Trying To Get That Pat Metheny Guitar Synth Sound - How To Do?

[jim-analog]

Greetings,

A bit of back ground info:

I've had and played Roland GR-300 and other gtr synths for over 25 years and am pretty familiar with the various early Roland models. As you've found, trying to build a polyphonic (hexaphonic) version requires 6x the circuitry and ends up pretty complex. As noted in another reply, the other big issue is the signal conditioning and the scaling of the resultant control voltage to track an oscillator at the 1V/Octave standard.

  Prior to the release of the GR300 (housed in a blue floor box that also required a gtr with a lot of circuitry in it), Roland made a rack unit called the SPV-355. This was a monophonic synth based on one of the "SH" series keyboard synths, but without a keyboard. It could accept a gtr, mic or any other similar source up to line level and did an "interesting" job of converting the pitch to a control voltage and the envelope to a gate function. I primarily use it to drive other more full featured mono synths, but if you play within it's limitations, it can do some cool but limited things by itself.

   I found one for under $100 at a music shop a long time ago. They had 5 or 6 of them NOS in a dusty rack that had been sitting there for many years. I probably could have gotten them all for a few hundred (wish I did). Well, I *would* suggest you get one, as it can do a decent approximation of the Methany tone using any guitar, but they are going for the better part of $1K now! The GR300 units are also pretty pricey and also require the appropriate gtr with the specific on board electronics. The cables/connectors that go between are getting expensive and hard to find as well (there are some work arounds for this).

The following product in the Roland gtr synth line was the GR700 (very large grey floor unit, that initially came with a goofy shaped gtr that had a stabilizer running parallel with the neck). There were a few other products released a bit after that, including a direct GR guitar to MIDI rack box and some other items. After those,  Roland simplified the system, particularly at gtr end. This allowed the use of a 9 pin interface on a round "DIN" type connector, rather than the 24 pin interface with the large rectangular connector. Many OEM guitar builders made "GR ready" instruments that used the 9pin standard (Fender, etc.). Of course, the 9pin and 24pin systems are not compatible without very rare (or home built) converter boxes. The 9pin systems have been undergoing continuous improvements for years now and are still currently produced.

I've heard that some of the newer (9 pin standard) GR hexaphonic units that use a digital synth engine have very good quality "GR300" emulations. You can probably find a used one for much less than the old analog units (that were fairly limited in tonal variety relatively speaking) and also have smaller, easy to add on hex pickups and electronics you can put on ANY guitar without permanent alterations.

If you go to Wayne Joness' page that was already mentioned:

https://www.joness.com/gr300/index.htm

He has the largest library of Roland GR info on the web. Plenty of sound examples, all the various brochures,  manuals and schematics for the early units (including the GR300 and SPV355). He has a comparison page with audio tracks of the "300" emulations done with more current units compared to the original.

If you have any Roland GR specific questions, I've been working with them for a long time and would be glad to try and answer. I have the prints for all the early stuff and actually at this time have a both a GR300 and SPV355 taken down for rebuild on my bench. I can get you plenty of inside pix or whatever else may help.

  My interest stopped not long after the GR700 era and I never got involved with any of the units after that, so you'd be better served at Waynes' page for info on any of the newer stuff. I still really like and use the old stuff all the time; it serves my needs perfectly. BTW, the old blue 300 is still probably the fastest and most accurate tracking gtr synth ever made, it's just limited by the primitive synth capabilities. The 700 and later versions, being MIDI compatible have no limit as far as sound possibilities, but they all suffer from conversion delay and tracking glitches to some degree.

Hope that helps you some and good luck with your project!

Regards, Jim

[jim-analog]

Greetings,

Correction to my above post. I couldn't figure out how to edit my post, so hence this one.

  Somehow I typed 9pin rather than the correct 13pin in regards to the newer type of Roland GR interface cable. Please just substitute 13 for 9 anywhere I mention in. 

Sorry for any confusion!

Regards, Jim

[pinkjimiphoton]

i believe its the vg99 that has the gr300 sims, its more a multi effect than a synth and requires a gk2-3 13 pin pickup.
wayne did 0-scope comparisons, its really close from what he'd written.
not gonna work in a home brew stomp, tho i don't see why ya can't change the shape of the waveform somehow to getcha in the ballpark.

[PRR]

> I couldn't figure out how to edit my post

On this forum, you get like 15 minutes grace-period to fix your typos, then the Modify button goes away.

[aron]

There's a pitch envelope on that linked video. To really nail the sound you need 2 oscillators. One of them have the attack rising then falling. Not sure if the guitar synth had "velocity" sensing. If so, velocity->filter envelope. That's key to changing the attack.

