A Good Direction for Digital Guitar?

Written by

Is this what your studio will look like in 2023?

The recent post on Auto-Tune for Guitar generated much interesting discussion, plus a few good Auto-Tune jokes. Thanks, guys! :beer:

The comments from smgear particularly impressed me, because he managed to assemble a wish list for a future digital guitar with more detail and clarity than I could have managed.

It’s always a good idea for smart, articulate musicians to sound off about the musical tools they desire. But that’s especially true now, as manufacturers grapple with technological change with varying degrees of success. People really are listening!

Anyway, smgear wrote:

Basically, the problem is that the majority of the manufacturers seem locked into old paradigms. The bulk of their new designs are intended to mimic something else (Line 6), combine/integrate technologies (Roland) or make the ‘art’ less rigorous (Antares). Those are all fine pursuits and have resulted in some great tech, but they’re basically locking everyone into a pre-1980 palette of sounds, functions, and expression. Quite honestly, after my momentary enthusiasm has passed, I just pick up one of my no-name beater acoustics or electrics and dig in because every day I find a new sound, attack, approach, or whatever that allows me to express something in a new way. It’s true that I could do that playing through those tools, but I don’t need them and they are specifically designed to conform my sound to specific realms rather than to let me explore new spaces.

So with regards to this batch of hex-based modeling tech, my general requirements are fairly simple. I want clean and discrete signals from each string (check), I want a serious multi-core processor AND a couple programmable on-board control knobs/switches (semi-check), and I want an easily accessible and intuitive interface that gives me full control over how each particular string or any group of strings is processed/routed (no check).

For example, wouldn’t it be sweet if I could program the lower two strings to have a drone-like building-sustain, set the width of the effect, and have percussion triggered when I pluck them. Or maybe do that on the second and third strings and leave the bottom just for a percussive trigger and the top 3 strings are normal. Or maybe those top 3 strings are set to ‘12-string’ timbre. Or whatever.

I can sort-of do all of that with the current batch of tech, but it would require components from several companies and about a week to configure each setting – and it would almost certainly fail on stage due to a cable or midi failure somewhere in the line when I trigger it.

Given the processing and touch screen capabilities we now have, I don’t see any reason (other than over-reaching patents) why one of these manufacturers wouldn’t just create an app that will allow users to have total control of the processing. Think of it in stages:

1. select a string(s)
2. select a tuning
3. select a timbre model (instrument/synth)
4. select a pickup model (pickup or mic)
5. add effect 1 (ambient, traditional pedal-based, distortion, etc)
6. add effect 2 (…)
7. add percussive trigger (select sample – static or variable to pitch)
(repeat for other string(s) )
8. add midi trigger
9. assign desired effects settings to controls
10. save patch

The pro-version would also let you dig in and design your own effects – simple enough to do if you have the processing power and an easy way to split/chain the signal and base effect types in the interface.

That would take some serious programming and processing to get that sorted, but seriously, why would any company be investing R&D in anything less? We don’t need more of the same. Give us something new to play with already!! And remember that our goal is to make music, not spend our practicing time screwing around with cables, interfaces, and indecipherable control displays.

Mega-thanks to smgear for sharing all his great ideas, and expressing them so cogently — and for free, no less! (Join the club, man! :satansmoking: )

So what do you guys think? There’s no guarantee that some smart manufacturer won’t swoop in and steal all our ideas — but at the very least, putting them out there improves our odds of getting the stuff we want!

Comments

20 responses to “A Good Direction for Digital Guitar?”

February 13, 2013

David

I’m all for looking forward. I had been trying to find as many sounds you could squeeze out of a guitar, that often didn’t sound like a guitar, since I started playing some 42 years ago.

The problem is today most guitarist want pre 1980s tones, and seem to reject anything that doesn’t look like it’s from the 70s at the latest. They reuse the same hand full of guitar tones over and over. it has to sound like a Strat, or a Les Paul, or a Tele. No one seems to want an identity anymore. Of course I’m generalizing, but this seems be the mainstream.

