you, Stepp admit to having had their fair share of problems in producing their own SCI frets - it is a very difficult design to get right - but the end product feels just like a metal fret, but won't wear down, and is certainly ingenious.
What happens is that the 'intelligent' fret knows instantly that it is being touched by a metal string and because of the electrical resistance between notes played on the same string, it can tell the difference between them. Since it is dealing with resistance, it also knows when a string is being bent because the resistance increases.
What this means is that the DG1 can accurately measure the change in resistance and, by means of software, have it alter the pitch, say by 10%, or the filter sweep by 12%. So that one parameter that the guitarist uses a lot - string bend - can suddenly control anything the DG1 software will allow. And that is just the starting point...
Since MIDI is not used to connect the guitar to the synthesizer voice circuitry, as it is on most 'guitar controllers', the guitar exhibits no apparent delay problems. Thus, it feels highly responsive when playing fast, fluent passages and hammer-ons, pull-offs to open strings, double bends etc are all reproduced faithfully.
A guitarist, however, effectively always plays in legato mode and strikes adjacent strings in quick succession when strumming a chord. A keyboard player does neither. So, to emulate a keyboard style of playing, the neck trigger switch located below the 'stepping man' logo on the DG1 selects Keyboard Mode, which permits the guitarist to trigger a note instantly just by pressing a string onto a fret. In doing this, you actually destroy a lot of the guitaristic elements of the DG1 sound, but when playing synthesized or sampled pianos from the guitar this mode helps produce a more authentic feel. It also frees the right hand to do such clever tricks as hammering on descending bass notes on the neck whilst the left hand is playing ascending chords. Beat that Van Halen!
Keyboard Mode takes some getting used to because guitarists naturally position their fingers on the fretboard fractionally before they actually hit the strings, which causes notes to trigger instantly. This is not the case in normal Guitar Mode.
The DG1 does not rely on a vibrating string to provide the information about when a note starts and stops either. On any guitar synthesizer, as on any guitar, it should be left up to the properties of the guitar string to dictate how long the sound lasts. With the Stepp DG1, the properties of the string are created in software. Therefore, if you want the string to sustain forever, you don't use an E-bow, you simply turn up the sustain value to maximum. And what you do to stop it sustaining is exactly what you would do on a guitar - you damp the string. You also obtain longer or shorter sustain by plucking the strings harder or softer - as on a guitar.
The pluck or 'Strum' parameter as it is called on the DG1 is software assignable, meaning that you can use the pick intensity, like keyboard velocity, to modulate any permissible synthesizer parameter - filter cut-off, oscillator mix, LFO rate, etc. The same goes for the 'wammy bar' or vibrato arm. It doesn't have to control pitch, it can just as easily function as a swell pedal controlling volume.
The crucial difference between the DG1 and a MIDI keyboard is that once the DG1 note envelope has finished, or the string has been damped, the note is then 'off'. This is not the case on a keyboard - regardless of whether the ADSR has completed its full cycle, MIDI does not accept a note as being 'off' until the held key is actually released. That's because MIDI is primarily a keyboard language and does not understand 'sound envelopes'; it comprehends only note-on and note-off information and is uninterested in what happens between the note being turned on and being turned off.
Problems arise on MIDI guitars because the poor guitarist thinks the guitar note is off when he can no longer hear it. But when he moves his fingers the note simply retriggers, since it has never been switched off as far as MIDI is concerned.
Stepp feel the current MIDI Specification is inadequate in this area and are campaigning to have it altered to take account of envelope characteristics so that MIDI note-off, in future, is related either to when a key is released (as at present) or to when the envelope stops. A sensible suggestion.
It is because the Stepp DG1 does not utilise MIDI to communicate its note information, that it successfully manages to traverse the inherent delay/tracking problems MIDI guitar synthesizers exhibit. Instead, it connects directly, via a five metre cable, to its own synthesizer expander module housed in the external LSU (Life Support Unit). This also contains the power supply plus all audio connections, and cleverly doubles as a guitar stand for the DG1. More Stepp ingenuity!