[pinkjimiphoton]

yes aron, guitar synths have velocity, tho they call it "sensitivity"... even the modern ones have it, so the early ones def should, as they were really just synthesizers with a guitar controller.
the last REAL guitar synth roland made was the gr1, which had knobs along the top you could tweak as well as digital menues.
since then they've kinda dumbed it down a bit. some sound amazing, but more like playing presets on a casio than classic analog synthesis. man. i miss my 2600

almost got what for me was at the time a holy grail, an arp avatar. but it was only partially working, so i passed on it finally tho i DID bid on it a couple times... it got too rich for my blood quickly, so i let a kid win it for a couple hundred bux. the whole system was at a local auction.

before i got into fx, i was really into analog synths. had a bunch of rolands from the juno series, 1, 6, and 60,  an sh101, the arp 2600, electrocomp 101 and a couple moogs, a micro, a mini, and a rat shack "rogue". i could never use both hands on one keyboard tho, so it wa spointless once i got into the rolaand guitar synths and sold/traded them all off. had wurly piano, c3 hammond, l101 hammond, a couple farfisas and a nice rhodes too.

guitars are MUCH easier.

sorry for the hijack paul!

[Rob Strand]

nteresting! I think you're right, my circuit has gotten pretty complex and it STILL doesn't work lol.

Seems like with some strategic filtering I might be able to get something I can live with.

Where is the info on the mod itself though?

If the guitar was a nice sine-wave the fundamental detector would be be easy.    However the guitar signal has a lot of harmonics.    If you put the signal through a limiter the harmonics stuff up the zero-crossings.  A compressor on the other hand changes the level but ideally it does not change the relative levels of the harmonics - so it doesn't fix that problem.

So one "strategic filter" is to pass the signal through a low-pass filter with a low-cut-off.  For sake of the argument 50Hz.   Take E string at 80Hz (approx)  the second harmonic is at 160Hz.   If the raw signal had equal levels of 80Hz and 160Hz it's hard to extract fundamental but after the low-pass filter the second harmonic become half the level of the fundamental and so the fundamental is emphasized and the detection process made easier.   If you pass that signal through a limiter the zero-crossings would be far better off than the raw signal.     The problem with the low-pass filter is if you have an E two octaves up at ~320Hz the signal level is 1/4 that of the E at 80Hz.   The filter still emphasizes the fundamental of 320Hz E just you have to handle the smaller signal level.  This is where multiple filters can help.

This is no means a final solution you will need other filtering.

Interesting to note, if you look at the front-end of the Roland unit is has a low-pass filter -> limiter -> low-pass filter. Then after that is the switched filter loop.    The filter frequencies are higher than 50Hz but that becomes more about making trade-offs.  It also has the switched filter stage to pick-up the pieces.   Anyway you can start to see why the Roland unit is like it is.  They definitely put things in to help reliability of the tracking.   It's a matter of how far you can back away from that before it falls in a heap.

[aron]

Someone told me that guitar->software works super well. Take the midi out and send it to the cheap low cost synths out there now. I know, not really DIY bUt it would work.

[pinkjimiphoton]

Someone told me that guitar->software works super well. Take the midi out and send it to the cheap low cost synths out there now. I know, not really DIY bUt it would work.

hit kvr!

[Paul Marossy]

More interesting info in the last few posts, thanks for all that everyone.

I made the Anderton Ultra Fuzz... wow that is pretty extreme, too extreme for me. Like it says in EPFM, it outputs a square wave. If I put a filter on the output it rounds the corners of the square waves and it sounds better to my ear.

I compared this to the Jordan Boss Tone with the "Trombetta mod" and tilt EQ & edge control tacked on the end, and these two circuits are quite different. The modded Boss Tone creates some very interesting wave forms... depending on how you set the controls of course, they can look something like a one-sided sawtooth with every other tooth missing as opposed to a square wave or massively clipped sine wave. Not at all what I was expecting to see. If I use the lo gain instead it looks more like what I'd expect.

In any case, the modded Boss Tone is quite cool. If I stay on the G, B & E strings I can get some cool horn-like sounds that sound good using lines like something a horn player would play. The sound I get is also very dependent on what pickups I am using. Seems to sound best with the bridge & neck pickup combo on my Strat (modded with a bridge on switch which gives me eight different sounds). That's a little more hollow sounding and seems to somehow be the most horn-like.