It seems to be fashion more than anything. As an example, you have the fake vintage guitars with the distressed finishes and rusty parts. For posing in front of the mirror? We have this new fad of having the metal pickup covers on humbuckers (players removed those for a reason! The pickups generally sound better without them). And then all the reissues of guitars and pedals from the 60s and 70s. Just a glance at some of the bands that played at the Grammys and I thought it was 1970! (too bad they didn’t sound like 1970). When companies like Gibson try to make something new, we get the Firebird X. What an ugly and misguided guitar!

Bassist have always been the ones into state of the art, cutting edge technology, but even here, you see a resurgence in vintage tones and gear; tube amps, flat wound strings, Jazz and P basses.

Now this other stuff, the auto tune and robot guitars; I don’t know what to make of that. It’s a solution looking for a problem. The auto tune plugin has been abused to the point that no one needs to have any real talent. As long as they look good, they can fix it in the studio. And even for singers that can sing, it’s sucking the soul out of the music. The human ear likes to hear variations in pitch. That’s why we use vibrato. Now I listen to recordings and the vocals often sound too stiff, because every slightly off note was ironed over until it is too perfect. Musicians are not doing this; the record labels are. They think if you don’t have the loudest song on the radio, that sounds like every other mix, it wont sell. And meanwhile record sales are at an all time low. Hmmmm. Wonder why?

Now we are going to have a whole breed of guitarist that both don’t know how to tune their guitars, and can’t even be bothered looking for the right notes! Plus they use out of the box tones on their Line 6 Pods, etc.

That’s my take on the whole thing anyway. Hopefully I’m not over reacting. 🙂

Reply
February 13, 2013

smgear

Those are good points David, but I think the issue really boils down to the related questions of marketing vs. market and looking ahead or looking behind. Eighty percent of the industry marketing campaigns are telling players to look backwards so it’s not surprising that many of them do. The high dollar relics are selling well, but I think most of them are being bought my middle-aged guys trying to reclaim their youth. The indie market has been trending towards dano/supro/magna/valco styles equally out of nostalgia and a desire to employ one of the few available ‘unconventional’ sounds. So you end up in what I call a marketing tautology – we made the vintage sounds, so we tell people to chase vintage sounds so they want vintage sounds, so we make vintage sounds, so we tell them to buy vintage sounds because that’s what we make….. You can’t assess people’s tastes when they have no other options.

The market isn’t likely to change much from within due the current advertising/power holders – because Fender, Gibson, Ibanez, Line 6, Harmon, etc stand to lose a lot if they encourage people to crave new sonic tastes. I’m in my early 30’s and most of my gigs are with players in their early 20’s. We want something different and we’re tired of being forced to choose from the existing palette. It’s true that a lot of cool innovations never caught on for various reasons, but at some point, a company has to decide to invest in the ‘logical next step’ and put their eggs in that basket. I think it’s very much a generational issue and it might well take 10 years or more to really catch on, but you have to take those risks in order to move the market forward.

So what I was proposing above is basically just a product that converges the existing tech in a common-sense and accessible way. It could be done in a closed system that only models vintage characteristics, but it could also be done in an open environment that would give players the ability to push their music in entirely new directions. The beauty of analog systems was that anyone with a soldering iron could experiment with new pedal designs and effects. The problem in the digital realm is that it requires a significant investment of time and tools to learn and experiment. So the experimenting has largely been done by a handful of engineers in a few companies whose mission has been to recreate the past with the current tech. I just want one of those companies to create a platform and interface that will allow the rest of us to experiment and use the technology in the ways that we want. Some just want to be able to mix and match more easily. Others want to be able to create new effects and textures. A good platform should satisfy both of those groups. It’s certainly not for everyone. But I strongly believe that there is a market demand for ‘accessible and reliable diversity’, and if some company provides a tool to satisfy that need, then a lot of the ‘vintage’ guys will jump on board when they see the possibilities it offers.