The DG1 doesn't need to measure string vibrations other than during the initial 'pick' period. Stepp recommend the use of the metal plectrum. The reason being that the plectrum tells the software that a pluck action is a 'peak' - and it registers a peak with capacitance - and recognises it as an intentional pick. It doesn't matter if the string continues vibrating after being plucked, the DG1 will not re-trigger, because the vibration without the capacitance is not viewed as a legitimate trigger. That's how the DG1 gets over the perennial re-triggering problem that has plagued previous attempts at guitar synthesis.
If you damp any string the DG1, being intelligent, knows that the string is supposed to be muted, and it adjusts the sound envelope accordingly to give a muted output. This muting occurs in software and can be transmitted via MIDI and used to mute DX7 voices from the DG1, for example.
At present, a plastic plectrum will not trigger the DG1, but the metal one supplied is easy to use. Eventually the DG1 will accommodate any plectrum, so I am told.
The final guitaristic device that the DG1 is able to recreate is a 'strum'.
If you have a vibrating string, what you would expect on a guitar is for the note to go off when the plectrum touches the string and stops it vibrating. When a guitarist strums, he actually leaves gaps between the notes. Now if there was no way of detecting how to turn the note off before you pluck a new one, you would start to lose those important gaps between the note envelopes and the definition of the guitar sound would be quickly lost.
So, on the DG1, when the metal plectrum touches a string just before it is plucked, the contact tells the computer to quickly turn off any existing note played on that string, before the pluck registers a new trigger. The beauty of this system is that you are blissfully unaware of what the DG1 is doing, since the very same arrangement occurs naturally on a real guitar, so it doesn't interfere with your playing. What it does is allow suitable synthetic sounds to be strummed rhythmically, complete with accented downstrokes if that's what you play. Ever tried doing that on a Roland GR-700 or Shadow?
This valuable feature was demonstrated beautifully when the DG1 was MIDI'd to an Akai S900 sampler loaded with an acoustic guitar sample. You could actually strum the sampled guitar from the DG1 - a physical impossibility using a keyboard - and it sounded bloody amazing!
Hopefully, by now, you will appreciate that the Stepp DG1 simulates the playing techniques of a guitar more accurately than any other guitar synthesizer known, and lets you impose natural guitar inflections on non guitar-like sounds to create new and unique synthesized guitar voices.
Tremendous thought went into the design of the synthesizer control section of the DG1 so as not to alienate the guitarist unfamiliar with synthesis.
To give an example: on the DG1, 'bend' can be a variable amount. So Stepp asked themselves how a guitarist would change pitch bend from a default value. The technical answer is that he would route oscillator frequency to be modulated by the string bend - which is fine for a synth player to understand, but the guitarist is probably 'lost' by such terminology. He just wants to turn a control that says 'Bend'. So the DG1 offers a DX7-like membrane touch-button (Stepp call it a 'cell') labelled 'Bend'.
To change the bend amount, or any other parameter, you go out of Play Mode and into Edit Mode by pressing the blue Play/Edit cell below the LED display. Immediately you are presented with an opening menu of flashing cell LEDs - anything that doesn't flash you can't edit at this stage. So, to change pitch bend the guitarist must enter Edit Mode and press the Bend cell, by which time he has already seen all the other flashing LEDs which hopefully act like a teaser and make him wonder what happens when he presses a different cell. Very quickly, and without realising it, the guitarist is gently initiated into the 'black art' of synthesis. Clever psychology on Stepp's part.
Selecting Bend gives you the current value in the bottom LED display - anywhere from 0 (off) to 100 (six semitones). There are 10 preset sound patches or 'programs' on the DG1 which you can edit but not overwrite, and on these pitch bend has a default value of 50 (giving three semitones when the string is bent halfway across the fret).
