Not going to get poly aftertouch of the GS1.

I think Montage Alg 14 can cover half of the GS1 (Carrier 2 in the GS1). Carrier 1, where two modulators are chained - the second modulator fed back to the 1st - then the carrier receiving the output of the first modulator (with 2nd modulator as fed-back input) is not something I can see is available. "Tapping in mid-stream" after the 1st modulator in a Montage Alg 14 type feedback loop.

Alg aside - I didn't hear anything in recordings of the GS1 that sounded far from what most FM EPs sound like. It doesn't "stand out" as having a totally unique fingerprint. The swirly sound I think you are describing is when you are in "Ensemble" mode - an effect. Some chorus or other effects should be able to put you in the "Ensemble" zone of the GS1. It's not the feedback loop that's doing that from the EP sounds I've heard.

Source for sound samples: https://www.native-instruments.com/forum/threads/yamaha-gs1-electric-piano-patch.50986/page-2

I also watched a youtube video of someone fiddling with a GS1. Much worse fidelity than the link above, so not bothering marking that as a source.

... follow up: Go to [CATEGORY SEARCH] and type in "GS". You'll find "GS E. Piano". Go figure.

If you look at the alg. chosen - it doesn't attempt to recreate the "architecture" of the GS - but using non double-chained-feedback loops, the Montage programmers came up with something they saw fit to call "GS E. Piano". This matches my first thought hearing the E. Piano that there is nothing wildly unique that the specific alg. / feedback would give under the GS series.

The two PARTs (1 and 2) alone sound un-Ens like and together create that chorus type effect. Looks like two copies of the same FM part with PART-level detuning +0.8 and -0.8 hz to give that detune chorus effect.

AF1 gives you more of that.

AF2 does something - not digging in. Superknob changes the character.

I'm not going to "A-B" the recording of the real thing vs. this performance - but you can see how close the performance gets to what you're after.

... adding: also, doing your own search you'd notice the "GS Tines AF1&2" which is likely another take on a GS series E. Piano It's sampled and therefore does not give the luxury of further FM programming manipulation. As an AWM2 performance (PART) - you can combine this as a static block with an FM-X layer to create something new and still retain some FM programming.

Richard wrote:

I've been trying to recreate the legendary GS1 swirly electric piano sound and finally came across the FM algorithm used in the GS1.

It is 2 pairs of 4 operators. Each pair of 4 operators is 2 carriers each with a modulator. Each modulator has feedback, but not to itself. Each modulator feeds back into the adjacent modulator's input. It's called cross-carrier-modulation.

I have to believe that this configuration was a key to to GS1's unique sound that quite is elusive to this day. Montage's selection of algorithms doesn't have this capability (unless I missed something) and was wondering if anyone knows why. Maybe Yamaha could implement it (firmware update feature?) so we can be backward compatible to the GS1 just like then DX7

Hi Richard,
We first have to correct you on some of the facts. First the GS1 was a dual four Operator system... yes eight total. There was only one algorithm and it featured four 2-Operator stacks. (See Algorithm 67). I can recall a conversation I had with Dr. John Chowning on the number of Operators and the potential. The two Operator stack was the backbone of all GS1 sounds.

Your idea that Yamaha some how forgot or didn't put an Algorithm in the MONTAGE borders on "conspiracy theory" thinking and I can assure that this is simply not the case. Fortunately, many of the original GS1, GS2, CE20, CE25, and DX-series programmers are still with us. (Hopefully, in the near future we can get interviews with some of them posted here on the site).
See the recent interview with genius programmer Scott Plunkett, he's been around with us since at least DX7II days... as long as I've been around here.

The number of Operators (8) and the four 2-Operator (Modulator:Carrier) stacks were enhanced with a vibrato (pmd) and tremolo (amd) which could be applied via the soft and sostenuto pedals- and there was an early version of the now classic Yamaha "Symphonic" effect- which is the secret to the swirling sound that everyone fell in love with... but this Effect Type while known to those close to the FM phenomena remains a "hidden" gem that has been in most every Yamaha Synth post formal effect processing become standard in synths (The DX7, TX816, etc., etc did not have Effects, period... not until the SY77 1989 did you have FM synthesizers with proper built-in Effects.

But there was that Symphonic 'effect' Type in the GS1 (and if there is a *signature* stamped on that coveted "GS1-sound" it is a combination of the two-Operator stacks with the movement of crafty "detuning" (there was a random detune) and the chorus-like thickening applied with the Yamaha "Symphonic" Effect Algorithm.

