It's been tried, and at least some guitarists like the result.
For instance, fairly recently I stumbled across the schematic of a famous (older) guitar distortion pedal that used exactly this approach - the guitar signal went through a simple JFET-based soft limiter before reaching the later stages (diode-based distortion). COVID brain-fog has made me forget the name of the pedal, unfortunately.
More recently we have the example of the Keely Compressor, which became hugely popular, and ended up on a lot of guitarist's pedalboards, usually as the first pedal in the chain. It's a clean compressor that doesn't add any audible distortion, and the compression itself became famous for being audibly very "transparent".
The reason this particular compressor became popular, is mostly because it uses a very sophisticated audio chip made by THAT corporation. The chip is able to accurately measure the RMS value of the guitar signal at any given instant, and this allows it to create a very accurate and well-behaved control signal, which in turn is able to provide just the right amount of compression at every instant (more or less!).
Crude diode-and capacitor circuits used in most guitar compressor pedals can't do this. For instance, the big transient at the start of a guitar note will trick most compressors into compressing far too much - or failing to compress fast enough. Either way, you'll hear an audible flaw, either a big "sag" in the sound, or a loud "pop!" at the start of each note. The Keely compressor, thanks to the sophisticated THAT chip inside, can do much better.
On the tube side of things, the late Fred Nachbaur's website had a little tube guitar amp called "Lil Spunky" that he'd designed and built. The circuit included an optional compressor/limiter feature - basically it rectified the signal voltage across the speaker, and fed that back to the grid of the input valve to vary the gain. Nachbaur claimed it worked well for him - and from a few clips I heard of his playing, he seemed to prefer very clean tone. He liked the added compression as it gave him more sustain, more uniform loudness from note-to-note, and more apparent loudness.
So it seems there is definitely a market for a guitar compressor, as long as it doesn't introduce too many unpleasant artifacts as a side-effect. Not every guitarist will want one, of course.
There seem to be as many ways to play the guitar as there are guitarists. Some great guitarists play with a feather-light touch, others whale away with abandon. 😀
With the latter group, you can often hear the pitch of the note go up at the instant the pick hits the strings hard, raising the string tension, then drop back down as the note continues. This is most noticeable if the guitarist also uses light-gauge strings (tension is lower to start with, so the percentage pitch change due to the hard pick impacts are bigger.) This odd-sounding pitch modulation becomes part of the sound of that particular guitarist.
I most recently noticed this when listening to some of David Simpson's playing (if you've never heard of him, he has a You Tube channel, and does some videos for the Anderton's Music You Tube channel as well). Simpson is a very versatile guitarist, but in that particular piece, he was playing something bluesy with a lot of intensity.
Personally, I'm more in your camp. I like to play with a softer touch, not only because I think it creates better clean tone (and less artifacts like fret buzz), but also because it leaves more room for playing dynamics and expressiveness. Start at 20% and you have room to raise the playing intensity. Start at 100% and you have nowhere to go but down!
Thank you! I appreciate the comment very much.
I'm not at all sure I could design a commercially successful guitar amp, though. I don't think I'm a good judge of the sort of guitar tone many younger players seem to want. Extremely high levels of distortion are popular with a lot of newer players, tones that often sound very harsh and unpleasant to me, not to mention robbing the player of any expressiveness.
IMO, if every note is so heavily distorted that it sounds like a power tool grinding on a tin roof, there's little room for any kind of subtlety - and without subtlety, there's little chance of expressiveness.
The other thing that happens with very high levels of distortion is that you can't play chords because they turn into a wall of discordant noise; you are limited to playing only single notes or "power chords", which aren't chords at all, since they only contain two notes, the root and fifth.
That in turn means that there are no longer any chord tonalities - every chord sounds the same, as there are no longer even any major or minor chords, never mind all the subtler sounds (suspended chords, major sevenths, etc). For me, a long string of identical-sounding two-note "power chords" quickly becomes boring to the ear - there's no harmonic variation at all.
As always, music is very subjective, so my preferences may be quite different from your own.