Stepp found little point in specifying string bend in semitones because guitarists don't always bend in precise semitone steps. The existing system allows you to set up individual maximum bend ranges for each string by relying more on feel, ie. altering the Bend value until you hear what you want. (Stepp use this approach successfully on most parameter specification and have christened it 'Active Performance Software'.)
Since the DG1 measures the bend by how far across the fret the string has been moved, bend no longer has anything to do with string tension (which can thus be set as preferred). Thus, you can bend the strings the same amount each time but programme the degree of pitch change to be different.
In keeping with the DG1's design philosophy, Bend is a 'soft' performance parameter which can be routed to control oscillator mix, filter cut-off or any of the myriad sound parameters possible - not simply pitch. At present, you can only bend three strings simultaneously, but I doubt many guitarists will find this restrictive. I didn't.
What is immediately noticeable when you first play the Stepp guitar is that it uses stainless steel strings all of the same gauge. Part of the reasoning behind this, I would guess, is that with so many possibilities open to the Stepp guitarist it is just as important to be able to bend the low E string as far as you can the top E. I found I got used to them very quickly. And as it transpires, both Allan Holdsworth and Lee Ritenour have recently opted for single gauge strings on their SynthAxe guitar synths. A case of follow the leader, perhaps?
The DG1's blue membrane panel is logically laid out at the top of the contoured grey plastic body. Having only one control - the parameter wheel - necessitates quite a bit of activity to compile the sound you want, though no more than with most modern synths.
A number of the membrane cells do interesting things aimed specifically at the guitarist. 'Tune', for instance, allows the string tuning to be adjusted differently for each string or equally across all six and can be stored along with each sound patch in memory. Each open string can be electronically tuned over six octaves - from C-0 to D-6 (without breaking!!) - allowing you to span a staggering eight octave range!
And because the DG1 is a totally electronic guitar, which does not rely on string tension, it will never go out of tune. Unless you programme it to! Still, the small knob to the far left of the vibrato arm, provides fine tune adjustment of overall pitch to match less fortunate instruments. The adjacent knob is the DG1 's master volume control.
Allied to pitch is a superb facility that creative electronic guitarists will surely find someway of incorporating into their repertoire. By selecting 'Chord' you can fret any chord, strum each string in turn, and have it stored as an 'open tuning' as part of the current program. Who needs a capo? Naturally, you are not restricted to storing only chord shapes you can physically fret - you can equally select 'Chord', then play any note on each of the strings in turn.
Used in conjunction with the 'Glide' (portamento) function, a synthesized blues Strat sound, and some suitably dexterous playing - this can produce an extraordinarily good simulation of blues slide guitar that you couldn't hope to get out of a Shadow or Roland guitar synth in a month of Sundays!
Lastly, if de-tuning a string reduces the amount of 'poke' it has, you can compensate for this by re-adjusting the volume of individual strings using the 'Vol' parameter button in the section of cells marked 'String'. I can see plenty of applications for this feature too.
On the synthesizer front, the DG1's nearest comparison is probably the Oberheim Xpander, given that nothing is connected for modulation purposes until you decide to assign it. It looks fairly basic in synthesizer terms, but in modulation terms is enormously complex, because you are still utilising the guitarist's dynamic expression as the last part of the equation.
A guitar after all has the same basic 'sound', so why does Eric Clapton sound different to Jimi Hendrix? The reason is that they are continually changing the shape of their sound in different ways while performing. Recognising this, Stepp placed considerable emphasis on the performance modulation abilities of their instrument when designing it - and it shows. Listening to Go West guitarist Alan Murphy play the DG1, his characteristic style still shone through - which says a lot about how accurately the DG1 reproduces a guitarist's individualistic phrasing.
Among the synth cells, 'Vol' sets the balance between the dual bank of oscillators (two per string) which can be modulated by any one of six modulation sources: LFO, Envelope Generators 1 & 2, Bar, Fret or Strum. These are selected by pressing their cell buttons, identifiable by their blue stripes and yellow LEDs.