But of course I could be wrong…

Reply
February 13, 2013

mwseniff

Actually something like Smgear talks about may already exist in the form of Keith McMillen’s Stringport for hex-pickup equipped guitars (violins). The software is Mac only it allows the use of any VST audio plugins from other mfrs. if you watch the demos the abilities of the software are very impressive. They have talked about a Windows version but it is very slow in coming

Reply
1. February 13, 2013
  
  joe
  
  Oh crap, I should of thought about StringPort — especially since Keith is right here in the SF Bay Area, and I did some work for him for SoftStep (which I still love and use almost daily).
  
  More info here: https://www.keithmcmillen.com/stringport/overview
  
  Now, this isn’t exactly what smgear is talking about either. The software suite is exceedingly complex, and soon the product need a USB update. But absolutely — Keith has been at the forefront of expressive digital music making for ages. Nice guy, too. 🙂
  
  Reply
  1. February 15, 2013
    
    mwseniff
    
    Have you ever played the Stringport system Joe?
    
    Reply
    1. February 15, 2013
      
      joe
      
      Only briefly, so I can’t comment much. If memory serves, the software approach is similar to the one in SoftStep: numerous small apps created in Max/MSP.
      
      Maybe it’s just me, but the multi-app method overwhelms my Mac-coddled, math-impaired brain. For digital gigs, I’m less concerned with having a large, continually available collection of realtime controllers than with being able to address one or two key parameters within a sound, but having the ability to address different preassigned parameters as I switch from sound to sound. So I tend to use MainStage as a do-it-all app for live performance.
      
      Reply
February 14, 2013

Oinkus

Didn’t Gibson make a Les Paul with a hex pup set a few years ago that came with a connected box with software and controls that allowed you to change each string to most anything ? HD 6x-Pro Guitar System and got to love that the box is called bob. It looks like it was a lot of work to go as far as smgear wants but might be there.

Reply
February 14, 2013

smgear

Well, I’m going to keep feeding this thread in the hopes of getting some more detailed input and response. Responses here and in the other thread do list some ~similar options, but each has some limitations to my ‘vision’.

For example, Max/MSP, StringPort, and Synthmaker are all really excellent packages for designing custom effects and hosting them. But in each case, you’re still burdened with layers of excessive coding or wrapping and routing everything through hosted software. Even if you can host them on a small/micro pc, you’ve got to run them through a bulky OS and the versatile, but processor-heavy VST protocols (or the MAX engine). So you end up with a bunch of cables, converters, a laptop, or at best a tiny pc with dedicated display plus the software costs. That would be fine for the studio, but it doesn’t give you easy and reliable live options.

On the hardware side, Raspberry Pi and Arduino seem to be the best diy candidates for hosting the processing, but both of those have their own limitations. For example, if you wanted to connect RPi to a hex pickup, you would probably have to find a 6X2 DAC and connect it via USB or have one of the manufacturers develop a protocol and have a digital i/o from their board that you could connect to – at which point you’ve got two boards to squeeze into the guitar if on-board is your goal. Arduino is nice, but from what I recall, the sampling rates and processing ability were too small to do high quality audio processing with.

There are a lot of knowledge gaps in what I’m about to type and maybe it all already exists or the existing protocols could be adapted. But if so, I’d appreciate someone clarifying the ‘industry’ for me.

So the first thing I want to see is a standard, efficient, and open coding language for storing effect/modeling file blocks. That could involve a set of standard library functions but each block could also include it’s own libraries for highly customized effects. Programs like MAX would ideally be able to compile programs and save them as blocks in this open format. These could be classified or tagged as effects, instrument models, pickup models, whatever.

Host apps should be developed that will run these programs on various ARM/Linux/whatever devices

A control app should be able to access the host app and allow users to arrange/select/save programs (the part I talked about in the original post)

If all that is feasible and a set of protocols has been developed, then I would like to see at least one of the hex pickup/processing manufacturers embrace it by either hosting the protocol on their processing board or providing a standardized digital i/o that would allow users to configure and run this on a separate board. The beauty of this architecture is that small boards could be installed in the guitar for running small amounts of onboard processing or larger boards or pc-based devices could be configured externally (essentially as stomp boxes – and with additional i/o) to handle more rigorous processing. It would be a scalable system which would be adaptable to many scenarios and allow you to develop a library of effect blocks that you could easily employ in whatever scenario you wish.