Now the GS1 was 5-6 years before I joined Yamaha (I was still in retail, where I sold three of those GS1s) but I was way (way) into everything I could learn about FM, but from what I subsequently learned that "Symphonic" Effect thing was developed to give greater dimension and to help create an ensemble type of sound. A little dab will do you... this was at a time fullfledged programmable Effects were not yet included in keyboards (it was unheard of at the time) so it was either On or Off. I think it was just an ENSEMBLE On/Off button.

If you remember the original GS1 - the Library concentrated on brass, string, mallet, flute, and the e.piano type sounds...

Each 2-Operator stack could define a complex sound component within the total result. So you could have an attack knock, a full bodied tone, a thin tine sound... etc. often what the first four Operators did was mirrored and delayed in the opposite set of four... while detuning (which got perhaps overused in the late 1980's FM 'programming') it was the results obtained with the Symphonic effect that gave the GS1 that sheen. The quality of the audio path didn't hurt either! When a keyboard is $10,000+ you can bet no expense was spared on the sound quality... it was like butter, polyphonic Aftertouch, walnut baby Grand like cabinet, inexcess of 200 pounds, leather booklet containing the program strips,

See the tutorial here in the "Synths" where we explore the potential of the two-Operator stack - "An FM-X Exploration" showing the wide variety of tones that can be recreated... from metallic to wooden, from hollow to dense...

While I try not to argue with people over what they hear (we all hear things differently to some degree) the facts are as I've stated them about the number and configuration of the Operators. And any thought that Yamaha is some how unaware of just about anything concerning FM (that would allow products to come out that couldn't reproduce what the GS1 did) is just NOT the case.

That said, the AFM engine (circa 1989-1992) so we're talking SY77, TG77 and SY99 introduced some extremely useful wrinkles to the whole FM programming arena... and it was not so much the ability to route AWM2 (samples) into an FM Operator (RCM), but the ability to create "nested feedback" loops... now that had potential.

The AWM2 sample directly modulating an FM Operator very (very) quickly goes to noise (elegant noise, to be sure, but noise nonetheless). But hopefully at some point we can get Dr. Manny Fernandez to talk a bit about that feature... "nested feedback loops" is an FM programming rabbit hole that opens the door to yet another set of infinite possibilities. You mention "cross carrier modulation" I'm not aware of this in the GS1 - but I can ask and perhaps find out more.

Definitely, read through Dr. Fernandez' articles here on YamahaSynth.com as he introduces you to basic FM programming using the refaceDX...
The little refaceDX is, yes, the equivalent of a GS2, except it has Effects and is programmable (and weighs a whole lot less).

Thanks for the questions, hope it helps.

Bad Mister and Jason,

Thanks a lot for the very helpful and thoughtful replies! I'm amazed how much time and effort you guys put in to answer my question and other questions in general!

Also, Bad Mister, thanks for the excellent historical background and detailed facts regarding the GS1. Your few paragraphs contain so much more info on the machine than everything I've been able to scour from the internet combined. It would certainly be great if one or more of the developers could be interviewed sometime in the future!

That would be great if you could find out more about the cross carrier modulation (and what it achieved sonically just out of curiosity). Attached is the block diagram I captured from the site Jason referred to in his reply.

Thanks again!!

It would certainly be great if one or more of the developers could be interviewed sometime in the future!

Please have a look at the links below:

Interview with Dr. Chowning the Father of FM

Basics of FM Synthesis by Dr. Manny Fernandez one the world's best FM programmers. Complete series of articles and downloads follows.

Behind the Synth -Scott Plunkett One of Yamaha's top programmers

Richard, I think this is a matter of terminology. I couldn't find references to the term "cross carrier modulation" in literature. The same limited search also didn't find "nested feedback loops" - although I do like this term as a way to describe what I was calling (my own language, no common terms) as feedback using chained modulators and also "tapping in" to the middle. The literature I found didn't cover the GS variant. When I looked at the same schematic (block diagram) you posted, I got what you were saying about "cross carrier modulation". Two modulators would supply output to two carriers - both to it's own "stack" and crossing over to a neighbor's "stack".