Routing any of the six modulation sources to any of the thirteen possible modulation destinations is logical and easy to accomplish - you select your modulation destination, Osc 1 (frequency) say, and unless already assigned, the LEDs of the six modulation sources will flash until one is chosen. The selected cell's LED subsequently remains lit whilst the other five go out. You can then change the depth of modulation by pressing 'Range' and turning the parameter wheel whilst listening to the effect. The Range LED window displays the current value (-9 to +9).
The DG1's oscillator waveform selection may appear basic - sawtooth or pulse only - but is sophisticated, as you can modulate both the waveform mix and the waveform itself, as well as the pulse width.
For example, let's say we have a sawtooth wave on Oscillator 1 controlled by a slow attacking envelope. The waveform can start off as sawtooth, go into a mix of both sawtooth and pulse, then into a pulse wave. Simultaneously, you can have pulse width being modulated by how hard you hit the strings (Strum), so as the envelope moves into the pulse section the pulse width opens and closes. Thus, from a few basic building blocks, the Stepp DG1 can create extremely rich, complex sounds that would be very difficult to emulate on most keyboards.
On the second oscillator, which can be mixed in with the first, pulse width could be modulated by the Bar, with the mix between the two oscillators being modulated on the frets by the degree of string bend. The aural effect of this I cannot possibly hope to describe adequately; suffice to say that it sounds amazingly full of life!
Oscillator 2 offers the same features as Osc 1 apart from 'Sync' and 'Offset', which allows the frequency to be accurately raised or lowered by up to two octaves. Thus, harmonies can be quickly programmed, or whatever you play can be duplicated one or two octaves below to produce an instant bass accompaniment. Small pitch offsets of a few cents produce ADT or 12-string chorus type effects, which are made all the more realistic by assigning a modulation source to dynamically vary the Offset amount.
The filter section of the DG1 offers control of Resonance, Cut-Off (Frequency) and Track. If you route Cut-Off Frequency to Envelope 1, which defaults to volume, you have control of the nature of the 'plucked' sounds because the filters move in the same shape as the volume envelope. If you disconnect the filter from Envelope 1 and route it to Envelope 2, then you can produce effects where the filters are opening up slowly but the plucked sound begins straight away.
'Track', as you might expect, adjusts the relationship between filter and oscillator frequency so that a higher note does not necessarily have to grow 'brighter' as the pitch increases, it can become muted instead. This may be more desirable with certain sounds.
Two ADS (Attack Decay Sustain) envelopes per string dictate the basic shape of the notes produced on the DG1, along with the unique Strum Mute and Damp envelopes. The left-hand Damp envelope overrides all others and is preset internally to simulate the effect obtained by holding your finger against a string without fretting it. And very realistic it is too. The right-hand Strum Mute envelope enables rhythmic playing and is programmable. It may also be stored as part of any program patch, so you could tailor the muting characteristic to individually suit every type of sound if you wished.
In toto, the Stepp DG1 can store 100 programs comprising 10 preset, 70 user-programmable and 20 'Split' user programs which can be dumped to tape or saved as system exclusive data. Split Mode is multi-timbral, like MIDI Mono Mode, and enables the DG1 user to assign individual programs to any string. Each string could therefore be used to play a different sound.
The creative potential here is enormous, though the effectiveness of Split programs will very much depend on your playing skills and the appropriateness of the chosen sounds. Still, it's probably the closest any guitarist will come to a 'one-man orchestra'.
Throughout the review I have emphasised that the DG1 does not need to use MIDI - it works best without it. But it does offer a MIDI interface on its expander unit and can be used to control other MIDI instruments, though many of the expressive nuances of guitar playing that you can achieve with the DG1's own voicing circuitry are, unfortunately, lost when MIDI is selected. This is no fault of the DG1 it must be emphasised; it is a fault of MIDI, since the present MIDI Specification has no means of interpreting intricate guitar-based performance information.
