Looking for software synth to go with MG2 that produces authentic Roland GR-300 sound

I don’t know if this forum is the right thing, because it’s really any other software forum, of software synths that emulates sounds of the past.

But since I think that there’ are a lot more guitarists trying out the MG-2 they MAY have had some experience with the ancient analog guitar synth from Roland, the GR-300. It has/had a special “turkish trumpet” sound that was frequently used by Pat Metheny, Robert Fripp, et al, in the 80s.

Thanks to its ramped phase locked loop waveforms it didn’t guess, the notes, and as such was one of the more playable guitar synths of the time, although limited in waveforms. Now, most people say “any freeware synth with a sawtooth waveform will do” but I do not agree with them. The difference between GR-300 and all others, is that in the front end of the note, the initial transient attack, there’s a “grit spit” similar to someone playing the trumpet, that when the note “is struck, blown” the very very first part is like something it can’t track and always interpret it as a “grit spit”. This was dynamically sensitive too on the GR-300 and if you played legato lines it was gone, but when you hit it with the pick attack, that peculiar and - very - unique “spit” came forward.

I wonder if there’s any VSTi plugin out there you can use which mimicks this one down to a tee? I’ve tried Zebra from U-he but it doesn’t really cut it. And loads of others, like Korgs legacy collection and so on and it seems very hard to produce that peculiar and “cheesy” sound. Many people hates it, but I think it’s full of balls.

1 Like

Thank you, I’ve checked those out by listening. I will try to use the KONTAKT 5 one, which is a sample though. But when just listening to them on the jjones site, it’s still all those “non-grit” sounds. Generic sawtooth, with some initial attack that is sort of “mimicking” pitch rise, and fall, or so to impersonate that true GR-300 “grit-spit”.

I remember when owning a GR-300 for a while, when testing it, I was “coaxed” by its sound to just tuck chords with my fingernails, as a classical guitar player, and left the pick behind, and didn’t strum. After a while I thought “Hey, I am only playing horn section lines all of the time with max 4 notes at the same time”… it was what it was good for.

Regarding Jjones I think it’s funny he has tried to mimick the sound on other synths both keyboard ones and “real” guitar synth ones, because he himself has stated:

_" It is not that the GR-300 ignores fret board misfires, or translates them into wildly inaccurate notes as some MIDI systems will. The GR-300 instead creates a unique analog synth equivalent. For example, the initial atonal pick attack is converted into a sound very reminiscent of the “spit” sound heard at the beginning of a trumpet phrase."

_“The trumpet comparison continues: the GR-300 has a waveform very similar to a sawtooth, with a brassy, aggressive tone. But the GR-300 waveform does something unique: it changes shape as the player moves up the fret board. An “E” played one octave above an open “E” string will not only be sounding at twice the frequency, but the harmonic content will be very different as well. This is the happy consequence of the brute force synthesis used in the GR-300. Inside the GR-300, the amplitude (volume) of each note is related to its pitch. The low “E” will have twice the amplitude of an “E” one octave above. Similarly, the high open “E” string, two octaves above the low “E” string, will have one quarter the amplitude of the low “E”. To eliminate this volume difference, Roland used a “chopper-gate” circuit to basically crop the top of the waveforms. The higher notes look more like a classic sawtooth waveform, while lower notes have a more rounded tone. The result is a rich, complex sound where every note played on the guitar across the instrument will have an individual sound. An unfortunate consequence of this design is that higher notes will have less sustain than lower ones.”

" 4 - Square Wave to Sawtooth Wave to GR-300 Wave

patent image

Other guitar synthesizers used different methods to try to detect the peaks of a waveform. The idea is that the time between waveform peaks will determine the pitch of the note played on the guitar. Roland took a different approach, and is probably the only company that tried zero crossing techniques to detect the pitch of a guitar signal. While it is possible to have several zero crossings in the guitar cycle (resulting in octave jumping), the adaptive filter is so efficient that this rarely happens. Here is an explanation of the diagram on the left:

A. Raw input guitar waveform

B. Square wave created by processing the input waveform through filtering and zero crossing circuit.

C. 1 uS pulses created by the leading edge of the square wave

D. Sawtooth waveform created by 1 uS pulses.

E. Distinct GR-300 waveform after “chopper-gate” clips the top of sawtooth waveform.

Phew! Such much processing ending up in a limited sound… but what an unique one and devoid of kludgey glitches, and octave jumps…

Oh yeah… love that Fripp GR 300 --> GR 50 trumpet. I have come pretty close using -->

–> JX 10 AM Synth (Trumpet) +
–> Transposer (Octaves =0 / Semitones = 12) +
–> FX - Delay (standard cfg) +
–> AMP/FX - Tube Amp (standard cfg)+
–> Cabinet - Cabinet 1 (std cfg)+
–> REVERB = HALL

PATCH / TRACKING / ARTICULATION
TRACKING --> NOISE GATE = 5
ARTICULATION --> BENDS = ENABLED (RANGE 12) / LEGATO = ENABLED / 100MS --> AFTERTOUCH = DISABLED
MIDI VELOCTIY --> GAIN=3 TONE=11 CURVE=14

There are still several adjustments which are required but, the above represents a good baseline.