-Gnobuddy
Thank you for pointing that out again. I agree completely with this opinion.
I bought various Line 6 and Digitech and Zoom boxes at intervals, starting in the late 1990s. IMO, every single one sounded bad, to varying degrees. The Line 6 stuff sounded worst, by far, and the Zoom stuff was least-bad to my ears, but this is faint praise, because overall (in my opinion), the sound quality generally varied from awful to mediocre. None could be described as good, never mind great.
As Shanx said, that changed dramatically in the last few years. To my ears, the Boss Katana amps were the first that broke the $500 barrier while still producing very good sound quality, at least for clean tone. That was in 2016, I think.
Then the Fender Mustang Micro came along (in 2021, I think) with pretty decent sound quality, and the Flamma Preamp with even better quality (but no onboard sound effects, only amp models).
This is almost exactly what I'm doing lately.
I run the output of my Flamma FS06 Preamp through a simple home-made attenuator (50k logarithmic pot-in-a-box) to bring the signal level down to the level expected from an electro-acoustic guitar (acoustic + pickup), then straight into the guitar input jack of my Katana 50, set to the "Acoustic" channel.
The Katana's Acoustic channel is designed to provide (reasonably) flat frequency response, and no additional audible harmonic distortion, so that you can play a plugged-in acoustic guitar through it. Boss probably used DSP to (somewhat) flatten the peaky frequency response of the 12" guitar speaker in the Katana.
Whatever those clever Roland engineers did, it works very well - check out the acoustic guitar sounds in this video (acoustic channel demo starts at 7:48 minutes):
To me, the combination of Flamma preamp and Boss Katana (set to the Acoustic channel) sounds really good. I like it much better than the "Crunch", "Lead", and "Brown" channels on the Katana.
The way I'm using it, the Katana 50 is basically acting as a (nearly) FRFR powered 50-watt loudspeaker, while the Flamma provides all the "tubey" sounds.
I say "nearly" because no amount of DSP magic can turn a 12" guitar speaker into a 1" silk dome tweeter; there is no way the Katana has extended flat treble response out to, say, 15 kHz, like a true FRFR (Flat Response Full Range) speaker. However, electric guitars don't need a response out to 15 kHz, or anywhere near it - my own experiments suggest that there is nothing good-sounding above roughly 5 kHz from an electric guitar.
More to the point, to my ears, the combination of Flamma + Katana produces excellent guitar sounds, with nothing lacking by way of high treble.
It's worth mentioning that a Katana 50 MkII currently goes for $270 USD at Sweetwater. I don't think you will find a powered 50-watt FRFR cabinet designed for guitar at anything close to this price - you'd have to spend far more. And the Katana, of course, is much more than just a powered FRFR (ish!) speaker - it's an entire guitar amp in its own right, and a very well designed and built one at that. It works very well with the Flamma Preamp (IMO), but also works very well on its own - it's been the best-selling guitar amplifier, worldwide, for over six years now.
A new Katana 50 + a Flamma Preamp together cost under $400 USD. The cheapest Tone Master that has more than a 12-watt power rating (i.e. the 100W Deluxe Reverb) costs about $1000 USD at Sweetwater. The Flamma + Katana gives you a lot more sonic options (7 amp models instead of one), and the considerably more limited Tone Master costs two and a half times as much - $600 USD more.
Earlier in this thread, both Shanx and I commented that the Flamma Preamp, set to its Fender Deluxe model, lacked the "glassy" high frequency response characteristic of some Fender Blackface amps. I recently discovered that those glassy cleans are are indeed available in the Flamma preamp - but only if I turn the volume pot on my "S-type" guitar to full! If I lower the volume even a smidge, those very high frequency "glassy" cleans disappear.
This particular issue is caused by capacitive loading of the guitar - either my guitar cable, or the input of the Flamma Preamp, has sufficient capacitance to "kill" the highest treble frequencies when the guitar volume control is rolled down (which raises the output impedance of the guitar, making it susceptible to this sort of loading effect).