The software world with FM simulators has an opportunity to "test pilot" some interesting FM structures. Since some are open source - they're also free to be tweaked to offer any feedback structure you would want. FM8 has an arbitrary algorithm construction - and also allows some arbitrary feedback structures - so it seems you can setup the GS-style using FM8 or experiment with other things. Not wanting to bring software to a gig, I would use these more in the studio for experimenting.

BM mentioned using AWM as an input to FM. Of course Montage does not have that - it is an interesting feature of the SY (and other) keyboards mentioned. Here, FM8 also allows for audio data to serve as an input the the FM synthesis rather than fixed waveforms (sine, etc). So you can experiment with this using FM8 at cost - or possibly other free plugins (not going to look).

I do like the option of being able to route your own modulator paths including feedback in FM8 instead of having a fixed menu of preconceived ones. Naturally software is going to allow for more flexibility since it's relatively easy to grow software resources vs. hardware. I'm not really running a commercial here for FM8 - but some of the things you have expressed wanting to do (if matching the GS structure is your goal - forgetting about matching the sound that the structure produces - because I think you can get the "sound" down using Montage even if the structure is different) can be done with FM8 including the side-bar discussion about modulating samples (arbitrary waveforms) through FM.

I'm not sure if any of the free FM plug-ins are as flexible (Oxe, Dexed, etc) as they, to some degree, seem to be concerned with emulation of the DX7 as a starting point. FM8 is more building-block oriented without trying to adhere to limitations imposed by one keyboard platform or another.

I haven't found any reference yet to the actual original voice settings of the E Pianos for the GS series. If carrier 1 is zero'd out - then you could have 100% alg. compatibility with the GS. But again, not sure that is in any way a limitation as you can get something that sounds the same using one of the 88 alg. in Montage - combined with multiple PARTs with even more of either the same or a different alg. mixed in.

Another tangent:

I know some of these "why didn't Montage include XYZ feature that was in an old Yamaha keyboard" or (as I prefer to table them) "I wish Montage would have included XYZ feature that was in an old Yamaha keyboard" threads may seem to bash Montage. I know any keyboard has to make some choices on what it will and will not include. You can't have it all - there's give and take within the technology, schedule, cost, etc. limitations at each generation. Rather than seeing these discussions as any dig on Montage - I think of these discussions in a more positive light:

a) Isn't it great that Yamaha's catalog of technical achievements is so great - they do have such a legacy of innovation that in 2017 you're looking at Yamaha's late 70s tech (released in '81 - but design phase may have begun in the late 70s - rounding down for dramatic impact) as still relevant synth tech today? Similar threads have discussed the FS1R (... or VL, or AN) in a similar light.

b) I love to see technology I was not previously aware of (like this GS discussion) "dug up" and now I understand some of the presets in Montage better (the "GS" ones - a couple of them). I get a kick out of seeing what other vintage synths can do which gives some inspiration for matching qualities of them when programming my own performances. It's also interesting to see how vastly limited some of these older synths are with respect to programming interface (a big dedicated computer system that only served as a voice programmer), storage media (magnetic strips), or limits in the dynamic modulation of parameters which now motion sequence/motion control gives us.

c) I think it's great to look at some ideas of the past and now with easier programming methods, more dynamic control options, more "computing" power/etc - how some great ideas in the past may be "retooled" for future use. We may never get FS1R formants back. May never get arbitrary algorithms to allow "constructing" the GS algorithm, etc. But I think it's good to take a second look and have the music community (players) chime in if there's anything of value they missed the 1st round due to problems of yesteryear that are now remedied by technological advances. I'm not sure the GS Electric piano sound serves as a great argument for requiring the feedback loop structure (since you can get that sound without it) - but perhaps someone more in the GS "know" can chime in on something unique that cannot be otherwise obtained now we even have the harmonics choices (odd, even, all - skirt, etc).

d) Often we arrive to a point where a suggestion demonstrates "Montage can model that". Not using the same methods - but we take a different path to show that the levers and dials available to us do reconstruct some vintage sound. Sometimes we show the opposite (DX7 glissando, etc) - but more often than not - there's a way.

Bad Mister wrote:
Your idea that Yamaha some how forgot or didn't put an Algorithm in the MONTAGE borders on "conspiracy theory" thinking and I can assure that this is simply not the case. Fortunately, many of the original GS1, GS2, CE20, CE25, and DX-series programmers are still with us. (Hopefully, in the near future we can get interviews with some of them posted here on the site).

