by J. Arif Verner
History of MIDI Guitar 101
While it can't be argued that the technological interest was there, the survivors of this technology are few and far
between. Without a question, Roland is the only company to corner the market, make a profit and continue to develop
its product line.
But the same problem that has plagued MIDI guitars remains: How do you get accurate MIDI data from a vibrating
string. The story is not simple.
When a string is hit, the guitar synth counts the cycles of the string's vibrations to determine its frequencies. This can
take up to 20 milliseconds, or 2 cycles.
The initial string attack needs to settle into a constant vibration for the converter to read data accurately, including
the pitch (key number), velocity (amplitude values of 1 to 127) and pitch bend (if activated). MIDI Noton messages are
transmitted to the converter after the amplitude of the vibrating string crosses a predefined threshold level. Conversely,
Note-off messages are sent after dropping below the threshold. The guitar synth's sensitivity control determines the
response of the threshold level.
The bottom line is, after a string is hit, there is a slight lag time until the converter can read the data. To make
matters worse, lower strings have longer delay times (add on a few more milliseconds) because they vibrate more
slowly.
However, the real problems occur when the guitar synth double-triggers and "glitches." These errors are caused by
the synth reading erroneous information from the guitar due to sloppy playing, excessive overtones triggering inaccurate
information, or sensitivity controls not set properly to name a few culprits.
The good news is that MIDI guitar synthesis has come a long way in remedying these problems. With clean, accurate
playing and a guitar synth properly set up, the results can be awesome!
In the quest for glory, many companies have attempted different approaches to harness this elusive technology. The
three most successful strategies have been pitch-to-MIDI (also called pitch-to-voltage), wired frets, and ultrasonic pitch
sensor scanning. As with most things in life, there are pluses and minuses in any decision. MIDI guitar synthesis is no
exception.
Over the last few years, the dust has settled and the clear winner is pitch-to-MIDI. Yet it could be argued that its
popularity is due to the cost-effectiveness and ease of set-up - as opposed to technological superiority. Pitch-to-MIDI
signals are triggered via a hexaphonic pickup. As strings are played, six different signals are sent to the converter to
analyze pitch and other corresponding data.
Roland's guitar synths have made their mark as integrated units with built-in sounds. This "plug-and-play" simplicity
is a major part of their popularity. But beyond marketing, internal sounds track quicker than sounds generated from
external modules - to the tune of 4 to 5 milliseconds faster. Roland uses a proprietary protocal to make the synth's
communcation with the sound module more efficient.
Shadow Electronics of America is another participant in pitch-to-MIDI development. Their SH 075 Quick Mount
MIDI System is a self-contained converter box and hex pickup. The whole unit attaches to the top of the guitar. Shadow
keeps their costs down by providing only a converter - there are no built-in sounds, so you need an external module.
Also new to guitar synthesis are Ovation, Fender, and Godin. These manufacturers are pre-installing hex pickups as a
factory option. Fender will be installing the Roland GR pickup on their Standard Stratocaster model. Ovation will
produce two GR-ready electric and acoustic guitars. Godin's flagship instrument is called the Multiac. This nylon
stringed guitar uses a custom designed hex pickup. This allows the transducer to have direct contact with each guitar
string. In addition, the Godin, Fender, and Ovation models use Roland's 13-pin output jack for interfacing with their GR
series guitar synths.
Taking a different approach with pitch to-MIDI is an instructional system called the Lyrrus G-Vox. Designed
primarily for computer use, this package comes bundled with software and hardware. The nuts and bolts of G-Vox is a
program called Riffs. Using this software, guitar players can tap into instructional programs by the likes of Steve
Morse, Adrian Legg, Carl Verheyen, and Greg Davis. With their hex pickup installed, guitarists can play at any speed,
see the music and the tab on the computer screen, and follow finger positions on the fretboard. Sequencers can read
standard MIDI files created by the Lyrrus G-Vox.
Wired fret systems present another unique approach to MIDI guitar construction. These are custom-designed guitars,
as opposed to the retrofit nature of pitch-to-MIDI. As the name implies, wired fret guitar synths use wires
attached to each fret through the guitar's neck. Playing a note triggers an electrical contact with the fret. Decoding
circuitry in the instrument then deciphers the signal data. This technique increases the sensitivity of the note response,
improving the tracking.
The most infamous instrument was the British made SynthAxe. This bizarre looking instrument commanded prices
of $10,000 and up. Here in the US, Zeta Systems is producing their own wired fret instruments. In addition to the wired
fret circuitry, Zeta uses a pitch-to-MIDI hex pickup to track the pitch bend data. As one might imagine, the
sophistication of these instruments involves a fairly complex setup.