The ultimate direction of this could also be a corollary modular hardware system where additional processors could be chained to provide a scalable processing solution for any individual application. I’d prefer if the system was totally open, but I’m not against manufacturers selling code blocks or producing source/processing modules that can be placed on the aforementioned scalable boards.

And then if someone would sort out midi over bluetooth or a similar NFC tech, we’d easily be able to sync up control devices/ipad apps/foot switches/etc without the need for any other cables.

For routing to external processing, you can always use a digital link cable, but one of these companies needs to get off their butts and produce a locking connector system with tension relief to make that secure – incredibly easy to do, but no one has done it yet.

So on the software side I want:
1. a straightforward, light, and open coding language for generating effect blocks
2. host applications for interfacing with a range of OS’s (mobile or legacy)
3. control applications that will run on mobile devices or legacy OS’s for programing effects chaines.

On the hardware side I want:

1. A standardized and open protocol for digital I/O
2. A modular board system running one of the light mobile OS’s with the following possible modules which could be added to suit various applications:
– Chainable/expandable processors
– digital I/O
– Analog DAC
– Proprietary source/processing modules
– Some form of NFC for communicating with controllers
– SD memory for storing blocks and samples

I know that much of that exists and I’ve left out some key stuff, but please enlighten me. 🙂

Oh, and Oinkus, Thanks for mentioning Gibson and Bob. Gibson, Line 6 and others have done much of this, but within closed systems that ether intentionally limited the possibilities or the control software was not intuitive and nearly impossible to manage. I want a system that ‘works’, is open, and lets users meaningfully interact at whatever level of detail/complexity that they wish.

Reply
1. February 15, 2013
  
  mwseniff
  
  I would love to see some reasonable standards for music tech but I emphasize reasonable. For example I have always had a problem with the standard for ext power on FX pedals. The current standard has the center negative and the barrel (outer) positive. If you use a daisy chain cable and one of the plugs touches the enclosure of a pedal you get a short circuit. Most supplies can take a short without burning up but they clamp down causing a loss of power to all the other pedals causing a loud pop or even making a pedal reset itself to off. I can’t believe there was a non-sleeping electronic engineer in the room when they made that decision. This is one of several bad standards I have seen over the many years I have serviced electronics. My point is standards are great if the right people make the standards up.
  
  I think you also need something other than consumer type OSes to run your system. An OS designed for the purpose of audio processing can be streamlined to better do it’s job without unneeded components. I used to do service on chemical analysis instrumentation and in the 80’s&90’s there were many instrument companies that had their own OS. The hardware speed back then was such that you couldn’t afford to waste processor cycles on anything when doing data acquisition. Nicolet had a very good OS that ran on their own computers it was finally replaced in the 90’s when Windows OS and a separate smart box to do the actual data acq. Windows acted as a control and data processing only the pcs acted as a control interface and data processing unit. I think a smart box with it’s own streamlined OS, ADCs, DSPs and DAC to carry out the actual work but use a different device for the human interface like a pc,iPad,Mac, android or whatever you want . This would allow the max power at minimum cost it also would reduce the learning curve because the user has a familiar OS for an interface.
  
  I would love to see this happen but if we want turnkey systems then someone is going to expend money and time developing the smart box and interface apps for the various OSes. I could deal with problems easily but many guitarists are technophobes that will never read a manual to save their souls. Otherwise we are creating a niche product that will never get anybodies investment paid back. I personally would love to have a system like this to play guitar thru but I can think of only a couple of musicians in my area that would not run screaming into the darkness after 5 mins. People like to plug and play not spend hours learning and configuring the system. I have personally run into snags on stage with overly complicated gear I persevered but most guitarists would shove it in the closet after one problem at a gig and get out the pedal board to gig with.
  