With due respect, I think Yamaha and other manufacturers does this all the time, in the sense of introducing good ideas, then abandoning them, and only occasionally bringing them back (I agree with Jason that there may be good reasons for this, but not all the time):

- The random detuning of the GX-1 and GS-1, missing from the DX7, reintroduced in the SY77, not present in the FS1R, back again in the Montage
- The pitch EG in the DX7, missing from most 4-op synths, except the DX21
- The looping EG:s of the SY77. OK, there's the motion sequences in the Montage, but it's not the same.
- The operator phase parameter of the SY77 - subtle, but eminently useful.

But complex waveforms and filters have been present in most Yamaha implementations since they were introduced (in the TX81Z and SY77, respectively).

I believe one could emulate the rumored GS cross mod on the Montage by using algorithm 11 or 12, and copying operators 4 and 7 to operators 6 and 3. With complex waveforms, op 1 and 2 could then be used as a substitute for the other two stacks, or one could use another part for that.

If the original programmers (and developers?) are still with us, surely they could tell us if the cross mod indeed was present, and why Yamaha chose that particular algorithm.

Bad Mister wrote:
The number of Operators (8) and the four 2-Operator (Modulator:Carrier) stacks were enhanced with a vibrato (pmd) and tremolo (amd) which could be applied via the soft and sostenuto pedals- and there was an early version of the now classic Yamaha "Symphonic" effect- which is the secret to the swirling sound that everyone fell in love with... but this Effect Type while known to those close to the FM phenomena remains a "hidden" gem that has been in most every Yamaha Synth post formal effect processing become standard in synths (The DX7, TX816, etc., etc did not have Effects, period... not until the SY77 1989 did you have FM synthesizers with proper built-in Effects.

I don't know if it counts as a "proper" effect, but the DX21 did have a 128-stage BBD used as chorus. It's very noisy 🙂

The GS series was the first commercially available product with FM synthesis. So it was, in terms of what followed, a prototype for a new generation of synthesis. The custom logic and bits and pieces Yamaha had available to them probably lent itself to the structure used. Usually the first out of the chute is more of a science experiment and then you refine on that. They had to cost reduce - the GS1 just about had the kitchen sink in terms of being like a Clavinova type fancy exterior with full pedals and polyphonic aftertouch etc. etc. Less ended up being much more (DX7 less cost and less, on some fronts, features; more sales). It's a balancing act.

The thing that "gets me" about the discussion of the GS-1 "algorithm" (modulator to carrier flow) is that I haven't seen an identified sound that is really needing the more limiting (albeit different) GS-1 structure. I'd be interested in learning/hearing if there is such a thing.

We do have half the structure of the GS-1 available.

When you make two copies of the same modulator and stack them - this isn't the same as feedback because there's just a single iteration of "the loop". You can take a 2-operator stack and use the modulator with feedback. Then go to a 3-operator stack, no feedback, but make the two modulators duplicates of the modulator in the 2-operator stack. You'll see the sound isn't the same. It is not true feedback - there's not a loop. There's just a single pass.

There are taller stacks - I didn't see if the 8-tall stack would sound close enough if I stacked 7 of the same modulator. I have a feeling like there's a fundamental difference - between what you are suggesting as a way to reproduce feedback and how the built-in feedback loops work.

Also, imagining how the phase of a signal works going through a loop - I'm not sure if phase shifts as it makes its way though a loop. Something that would not happen when you stack - the phase would (maybe) act differently. This is an area of detail I'm not solid on with respect to FM. The difference may be there but too subtle to really be a difference you can hear. Given processors - even in the 80s - are much faster than sound - this isn't likely to make much of a difference.

I believe the lack of the GS-style structure makes it harder to program (because you just can't copy the magnetic strip programming verbatim) - but the flexibility of the Montage system very likely allows for an indistinguishable sound from what you can program on the GS series. I could be wrong - but haven't heard an example of anything yet that didn't remind me of the character of Montage FM.

Jason,

Great points about using FM8 for experimentation! I've heard of it but never used it -- I'll check it out. I'm also planning to use Matlab to try some experimenting with different routings since I'm comfortable with the equations and it gives full flexibility to program anything.

On your tagngent comments I fully agree. My intent was in no way to bash the product or to call into question the motivations of Yamaha -- it was just an objective question -- maybe I should have posed the question more like "can the Montage to this....?" rather than "why didn't the Montage do this....?".