Ultrasonic pitch sensor scanning was another design used in MIDI guitar technology. Yamaha and Quantar produced
these custom made instruments in the late '80s. These guitar synths used a technique in which high frequency ultrasonic
sounds were transmitted across the strings. The fret positions were determined by a reflected wave - much like a
musical "sonar" system. These units were only MIDI controllers, and didn't generate any sounds of their own. Also,
they used six strings of the same gauge (approximately 0.4 mm/ 0.16") for better tracking. Ultrasonic scanning and
wired fret guitar synths tend to have quicker tracking than pitch-to-MIDI. However, pitch-to-MIDI has the smoothest
pitch trend response.
System setup
Installation of the pitch-to-MIDI pickup is fairly easy In most cases, you can use your own guitar. The hexaphonic
pickup is approximately 4" long. Position it as close to the bridge as possible (do not place the pickup more than 20
millimeters, or 13/16" from the bridge). There are six sensors on top of the pickup. Align them under each of the
corresponding strings. The best distance between the pickup and the strings is about 1.0 millimeter or.O4" when the highest
notes are fretted.
Roland's Control Unit is part of the hex pickup's configuration. This unit is a small box-like housing used to drive the
guitar synth. It includes several controls to adjust various synth settings. Double sided adhesive tape can be used to
attach the pickup and the Control Unit to the guitar's body. However, this could be bad for a guitar with poor finish. If
you wish to drill holes for a more permanent installation, make them about .O8" or 2 millimeters in diameter and 2 to 3
millimeters deep.
To route the guitar's audio cable, there are two choices. Roland provides a small cable that plugs directly into the
Control Unit from the guitar. But some guitarists prefer to go directly into the amp for greater control. Either way works
fine for mixing the two signals.
There are several considerations when configuring your MIDI guitar. First and foremost is the choice of playing
mode. The options are mono and poly mode. The term mono mode is somewhat misleading since it does not refer to the
standard audio definition of monaural. In the world of guitar synthesis, mono mode assigns each string to its own MIDI
channel - kind of like six different synths playing one note at a time.
While using this mode with pitch
bend, make sure your guitar is perfectly in tune. The synth's circuitry is designed to read the exact pitch of each note
you play; your synth will faithfully respond to every tuning discrepancy from your guitar. Also, not all external modules
can address mono mode. So check the owner's manual if you are connecting multiple units together.
As opposed to mono mode, poly mode puts all the data on one MIDI channel. In poly mode (or with pitch bend off in
mono), you have a +50 cent range of tuning error before the synth detects it.
So which mode is best? It depends upon your application. Mono mode provides the greatest control of the overall
sound. If you do a lot of chord work and pitch bend, then mono mode is your best bet. This is especially true of live
performance work.
For sequencing, poly mode has several advantages. Most importantly, it uses only one MIDI channel. Unless you have
a multi-port interface, your 16 MIDI channels will quickly be devoured with a multi-timbral sequence. Also, editing
mono mode can be a daunting experience with controller data scattered over six different tracks and channels. (Some
sequencers allow you to merge multiple MIDI channels into one. This will simplify your life.)
Poly mode is better for sequencing projects without pitch bend or large multi-timbral projects. You should also
consider using poly mode if you are only using pitch bend on one string at a time (as opposed to chords). Sequencing
bass guitar parts is a good example.
Another important set-up option is the pitch bend range. Standard choices are 0 (also called Chromatic mode) to 24
semi tones, giving you several octaves to play with. If you find your slides and bends sound too angular (sometimes
called "stair-stepping"), lower your bend range. When working with external modules, make sure they are set to the
same range. If your gear is older, you may not be able to set a bend range of 12 or 24. Also consider the sounds you are
using. Some sampled patches like piano, organ, brass, and bells sound terrible with pitch bend - so turn it off.
In addition to configuring your controller and sound modules, you need to consider the logistics of your sound system.
Unfortunately, being a MIDI guitarist puts us in a hybrid world somewhere between keyboards and guitars. The rules
need to be reinvented. You can't expect optimum results by sending all of your sound into a guitar amp. These amps are
not designed for the dynamic range of synths - they're designed for the punchy sound of guitars. Synths should be routed
to a pair of keyboard speakers.
In an ideal world, you would have a three-speaker system. This would consist of stereo speakers for the synthesizer,
separated by your guitar amp in dead center. And yes, you'll need to tame this configuration with a line mixer. No one
said technology comes cheap. However, if you do it right, it will sound great!
Sequencing
Part of MlDl's versatility is sequencing. Yes, there is a learning curve, but the extra effort is well worth it. Map out
your tracks beforehand. Which ones will have pitch bend and which won't? If pitch bend is on and you hit an open
string, your sequencer will read a whole plethora of controller data. This data eats RAM like a hungry dog, and in no
time, MlDl's bandwidth of 31.25 kbps becomes congested. If you're using a hardware sequencer, this could spell
trouble, so check the memory requirements of your unit. You can turn the pitch bend off on the sequencer or your MIDI
guitar.