The guitar itself doesn't seem to be at fault - it's a Sire S7 FM ( https://sire-usa.com/products/larry-carlton-s7-fm?variant=34034064654474 ). Larry Carlton was involved in the design, and the pickups have that sort of LA studio sound - very bright and clean - if you leave the tone pot all the way up. The tone control is very well designed, and more muted tones are readily available if that's what you want.
As an aside, this is the first electric guitar I've allowed myself to buy since 2012 or so. Bought online during the pandemic, I waited over ten months for it to arrive, and came very close to cancelling my order. Literally the day before I'd decided to call and cancel, I got an email from the music store saying the guitar was finally in stock, and did I still want to buy it? Well, you all know the answer to that question now!
One final comment about the use of the Flamma Preamp to provide my "tube amp" sounds: using a line-level device like this gives you a lot of additional versatility compared to my previous "classic" Fender tube amps. It lets you put additional effects pedals either before or after the "amp", just as though you had a guitar amp with an FX loop. I've chosen to put a guitar tuner and overdrive pedal between guitar and Flamma, while modulation, delay, and reverb pedals go after it.
The last pedal on the pedal board is a drum machine + looper. With the Katana set to its (nearly flat response) acoustic channel, it does quite a nice job of reproducing the drum sounds. No need to feed the drums into a separate PA or FRFR speaker - another big win!
So now I can have my cake and eat it too: distorted lead guitar, clean drums, clean looped rhythm guitar, clean or distorted looped bass guitar, etc, all coming through the same speaker. Works great at very quiet living-room volume, and works equally great at the levels used at our weekly music jams (which are not very loud by most e-guitarist standards.) Probably won't work well if you play very loud, though.
I should probably state very clearly at this point that I'm simply reporting on a setup that makes me very happy. It may not make you happy. And I should also point out once again, that I have no personal stake in Roland, Boss, Flamma, or any other musical product company.
-Gnobuddy
If you have 27 minutes to spare, I think you'll find watching this video well worth your time:
I know parts of the video are slow going, but I think it's worth watching the whole thing, because the lesson Hovak (the father) is teaching his little daughter (Ellen) is stunningly important. He's teaching her to first hear a sound in her head, then find and play that same sound on her bass guitar. And to never play a sound on the instrument simply because your fingers happened to find it.
How many guitarists ever do that? Most let their fingers do the work - they play whatever notes fall under their fingers, not what they hear in their heads. Once you've put in some time learning your scales, it is very easy to find and play easily reached notes from the scale fingering patterns you've memorized. And the result is the typical "noodling" or "widdling" guitar licks and solos we hear time after time, from the vast majority of guitar players. The sound of flying fingers playing memorized scale patterns and licks, rather than playing musical phrases that started out as sounds in the guitarist's imagination. The sound of guitar, not the sound of music!
I've been playing guitar for a long time now, and I really wish someone had taught me this lesson 35 years ago, when I was first starting. Instead, I've been clueless all this time.
For years I've been working at "hearing" the sound that's about to come out of my guitar before I actually play that note, and I've gotten pretty good at that. But that's not the same thing as hearing a musical phrase entirely in your head first, and then turning that same phrase into actual sounds from your guitar.
But while there's life there's hope, right? I'm not dead yet, so from now on, I'm going to remember Hovak and Ellen's lesson every time I pick up a guitar to practice.
Incidentally, I've watched some videos of David Gilmour (of Pink Floyd) during behind-the-scenes practice and sound checks and so on. Often you can see that Gilmour was actually singing (scatting) the notes with his mouth as he played them on his guitar. Sometimes you can actually hear him sing each note as he plays it.
The full significance of that entirely eluded me, until just a few days ago, when I watched the video of Hovak teaching Ellen how to improvise for the first time.
Now I know one of the many reasons why David Gilmour is almost without peer as a guitarist. He plays what he wants to hear, not what falls under his fingers.