This is an amazing instrument and I am blown away by the capability (including programmability) and sound quality. I'm looking forward to trying some of the things you suggest. I haven't checked for sure, but if I can assign part and carrier detuning parameters to superknob/FC7 or other modulators like the wheel or aftertouch, it might be a nice way to bring in or change the the effect.

By the way, my favorite example of the GS1 EP is on Yellowjackets "Revelation" on their Shades CD (YouTube has it in case you want to take a listen). There is an organ with Leslie mixed in that makes it sound even cooler. In the last maybe 20 seconds of the song the sustain sound is just unbelievably cool. *That* is a sound I'd like to achieve. The in and out beating almost sounds like a very slow Leslie but it's probably a detune beating along with the tremolo pedal. Maybe the aftertouch is modulating something as well.

Great discussion!!

It's hard to pick out for me since my computer setup is low-def and there's a lot stacked on top at the end there. I didn't make any attempt to pick out a good starting tone. In fact, I happened to have some FM performance that I had switched around the Alg - not because I wanted to - but messed up. So it's a kind of strange start - I just put the rotating speaker over it - because mostly the rotating speaker makes anything sound like ... it has a rotating speaker. I'm using the old rotating speaker effect - not the new souped up one.

Since the Leslie inside my transistor organ made its plastic sine waves sound much more like gravy - I think a rotating speaker does a lot all by itself without having to worry too much about what's under. There's more leeway for having something imperfect as the pre-speaker sound. This FM sound is too bright, but it's not hugely off given how different this sound is than probably the YJ sound before the Leslie.

https://clyp.it/x5vtjxpt

My intent was in no way to bash the product or to call into question the motivations of Yamaha -- it was just an objective question -- maybe I should have posed the question more like "can the Montage to this....?" rather than "why didn't the Montage do this....?".

It is a much better way to approach the subject. Because the engineering team is not tasked with necessarily repeating something from the past (the past, something already known is a comfort zone for customers... they fall in love with the way it used to be done). The engineers, perhaps, are targeting something new and heretofore, not possible. They may feel like ...we've been there done that and have the t-shirt! So their focus maybe looking in the opposite direction.

It's interesting, when you do marketing surveys you discover people will design you a product that is a combination of features from several existing or previous products; talking with engineers... they are thinking 'hey, this opens the possibilities for some new music'.

There are several interviews with the engineers you can read and watch - and what you take away is that while there is a tip of the cap to past methods and features, they are overall looking to push the envelope towards what's new and different. Sure one of the things FM-X can do is recreate some of the sounds from the past. But that is (thankfully) not the focus...

The programmers, then can recreate some of the classic sounds... it's funny, what we remember, and I'm so certain you can recreate most sounds, and they will sound better - the hardware is simply that much improved.

For example, the single Feedback loop in the 88 Algorithms is strictly to bring compatibility with legacy sounds that used that setup exclusively. The FM-X engine has that Spectral Form and Skirt system that allows a single Operator to create an array of complex source Waveforms alone... which is like having the ability to feedback any Operator.

It's the ability to move and change the FM-X sound that is exciting.
Hopefully, in the future we can get some of those early programmers to talk about this subject... I'm sure in each generation there were things that survived and other that did not. When you get into the details you start to realize that each generation is very different.

Jason wrote:
It's hard to pick out for me since my computer setup is low-def and there's a lot stacked on top at the end there.

Yeah, you probably need a higher fi system to really pull out the effect. I've been playing with the GS1 EP and EP Tines performances on the Montage and to my ear they don't really capture it (could be just me 🙂 ). I get a better approximation with the rotating speaker effect on many of the other FM Pianos like you mention.

I went back to the patent that covered at least some of the GS FM implementation (US 4249447) and it gives quite a bit of detail on the "cross-modulation" configuration. Specifically, Figure 19 shows the modulator feedback configuration that matches the description in my previous post. Subsequent figures show the spectral shapes that are achieved, but it's not clear what they mean subjectively without implementing them and listening. The best we have is:

'It has also been confirmed by experiments conducted on the test device that the above described effect is remarkable when the range of variation of β1 and β2 is set between 0 and about 1.5." -- The patent describes the effect through a couple of previous paragraphs that are difficult to envision.

It sounds tempting!!! Then again without additional information, the Montage may recreate this "remarkable effect" through some other means. Oh to pick the brain of Tomisawa-san would be a delight!!