Some sequencers have the ability to thin controller data. This might be an option - or rather a last resort. Realize that
you will be compromising some of the subtleties of expression in your performance. Then again, you may not have a
choice if you're out of memory.
To improve tracking even more, play your guitar parts in upper positions whenever possible. So if you are playing a
bass line, try it on the top strings and drop the patch down an octave or two. Since the wavelength is shorter on the
higher strings, it takes less time for the converter to recognize notes, which translates to a quicker response.
You might experiment with lighter gauge strings to see if there is a difference. There should be. And if you find that
open strings are glitching, try playing them as fingered positions. (Obviously, this doesn't work with the sixth string!)
Also, experiment with your sequencer's timing resolution (pulses per quarter note). Lower resolutions like 48 or 72
quantize the music more than higher resolutions like 480. This may be helpful if there is a slight lag time and notes
need tightening up.
If double-triggering or glitching is a problem, set the synth's threshold levels higher. If notes cut off too soon, then
lower the amount. This technology is a real balancing act; when the tracking speed increases too much, it may be at the
expense of faulty pitch recognition. Consider setting your sequencer to elimi nate notes with short durations of 15
milliseconds or less. The same holds true of velocities lower than 15. Try using a heavy pick. It will send a stronger
signal to the converter.
Finally, check out your MIDI guitar's Local Control settings. When set to On, the synth's internal sound generator and
guitar control section are connected. When set to Off, the sound generator is disconnected from the guitar control
section. Set Local Control to Off when sequencing. This will prevent the problem of data from the MIDI guitar colliding
with the same data as it comes through the sequencer back into the sound module. Don't forget to turn Local Control to
On when you're done sequencing, or you won't hear anything the next time you play!
Sound ideas
MIDI guitar has its own personality, and designing your own patches is half the fun. Here are some starting ideas.
Choose a guitar sound on your synth and transpose it up an octave. Add a slight touch of delay and mix in your regular
guitar. You'll get a neat 12-string harmonizing effect. If you're looking for a massive wall of sound, try running your
guitar through your favorite distortion unit and layering a brass patch underneath it.
Another popular technique is to lower the fifth and sixth strings down an octave on your synth. This will really fatten
up your sound. In fact, you can add a bass patch to strings five and six and piano on strings one through four. For a slow
chordal piece, try layering a clean chorused guitar with a light string patch on top. If you have enough polyphony (or
several other sound modules), try choral tunings. For example, tune one sound to a third and another to a fifth. As you
hit a note, your guitar and synth will play a major chord.
With your guitar synth, you can easily design open tunings. However, keep the volume down on your regular guitar!
Two popular tunings are D-A-D-G-A-D and D-G-D-G-B-D. You can even tune to a chord. Try tuning to an E or A
minor chord for starters. Since you can save these tunings with your favorite patch, there's no need to retune your "real"
guitar. Another possibility for open tunings is to use sostenuto with improvisation. Use the first and second strings for
an improv line and the lower four
strings with sostenuto as a drone. Realize that sostenuto acts in a similar way as a hold pedal. So pick a sound with a
long release time to its envelope - a thick string pad should work fine.
Don't forget to try sequencing with a drum kit patch. Each note will be assigned to a different percussion instrument.
Or to stretch the limits, try a different sound on each string. You'll need to use mono mode for this. And remember, with
synthesizers you need to play-the-sound. For example, use a strong picking attack for a sharp piano sound and keep your
speed runs to a minimum with delicate choir patches. You get the picture. Also decide on your primary voicing. Are you
using the synth for your main timbre or as ornamentation to your regular guitar sound? Conceptualize your plan of
action to maximize your sound.
Aftertouch
It's been an interesting 20 years since the days of the ARP Avatar. And even though MIDI guitar synthesis has
improved drastically, there's still plenty of room to grow. Will it ever become a perfect technology? Time will tell. So
stay tuned - the best is yet to come.
Midi Guitar Synthesis:
Yesterday, Today and Tomorrow
It was 1976.
The advertisement read,
"And it came to pass that
guitar players would
have the same performance potential as
keyboard players, for the
Arp Avatar would prove
it." Legendary synth
pioneer, Arp Instruments,
was riding high on the
success of their analog
synthesizers. The company
went out on a limb
developing a monophonic
guitar synth called the
Avatar. Unfortunately it
didn't work and the
reputation of guitar
synthesis was tainted for
years to come. However,
it could be said that the
Avatar was the birth of
Guitar Synthesis. It was
also the death of Arp...