-Gnobuddy
This what I simultaneously love and am infurated by in audio (generally): that we're all separated by a common language🙂. Yes, it's a high gain amp and if the knobs are all set to eleven, it's a fancy squarewave generator.
If you wind the gain knob down on this sort of circuit there's all sorts of options (but beware the noise floor) and I had a USB-Audio interface recently that overdrove in the same way. And the musicians in the room loved it. Took me a while to dial in the Marshall-clone to get the same tone live - too close to zero on the gain knob for reliable dialing in so a "box" may be on the way.
But I agree broadly, even if we're using language slightly differently.
For a wander down the wrong stack: a serious and tounge-in-cheek review of some very modern but very old amplifiers
You need a "treble bleed" on your volume knob. I've settled for (IRRC😵) a series RC pair between input and wiper. Took a little mucking about to get values that worked. For me it's the first 1/2 turn that I focus on as this is where it spends most of it's life.
Agreed.
Yes, because after well over 30 years of hardware and software development, the "new" tech has finally stopped sounding like crap, and at the same time, price points have dropped below the "old" tech.
From now on, Moore's Law will continue to drive up the quality of DSP guitar amp emulation, and drive down the costs.
Win-win.
A big part of the reason for this is Moore's Law, which has given us huge increases in computing power per dollar. Now you can run a reasonably complex DSP model on cheap hardware, which means you can do better than the crude DSP amp models that have been thrust down our throats for decades.
I recently bought a 32-bit, dual-core microprocessor that will happily run at up to 133 MHz clock speed - for only $5.76 CAD, retail price in quantities of one. That sort of computing power, at that sort of price, was unimaginable even a decade ago.
The chip in question? A Raspberry Pi Pico : https://www.digikey.ca/en/products/detail/raspberry-pi/SC0915/13624793
Some specs here: https://www.raspberrypi.com/documentation/microcontrollers/raspberry-pi-pico.html
As I wrote on this forum at least two years ago, a watershed moment in the history of electric-guitar amplification has arrived, without much fanfare, during the last few years. For the first time we have DSP-based guitar amps that sound good, and cost less than their tube competition. From now on Moore's Law will continue to drive up the quality of DSP guitar amp emulation, and drive down the costs.
At the same time, events like the COVID pandemic and Putin's invasion of Ukraine have put tremendous strain on the last remaining vacuum tube factories. Tube availability has dropped, tube costs have soared, and the future looks very uncertain.
Every tube amp manufacturer with even half a brain can see the writing on the wall now. If a manufacturer sticks to making only tube amps, they will be out of business soon. Simple as that.
For commercial production, I agree.
Yes, but you completely missed another reason, which was discussed earlier on this thread: this is a DIY forum, and trying to build a DIY DSP-based guitar amplifier is a much more difficult proposition than building an analogue one.
Not only are the skills required very different, but without an existing code-base of millions of lines of software, a hobbyist is unlikely to develop a good-sounding DSP-based guitar amplifier on any reasonable time scale.
By contrast, I can walk into my hobby room, dig out a couple of tubes from their cardboard boxes, heat up my soldering iron, fire up my oscilloscope and signal generator, and come up with a brand-new circuit that sounds at least halfway decent within a few hours (maybe a few days) of tinkering.
To summarize: for new commercial production, it really only makes sense to use DSP amp modelling, and class D power amplification...unless you're going after a niche market, people for whom "analogue" is religion, and "digital" is automatically bad. Those people will buy expensive analogue products like this:
But for DIY, it still makes far more sense to build an analogue guitar amp than to try and do it digitally.
And if you're going to build an analogue DIY guitar amp, it's a lot easier to get good sounds out if it if you actually use tubes. This is probably the only point on which you and I disagree. (Yes, I have read your book, many years ago.)
Certainly not every analogue solid-state guitar amp sounds bad - some sound quite good:
But it's a lot harder to DIY a solid state guitar amp that doesn't have various undesirable sound qualities, usually in the form of clean tones that are "too clean", and distorted tones that are harsh and abrupt.