Another interesting tidbit to glean from the patent is that averaging had to be used in the feedback to prevent a numerical anomaly called "hunting" (small oscillations). Useful to know for any of us coding our own Algorithms 🙂

Do the liner notes say they're using a GS1? I'd be more inclined to start with an organ sound than an EP for the ending. That's the general effect they're after and delivered by a tonewheel organ + rotating speaker. Only reason why I started with the EP was because it was already loaded for something else I was working with.

"range of variation ... between 0 and about 1.5". I'm not sure if this means vary both between 0 and 1.5 - so say you have 0.5 increments:

beta1 = 0, beta2 = 0
beta1 = 0, beta2 = 0.5
beta1 = 0, beta2 = 1.0
beta1 = 0, beta2 = 1.5
beta1 = 0.5, beta 2 = (all in above list)
beta1 = 1.0, beta 2 = (all in previous list)
beta1 = 1.5, beta 2 = (all in previous list).

... of if the statement means the difference between the two (delta) should be between 0 and 1.5.

... previous list, but also:

beta1 = 2, beta2 = 3.5 (as an example) .. and so forth

At any rate - if either beta value is 0 - then this simplifies the resulting effective modulator/carrier/feedback relationship.

There are two things here. One is simulating a GS1 structure. I think you can do this with either Montage or, more directly, with something like FM8. Some nuance DAC bit depth/etc. may not be covered - but at least the general feedback structure you can experiment with directly in software that supports that kind of setup.

Montage can approximate the GS1 I believe fairly well as far I can tell from the presets featured that I've seen thus far.

Then there's this organ type sound at the end of the tune - not sure if it's GS or not. Even if it is GS - the effect of using a real rotary speaker is at least half of this sound - since standing in front of a fan even your own voice sounds cool chopped by a rotating mass. So that's something if you want the real deal can invest in a leslie of some type. There are effect pedals - and the new Montage rotary speaker emulation is better than previous. But if you really love this sound - then there is an option to go for broke.

Yamaha has their own non-drum-rotating version with little speakers spinning around. An interesting piece of gear. Most of the online demos of it are guitarists who I think could have gotten a chorus pedal and sounded the same. I'm not the biggest fan of guitar+rotating speaker.

Jason wrote:

Do the liner notes say they're using a GS1?

Indirectly. Here is Russell's quote regarding "Revelation":
"The synth used on that recording was a Yamaha GS1. It was an early digital synth predating the DX7. It did indeed have a very distinctive sound, with polyphonic touch sensitivity and a cool built- in chorus that had almost a leslie-like effect."

He also used it on "Talk to your Daughter":
"The keyboard sound on "Talk to Your Daughter" was the Yamaha GS-1. It was an early digital synth that actually preceded the DX-7. The patch I used was an interesting blend of e-pno and organ all wrapped in one with this very cool aftertouch. I think it would be difficult to get that sound without the instrument."

You can see that he is referring to that Ensemble effect that Bad Mister described in detail. The phasing created by the ensemble detuning sounds like a leslie to me too.

I'll probably experiment more with operator/part detuning, rotary effect, and maybe there's a good chorus setting that would work.

But I agree, I don't think that the cross modulation Algorithm was especially key to the EP sound. But it is an interesting feature (they thought enough of it to Patent).

Talk To Your Daughter (studio cut) is easier to pick out - and represents something closer to that familiar FM sound.

https://youtu.be/vM0CWfVZczY

Liner notes do say GS-1

https://www.discogs.com/Robben-Ford-Talk-To-Your-Daughter/release/6323535

The site is easy to search liner notes/instrumentation. Chick Corea shows up with a GS-1:

https://www.discogs.com/Chick-Corea-Touchstone/release/1686740
https://www.discogs.com/Chick-Corea-Again-And-Again-The-Joburg-Sessions/release/4036023
https://www.discogs.com/Eddie-Gomez-Featuring-Chick-Corea-And-Steve-Gadd-Gomez/release/2975199

Herbie is in the mix too, even though now he's a "K" guy:

https://www.discogs.com/Herbie-Hancock-Future-Shock/release/115376

Dave Grusin ...

https://www.discogs.com/Dave-Grusin-Night-Lines/release/10329646

Only Yellowjackets album with GS-1 mentioned I could find:

https://www.discogs.com/The-Yellowjackets-Mirage-Trois-Club-Nocturne/release/5988124

...liner notes are not definitive. Just an interesting resource using discogs.