  Reply
  1. February 15, 2013
    
    joe
    
    Interesting idea about the dedicated operating systems. And it makes me think of the mini-trend I observed at NAMM of “smart” pedalboard switching systems — digital control of analog boxes.
    
    Reply
February 15, 2013

Digital Larry

But I don’t want all that! I just want to SING!

Reply
1. February 15, 2013
  
  joe
  
  Hee hee. https://www.youtube.com/watch?v=g3YiPC91QUk
  
  Reply
February 15, 2013

Digital Larry

On a serious, insomnia-fueled note, I think we’ve seen (can’t recall the product name) that the hex pickup is no longer needed. Someone’s developed a DAW plugin that just takes a monophonic guitar input and combs out all the notes you’re playing and seems to do it really fast. While I don’t know for sure, I think this may be accomplished quickly in DSP using the “cepstrum” (spectrum of a spectrum) method. So if you use a dedicated DSP you can zoom in on the CPU cycles you need without OS overhead of Linux (Raspberry Pi) or an underpowered processor (Arduino – although they can apparently do FFTs at 44.1 kHz, so…)

For the rest of it (sound/FX generation) you will probably find that the sound designers (those who bridge objective intent with the available tools) will be a rarified bunch, and that the tools available to the average consumer will need to be greatly simplified. What I would hope is that we’ll be able to go beyond emulations of known sonic quantities and celebrity presets, while avoiding potential dead-ends of useless algorithmic combinations. Now some of you are probably saying “hey give the the dead ends!” so it should be possible to get those if you need them.

I’m going to go back to my (boring by this point in time) personal POV. When I goof around with synth patches playing a keyboard, I don’t feel any sense of loss in the process. When I have played around with MIDI guitar, that is not true. There’s the problem of triggering latency, and in some future case, the system trying to analyze whatever is really going on in that first few milliseconds and then boiling that down to a series of control messages that are then sent to a waveform generation engine for reconstruction. It sounds exhilarating from a design perspective and excruciating from a playing perspective.

But I don’t want to dismiss this entirely out of respect for smgear’s zeal. So sir, what do you expect to come OUT of the thing, once all is said and done? Is it:

a) An easier and universal way to quickly access all sound processing strategies past and present?
b) A completely flexible system for doing “anything”?
c) Faster, Mae! Play faster! Faster! (second pop culture reference in as many posts)

Reply
February 15, 2013

smgear

🙂 Yes. I want all those things… tied together with a bow on top. You raised some good points that I need to do some research on. Some of the new polyphonic tech and algorithms are indeed sweet, and perhaps sufficient…. but they don’t let me configure my lower two strings as a bass…….

Anyways, it’s good to vent and to get feedback from the likes of ‘yous’ who know more about the technical side of things.

Reply
1. February 15, 2013
  
  joe
  
  Another thought: While Fishman’s TriplePlay is a MIDI-only hex product, it uses a difference paradigm than the one smgear suggests in the comment quoted at the top of this post. He talks about a workflow in which you proceed string by string, which is exceedingly logical. But in some ways I prefer the TriplePlay approach: You have up to four virtual instrument channels. You can play them simultaneously, or assign specific ones to specific regions of the neck by dragging boundary markers on a virtual fretboard.
  
  I need another cup of coffee before I can contemplate how this would work with audio-based patches, including ones with radical pitch-shifts. But for assigning basic MIDI splits on the fly, the TriplePlay interface is really fast and intuitive.
  
  Reply
February 15, 2013

smgear

Good stuff! Thanks for the comments! I admittedly confused the issue(s) by going into the hardware side. There are many ways of doing this, each with particular advantages/disadvantages. To clarify what I was talking about with the different apps, I was just describing how the hardware/os/open effect file format could connect on the backend. In actual practice, you wouldn’t have to think about that at all and would just be using the main control interface for chaining, switching, and adjusting settings.

In many ways, the existing ‘tech’ is sufficient to get very creative, but the limitation is usually the interface/control, and occasionally the processing capacity. Designing an interface that is intuitive and also permits detailed configurations is certainly a challenge, but would only take a matter of days to sort out if they kicked the engineers out of the room that week and brought in a nice sample of gigging musicians and some actual designers.