The "new tech" has also been "bad tech" for a very long time. Cheap but bad-sounding DSP-based products have sold very well to beginner guitarists for a long time. For instance, Line 6 started selling their horrible-sounding DSP amps in the late 1990s, over 25 years ago now. But more experienced guitar players, with more developed ears, stayed away because they sounded bad.
That's finally starting to change. There are now some DSP based amps that sound great, at excellent price points. (But there are still overpriced and bad-sounding ones available, so Caveat Emptor still holds.)
I don't think you and I actually disagree on the most important points. Hopefully you don't actually believe that every digital product is better-sounding than every tube amp, though you sometimes sound as though you actually do believe that.
I remember a friend who bought an expensive and new-fangled digital phone answering machine in the early 1990s. She was horrified to find out that the digitally recorded audio from her new gizmo sounded much worse than the cheaper analogue micro-cassette answering machine it had replaced. Yes, there was no tape to wear out or break - but 8-bit digital audio at the low sample rate enforced by memory technology of the time sounded horrible!
-Gnobuddy
I kinda want to pin down the source of the problem before modding this particular guitar.
It may be that my Amazon-sourced generic guitar cable has more capacitance than it should, or that the Flamma preamp has more input capacitance than a typical guitar amp. If one of those is the case, I'd rather fix that problem than mod the guitar.
Funny thing, I have "LP style" guitars with humbuckers (and higher-value volume pots) that don't seem to suffer from the same issue. Maybe those pickups don't put out those high frequencies in the first place, regardless of volume pot setting.
-Gnobuddy
No, and yes. 😀
No, it doesn't have to be that way - that's how the guitar pickup and 1930's speaker technology evolved together. Nobody knew how to make flat-response speakers then anyway. And letting the speaker have a big peak in its frequency response also made it louder - which was important at a time when even a few watts of audio power was very expensive.
Yes, most of the new DSP-based solid-state guitar processors are designed to be fed into a "Flat Response, Full Range" loudspeaker, usually shortened to "FRFR speaker".
Because FRFR speakers tend to be far less efficient than a peaky old guitar speaker, these rigs tend to need a lot more audio power. Fortunately, huge amounts of class-D audio power are now quite cheap. So you can throw, say, 400 watts of audio at your FRFR speaker, and so manage to be as loud as 40 watts of tube power into an old-school guitar speaker. 😀
Nothing wrong with your idea. It's already in wide use. To put it as politely as I can, those other guitar people should open their eyes and look around the current guitar landscape a little bit. It's not all Blackface Fender Deluxes these days! 😀
Not at all, and it's quite common now. The AxeFX, Fractal, and Kemper processors are all usually designed to work with FRFR powered speakers (often just powered PA speakers).
Those three doodads are all insanely expensive, but similar technology has been dropping down to much more affordable price brackets in the last few years.
Here is a good example:
It's both possible and practical - and already in wide use. It does need a lot more watts to get equally loud, but thanks to class D power amps and switching power supplies, that's no longer an obstacle.
I think it's fair to say that there were good engineering reasons for using peaky-but-efficient speakers until just a few years ago. Of course conservative non-techie musicians might not know about the engineering at all, and just focus on tradition and history instead.
What's changed recently is the underlying technologies. Instead of a handful of vacuum tubes, you can now build an SS guitar amp/processor with literally tens of millions of transistors inside, using mathematical operations to create audio effects. Using DSP, it's very simple to create just about any kind of frequency response you want. You don't have to rely on the speaker creating it for you.
At the same time, developments in class D amps and switching power supplies - and powerful PA speaker systems - have made it possible to throw hundreds of watts into speakers, and to make speakers that might actually survive those levels of power.
Put that all together, and you have the possibility to create an off-the-shelf guitar rig based around an FRFR speaker, like the one "Digital John" (aka John Connearn) put together in that video.
Does Connearn's digital signal chain and FRFR speaker sound as good as a more traditional all-analogue guitar amp and effects pedals on the same budget? Evidently it does, at least to him. It may not, to you!