Nevertheless, some of the triggering problems associated with MIDI guitars can be overcome (or at least disguised) on the DG1 when MIDI is used, provided you remember to damp the strings before changing chords or notes (to signal a note-off command in MIDI). This task isn't so difficult to perform either - in fact, many electric guitarists damp strings between notes when playing at high volume levels anyway. It just takes practice.
The instrument functions in Poly or Split Mode. In Poly Mode, all the string information is transmitted on any one of sixteen MIDI channels selected via the MIDI menu 'Channel' option. In Split Mode, each string is assigned its own MIDI channel for multi-timbral playing, and these are preset to channels 1 to 6. Split Mode must be used if you wish to record your DG1 performances on a MIDI sequencer and play them back through MIDI In.
Along with Channel selection, the DG1's MIDI menu lets you activate Program Change so that DG1 programs will select the corresponding MIDI program remotely. If you wish to transpose notes played by whatever MIDI instrument the DG1 is controlling, this too is possible using the Tuning Offset option, whose range is plus or minus two octaves in semitone steps.
There's still plenty of scope for injecting expression into whatever MIDI instrument the DG1 controls, as its real-time performance controls - Bar, Fret and Strum - can be assigned to control MIDI pitch bend, aftertouch or any of 31 other MIDI controllers (such as breath control) that a connected device may make use of.
The final MIDI option is Program Dump which allows you to download DG1 programs to a sequencer or transfer them into another DG1.
The Stepp DG 1 has been well designed to feel and play as much like a guitar as possible - even though it is made of thermoplastic. It is nicely balanced, with a comfortable average width neck. The inlaid red LED fret-markers should look great on stage but, unfortunately, don't make up for the necessarily flat fingerboard which will take a bit of adapting to, as will the split string arrangement.
Bending strings on the very high frets proved difficult as they are close to their anchor points and taut. Then again, you don't need to bend them as far because you can set the bend range electronically to give a corresponding amount of pitch bend for a smaller movement of the string.
Personally, I don't care for the shape of the DG1 - its angular look reminds me too much of Bo Diddley's guitar. Though its sleek lines and satin grey finish give it undeniable sex appeal - and that's an ingredient not to be overlooked in a guitar.
Stepp's biggest problem, however, will be the mentality of the player and some guitarists will undoubtedly pick up the DG1 and be more concerned about the fact that it doesn't feel like their Strat or Les Paul. But why should they expect it to? A Spanish guitar doesn't feel anything like a Les Paul, yet it is still a valid instrument.
The Stepp DG1, in contrast, is a marvellous instrument that genuinely pushes forward the frontiers of your creativity as a guitar player. It encourages experimentation with new sounds and playing techniques, but doesn't force them upon you. Most important of all, it gives you back the natural expression you lose with a keyboard synthesizer.
It is hard to put a price tag on such an inspirational instrument, but when you consider that you get a sophisticated six-voice, dual oscillator synth along with the guitar, £3500 plus VAT begins to look a distinct bargain. After all, the nearest comparison to the Stepp DG1 is the SynthAxe and it is considerably more expensive at £6000 plus VAT, relies on an additional synth for its voicing, and requires the guitarist to learn new techniques to operate the trigger keys.
The Shadow MIDI guitar controller, alternatively, is priced well below the Stepp at £1000, but you must add on the price of a guitar and polysynth to approach the facilities offered by the DG1. Even so, the Shadow isn't a patch on the DG1 in performance terms, being so restrictive that it frustrates the hell out of you as a guitar player. The Stepp DG1 never does that. It always makes you glad you're a guitarist.
Stepp believe the DG1 is the natural and logical step in the evolution of the guitar. Having played it, I must agree.
Further details from (Contact Details)
Review by Ian Gilby
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!