[Mark Hammer]

Not that I wish to make this more complicated for you (and I'm really glad the Trombetta thing is partly sating your urges), but you might want to consider adapting the Beausoleil Square-wave shaper to the Ultra-Fuzz output to get less harsh waveforms.

[Paul Marossy]

Not that I wish to make this more complicated for you (and I'm really glad the Trombetta thing is partly sating your urges), but you might want to consider adapting the Beausoleil Square-wave shaper to the Ultra-Fuzz output to get less harsh waveforms.

Thanks. This looks like a Tim Escobedo schematic. I'm thinking of the Ultra Fuzz as more of a building block for a more complex guitar synth that I might try to pursue in the future but I wanted to at least hear it in its naked form.

[Mark Hammer]

Correct.  The original, as indicated, was an old article in POLYPHONY, which Tim adapted.

The Ultra-Fuzz probably should be thought of as a building block or starting point for something larger.  This circuit simply shows how to turn a square wave into something else with a different set of harmonics.

[Paul Marossy]

Correct.  The original, as indicated, was an old article in POLYPHONY, which Tim adapted.

The Ultra-Fuzz probably should be thought of as a building block or starting point for something larger.  This circuit simply shows how to turn a square wave into something else with a different set of harmonics.

I was wondering if it would sound any good to take the Ultra Fuzz output and convert it to a triangle wave or a combo of the two. That was kind of my original idea on my guitar synth concept.

[pinkjimiphoton]

paul,
with any of these "brass" emu kinda things, take a clue from  your tagline.

play with your fingers, not a pick. i think you'll find your control over attack will have much more effect,
you can "coax" tones out with your fingers that are impossible with the attack of a pick.

misbiased fuzz like the t-bone circuit you can coax slow attacks and some very cool envelopes with your fingers that are
literally impossible with a pick.

when i worked up my model 1 driver thing years ago, i had a guy who HAD to have one.
so i built him one, exactly like mine, same pcb and everything. sent it to him.

he went batshit. "i can't get those sounds out of this, its just another crappy fuzztone".... i told him to use his fingers. suddenly he could get all the shit on the damn demo video, if he set the knobs right and played with his fingers.

he wanted his money back. it was more than 3 months after he'd recieved it. i said dude, if ya contacted me right away, i'd gladly accept a return, if it weren't a special order... but 90 days later? no can do. i had to order parts specifically to build it for him. never heard from him again really. oh well.  he settled down when he tried it with his fingers instead of a pick.

try it and see if it matters to you, too.

[Paul Marossy]

paul,
with any of these "brass" emu kinda things, take a clue from  your tagline.

play with your fingers, not a pick. i think you'll find your control over attack will have much more effect,
you can "coax" tones out with your fingers that are impossible with the attack of a pick.

Depends. In this case, I actually get better results using a pick. But I do use my fingers a lot in general. Just depends on what I am going for exactly.

[Paul Marossy]

One last comment before I consider this a closed case for me.

Now I can get some pretty decent Metheny-ish sounds with a Stratoblaster and modified Orange Squeezer in front of it and a delay behind it (speaking of the Jordan Boss Tone with "Trombetta mod" and a tilt EQ & edge control on the output). The tilt EQ and edge control help to focus the sound more specifically, so it seems to actually be an instance where the useless tilt EQ actually has a use. A side benefit is that I also get sounds that are reminiscent of my fretless electric experiment when playing on the E, A & D strings. Maybe I'll do a video on it that no one is going to watch, maybe not.

Anyway, thanks for all the help and interesting conversation. I learned a few things.

[Mark Hammer]

I always recommend to parents of young children that they should keep a list on the fridge door of all the foods they introduced to that kid that the kid loves.  That way, when you try and introduce something new to a young picky eater, you can point to the list, rattle off the names of things, and establish your cred as a reliable source.  "Remember when I told you about chocolate, and pudding, and ice cream, and apple slices, and French toast, and noodles....".

It's good to have a reputation.  Enjoy the broccoli, Paul.  It may have seemed strange, and not what you were hoping/looking for, but it tastes good. 

[Paul Marossy]

It's good to have a reputation.  Enjoy the broccoli, Paul.  It may have seemed strange, and not what you were hoping/looking for, but it tastes good. 

I like broccoli. I guess for me it would be squash. Yuck.

[pinkjimiphoton]

mmmmmn. squash. pick it, slice it, throw it in a pan with a little butter, add salt and pepper, shit , i could live on it!! 

that said... i AM weird.

also, that said, congrats on finding the tone ya needed.

i may only be one, but i'd love to be one of the only people to never see it. i lie. i would love to see video, but don't do it on my account alone.

rock on, paul!