There is another (relatively new) category of guitar doohickey designed to feed into actual Hi-Fi speaker systems - attenuator systems that let you play your monstrous old tube amp, at living-room friendly SPL levels, though little Hi-Fi studio monitor speakers. This video will tell you more:
-Gnobuddy
I just got a Peterson tuner. Got a deal cause it's missing the battery door, something I can handle... I got it because it's a virtual strobe tuner. I built one in college for a good mark in a electronic instrumentation elective using LEDs and this reminds me of that, long gone.
Because it's a strobe, you can see the effect of the string bending sharp just after the pluck. I like it way better that my other tuners that manage to put up a letter when they figure out things - another algorithm again. This Peterson puts up a letter too, but I mostly look at the strobe pattern. It lets me know what's going on faster by looking at the pattern behavior; I can take it from there to get the string tuned ;')
Dave Brubeck (piano). Many examples. Extended passage in Balcony Rock he hums louder than the strings. And he doesn't mean to do it, he just does (I've been in the room with him). OMG: everything is on Youtube. (I've hoarded a badly damaged LP most of my life; guess that's trash now.)
Perhaps not anymore. This article is coming up on a year old. A few hours of time as well, to go from unboxed board sets to a working amp model? https://towardsdatascience.com/neur...-audio-raspberry-pi-guitar-pedal-bded4b6b7f31
When I was in school +20 years ago the majority of hobbies involving electronics was not analog electronics, certainly not anything guitar effects related and even less tube related. In my course I was the only one who even knew what vacuum tubes were. Folks were more into digital electronics and microcontrollers.
No. I don't believe that DSP is really that difficult for a future generation of DIYers: some basic microcontroller & digital knowledge and you're already quite there on setting up the digital part. In analog realm one has to know how to build decent CODEC interfaces and that's about it. Certainly easier than gathering a vast knowledge base of analog signal processing techniques and circuits. Kn analog most have skills limited to building a TS clone and swapping different diodes; maybe they can program much better?
No. I don't believe that DSP is really that difficult for a future generation of DIYers: some basic microcontroller & digital knowledge and you're already quite there on setting up the digital part. In analog realm one has to know how to build decent CODEC interfaces and that's about it. Certainly easier than gathering a vast knowledge base of analog signal processing techniques and circuits. Kn analog most have skills limited to building a TS clone and swapping different diodes; maybe they can program much better?
"In analog realm one has to know how to build decent CODEC interfaces and that's about it."
That could well become the bottleneck. You have to understand something about analogue concepts such as noise optimization, shielding, overload margins and sensible gain values to properly go from an analogue guitar signal to a digital representation. When I see how many analogue-oriented people have difficulties with those things, it is hard to imagine that someone who's only interested in microcontrollers will ever get that right. Then again, maybe they can team up with someone who does understand analogue and mixed-signal circuits.
That could well become the bottleneck. You have to understand something about analogue concepts such as noise optimization, shielding, overload margins and sensible gain values to properly go from an analogue guitar signal to a digital representation. When I see how many analogue-oriented people have difficulties with those things, it is hard to imagine that someone who's only interested in microcontrollers will ever get that right. Then again, maybe they can team up with someone who does understand analogue and mixed-signal circuits.
Still tubes rules. Those Andertons reminds me of UK TV-shop 😂 Id like one of each UK/US pedal.
Cheers!
Cheers!
Some Swedish guys simulates amps just by giving the simulator the diagram and components. Simulating all speakers inkluded takes a crazy amount of variables. Far far from easy. Otherwise they would have made good ones 25 years ago. They made good even great dsp effects then, but not good speakersimulation.
Cheers! And get those tubes cookin!
Cheers! And get those tubes cookin!
That presumes the simulator understands the components and understands rounding errors.
As of a few years ago (and I'm far too lazy to look up the thread, but it's here) MOSFET models in spice really only worked for a subset of operating conditions - those that were commercially important, to be sure.
The odds of them understanding the behaviour of speakers reflected back through transformers to a a pair of (not quite) matched pentodes is basically zero.