Joe raised a good point about the fishman. While it’s not an approach that I would have considered for my needs, I can certainly see the appeal for others. But if the hardware can support 4 processing chains, then that is powerful enough to handle the basic scenarios that I envisioned – if only the interface allowed me to configure it that way – and I could do it wirelessly from my ipad during the gig. I don’t have a strong opinion either way over the relative merits of a midi based or analog piezo based hex systems. This generation of either approach seems to work really well in most scenarios. Ideally a hardware system would accept either input.

Like I said at the beginning, many of the products out there have the capacity to do what I want, but they have purposefully designed them to offer a fairly traditional tone pallet and are closed systems. I’m assuming the main technical rationale is system stability and preventing users from overloading the processor with complex combinations. I get that. But at the same time, there are dozens of companies with extremely powerful and compact hardware that could probably be re-purposed for an open format like I’ve been describing with relatively minimal effort… I think. Does anyone know if line6, digitech, boss etc, are all using the same core hardware and coding languages or if they are all proprietary?

mwsennif is certainly right about keeping this simple. I hadn’t intended for end users to have to deal with the hardware side at all. I just brought that up as an example of a product someone could develop that would open up digital effects for the rest of us. Mid-level guitar manufacturers could throw it in some guitars, some of the boutique pedal guys could build interfaces around it with preloaded effects, it could be synced to the digital – analog pedal switchers, or be available for people like us to tinker with. But in any case, the shipped product would be a fully functioning system that you just connect to your pickup or output and configure your desired settings with the aforementioned app. If you did have some gear running that hardware and had a program like Max/MSP which would (ideally) be able to compile in the format, then you could design your own effects if you wanted, drop them on the SD card or install them through the control app and rock away.

I agree that a dedicated OS would be ideal for the hardware. I’d guess that one already exists for use in many of the current digital multi-effects pedals. I have just never seen any actual data on how those are designed and whether or not there is already a standard in use between manufacturers. I assume that they would claim there isn’t, but in reality they’re using the same chips and language with minor customization. Is anyone authorized to comment on that?

so to hopefully simplify my ramblings. I think it would ideal if there were:

1. More intuitive and customizable controls for the current hardware in order to break free from some of the traditional formulas (extremely easy to do)

2. A standard and open effect/processing language/file format and corresponding development platform so that above-average Joes could create or customize their own effects/models (would be lovely, but unlikely to happen)

3. Either a new hardware platform that would support the above in a range of settings/usages, or some industry adoption of the open effect format so that we could run our new programs on or alongside the existing tech. (I think the former may be profitable for someone to develop, but the latter is extremely unlikely to happen, if it is even possible)

In reality, I know that there are a thousand valid objections to why those won’t or can’t happen. But I feel that the art and therefore the industry would be best served in the long run if they moved in that direction.

Reply
February 16, 2013

mngiza

Unfortunately, when the brilliantly-conceived “smgear” system is made available in the year 2016, I will play the same old Eric Clapton licks I learned in 1967 through it.

Reply
1. February 16, 2013
  
  smgear
  
  yeah,… but they’ll sound way cooler. 🙂
  
  Reply
February 22, 2013

bear

Tried to get the Paul Vo LEV-96 prototype into the discussion a couple days ago but my comment evaporated. But check it out:

https://createdigitalmusic.com/2013/02/guitar-innovation-then-and-now-paul-vo-reinvents-fretless-acoustic-guitars-videos/
https://paulvomusic.com/2012-forward-concept-for-the-future/

We’re not just talking processing or sound triggering from your normal note attack. We’re talking having your strings behave and play in new ways as the tech manipulates the fundamental and overtone series. Combine this with any of the digital guitar processing ideas or the Triple Play.

Really hope Vo can get this out in a more musician-affordable package than the Moog guitars he developed.

Reply
1. February 22, 2013
  
  joe
  
  Thanks for posting that CDM link, bear. I was going to mention it as well, but my attention span evaporated. 😉
  
  Reply