Take Care my friend.

SAW8

Thanks SAW8. I have tried the JX but not exactly those settings but something similar. The main thing most post GR-300 synths lacks is that “trumpet spit”. But I will try that one.

What I think people misses out (samples is really out of question here, they will NOT produce anything close at all) is whenever trying to cop that sound misses out on methods to mimick that initial quick 5 ms of a “blur” or “fart” or “spit” …“grit”.

The thing with the GR-300 when I had it for a brief period, it wasn’t good at “strumming” any sound at all, it was almost best at monophonic lines but didn’t wreak havoc if you accicentally 2-3 notes or 6.

The main difference from all subsequent guitar-synths and software impersonations, is that the “trumpet spit” is only heard whenever leaning in with the pick. If you do legato playing with hammer on and pulloffs with the left hand, the waveform sounds/works just like any othere regular sawtooth.

On the other hand, whenever trying those sounds out today, you tend to go into some kind of automatic mode, and do play Metheny licks to no end, because that sound, just inspires you naturally to be steered in that direstion. Or Fripps stuff. Fripps licks are more angular, linear, pointillistic, and contains no jazz phrasing at all like Metheny does with his “glissandos” before hitting a note.

I came to the conclusion what intrigued me the most with the GR-300 sound/waveform was its ability to separate picked/plucked attack sound from the legato ones. Whenever plucked/pick the “trumpet grit-spit” was there immediately, but when resorting to playing legato lines only, it was not there. Hence the ample use of Metheny like phrasing. the gliding notes didn’t go “thpbweee-thpbweee” (2 notes) but “thpbweee-eeee” (2 notes the eeee is legato, glide playing).
That’s the best onomatepoetic description I can give you, alas.

Hey Mats,

trumpet spit” - Noted. and exactly pointed comment. From a technology prospect we have exceeded the capabilities of the earlier days of the Guitar Synth. The “trumpet spit” (love that descriptive term) is even near impossible to achieve on the GR-55 and GP-10 but, I am very optimistic about the ability of MG2 to configure this simply due to the vast array of controls, effects, and nuance management built in. BTW, I actually met Fripp in the late 90’s, and found that was a bit discontent about the inconsistencies (tracking, complex manipulation/configurations) of the early GR-XXX units. Irony, gotta love it :smile:

Yes, now when the above description is made, how that filtering and chopping off works, it wouldn’t be that impossible to achieve inside any software, digital modellerer or even a bona fide DSP.

The thing is at the time, all other Roland guitar synths, even the “first” GR-500, later 707, and so on, didn’t sound the same. You instantly heard that it was a Gr-300 anyone was using. Not the “old” one, or the newer one.

Allan Holdsworths behemoth Synth-Axxe controller, sounded exactly like any keyboardist playing, because that one hooked up to any external synth and played sounds verbatim, and you could not tell any difference if it was executed on keyboard or guitar controller. Perfect triggering of fast notes, because it didn’t rely on what pitch the strings had at all. That’s when I stopped listeing to him. He had a keyboard player in his band too, equally skilled, and it was nearly impossible to hear who did what and when. So it kind of went boring… after a while.

Remember one thing, that whiel Fripp could use the Gr-300s more alternate sounds with slow attacks, and automatic vibrato (I hate his synth solo in the middle of Elephant Talk, it can’t get any cheesier, it sounds like mrs Miller is trying to sing “I could have Danced all night” a musical number from the 50s…eeww!) Pat Metheny basically use one sound only, all of the time, maybe jumping up an octave or so. It’s pretty much one preset… that “turkish trumpet” (a phrase coined by Fripp by the way).

https://www.vguitarforums.com/smf/index.php?topic=17528.0

BTW it is Joness, who runs that site, and are infatuated in Gr-300 who coined the phrase “trumpet spit”. Not me. I thought it was a transient attack reminiscent of any similar horn sound, brass instrument. As he said the only instance of where he has detected that they use “zero crossing” as a means of triggering or pitch detection.

Yep, that’s messrs Fripp in a nutshell I alwyas refer to Fripp as messrs, because he always alludes to himself in interviews in third person. Which is a big alarm bell going off for me regarding some abbreviation letter ailment. He is the bona fide fakir, masochist, self-agonising. Still doing and executing things on him that he doesn’t really like at all, using gear that shows a lot of idiosyncrasies technologically, and airs his views on the shortcomings, but still uses them to no end, even if it causes him major gripes.

1 Like