Anyone have reference to read, explaining how the internals of popular modelling amps work?
My GP-10 simulates many guitars but synth sound is just one, where as my GR-1 simulates many synth tones including guitar but no real guitar sound.
Both use the same pickup.
Regards.
My GP-10 simulates many guitars but synth sound is just one, where as my GR-1 simulates many synth tones including guitar but no real guitar sound.
Both use the same pickup.
Regards.
Do note as well that I'm making a distinction between digital "modelling" units and amps (that specifically emulate or "model" circuits and / or their characteristics), which appeared around the late 1990's and first units with digital processors to generate distortion or other effects, that appeared around the late 1980's, about a decade earlier.
With the latter I mean units like, say, Yamaha REX 50 or Digitech GSP 21, which make no claim whatsoever to even model anything. For example, REX 50 simply comes with a "distortion" effect feature (which I believe is fully digital), while GSP 21 in turn features several different distortion effect styles like "Metal tube", "Over-drive", "Rock tube" and "Heavy sustain", which as is are pretty generic descriptions a'la "Tube Screamer" and only imply a vague impression of doing something that some tube circuits do (i.e. distort).
Furthermore the distortion effects of several Digitech units were actually all-analog (with mere digital interface for control), and some actually even had genuine tubes to produce some of those distortion effects. (This didn't stop people from complaining how "digital" they sound).
So we have at least a whole decade worth of evolution before digital distortion effects even get upgraded to specifically "model" or "emulate" tube circuits. In terms of evolution consider that there is a similar time frame period between first Line 6 amps and the first Axe FX unit.
With the latter I mean units like, say, Yamaha REX 50 or Digitech GSP 21, which make no claim whatsoever to even model anything. For example, REX 50 simply comes with a "distortion" effect feature (which I believe is fully digital), while GSP 21 in turn features several different distortion effect styles like "Metal tube", "Over-drive", "Rock tube" and "Heavy sustain", which as is are pretty generic descriptions a'la "Tube Screamer" and only imply a vague impression of doing something that some tube circuits do (i.e. distort).
Furthermore the distortion effects of several Digitech units were actually all-analog (with mere digital interface for control), and some actually even had genuine tubes to produce some of those distortion effects. (This didn't stop people from complaining how "digital" they sound).
So we have at least a whole decade worth of evolution before digital distortion effects even get upgraded to specifically "model" or "emulate" tube circuits. In terms of evolution consider that there is a similar time frame period between first Line 6 amps and the first Axe FX unit.
A good introduction is, for instance, "A review of digital techniques for modeling vacuum-tube guitar amplifiers" by Jyri Pakarinen and David Yeh. You can also google patent search related documents.
None of the material will delve into very details of the "code" though, and often patents indicate that methods of optimization are far more important to patent than specific characteristics to emulate or model. IME it just portrays that there is nothing esoteric or magical in operation of tube circuits. The designers already know how they work in good detail and model that, the real question is how "efficiently" in terms of computing time, memory, and et cetera.
So are great sounding acoustic spaces, silent speaker cabs and virtually many other devices and things that introduce an effect that digital technology allows to pack into a format that you can basically carry in your pocket.
Yes, we are on somewhat at another subject. ...and then again not. Think how many things digital has made more accessible, easier and convenient and weight that against how good the "real deal" was really in practice.
Digital signal processing has, for instance, made home recording much easier and at the same time better quality. You weigh this against recording a deafeningly loud speaker cab in a 20 sqm room, or affording that real good studio at any given moment and with variety of cabs and microphones.
Or think how many amps and effects one can access with a push of few buttons, and compare that to having to buy or loan and load and carry all that stuff.
Yes, we can find very good and fascinating examples of any old technology but one was remember there were reasons why alternatives to it were developed in the first place.
Yes, we are on somewhat at another subject. ...and then again not. Think how many things digital has made more accessible, easier and convenient and weight that against how good the "real deal" was really in practice.
Digital signal processing has, for instance, made home recording much easier and at the same time better quality. You weigh this against recording a deafeningly loud speaker cab in a 20 sqm room, or affording that real good studio at any given moment and with variety of cabs and microphones.
Or think how many amps and effects one can access with a push of few buttons, and compare that to having to buy or loan and load and carry all that stuff.
Yes, we can find very good and fascinating examples of any old technology but one was remember there were reasons why alternatives to it were developed in the first place.
I was on another forum looking for the thread asking on modelling and could not find it. Then I saw the tab open for this page. Sigh.
I just did a quick search, other combinations may bring up other papers on digital modelling.
https://www.google.com/search?client=firefox-b-d&q=paper+modeling+algorithem+guitar+amplifier
It is not easy to hear if the bass guitar sound starts to distort.
Many bass guitarists even have problem in hearing bass notes or tuning.
But it is easier to hear onset of distortion with a regular 6 string guitar.
Regards.
Glad that you can do it. Still as I said, many bass players have problem in hearing bass notes or tuning.
My guitarist can write down chords just by hearing the song without checking with a guitar. His pitch memory is so good.
Many piano or keyboard players also can do it.
Never seen a bass player like that. Not offending anyone.
Regards.
My guitarist can write down chords just by hearing the song without checking with a guitar. His pitch memory is so good.
Many piano or keyboard players also can do it.
Never seen a bass player like that. Not offending anyone.
Regards.
Yes, of course there's the stereotype "failed guitarist" bass player and I'm sorry yours is one. But I hardly think I'm alone, or even unusual, in having good ears as well as big hands. Or is it just me and Paul McCartney that can hear the difference between E and F?
Not offended by the way. Amused by your unfortunate experiences/prejudice though.
Not offended by the way. Amused by your unfortunate experiences/prejudice though.
When you watch people play or tune, you should have realized what I said.
But this should be enough and we go to the forum topic.
Regards.
But this should be enough and we go to the forum topic.
Regards.
Yes, back on topic.
I can confirm that my basses don't ever clip the input of my line 6 modeller, and I'm fussy about unwanted distortion.
I can confirm that my basses don't ever clip the input of my line 6 modeller, and I'm fussy about unwanted distortion.
Just checked the service manual for Line 6 bass pod.
Max signal level after the first opamp is 5 volts peak to peak.
Unlike normal preamp IC, The gain is reduced, less than 1.
Probably it will never overload for a active pickup also.
Regards.
Max signal level after the first opamp is 5 volts peak to peak.
Unlike normal preamp IC, The gain is reduced, less than 1.
Probably it will never overload for a active pickup also.
Regards.
Yes. That is very example of what I wrote about earlier. A good 5Vpp (or 2.5Vp) headroom for inputs that Rod Elliot expects to be in 500mVp range and which at very hot pickups of a bass guitar might be in the 2Vp range.
Consider that against typical 1 - 2 Vp input headroom limits of tube amps, of which naturally no one complains about clipping during hyperbole of praising their magnificient clean tones or overall linearity of triodes.
Note also the "DAC stacking" from a sidechain with 16x gain for lower amplitude input signals. The device basically alternates between AINR and AINL inputs and selects the proper one based on input amplitude. So lower amplitude inputs are not even attenuated but amplified.
There would be no point of amplifying had the CODEC been operating close to its headroom limits constantly. So, instead we have a very clever system with careful attention to dynamics of the input: input signal's transients are attenuated so that they do not overdrive the input while low magnitude input signals are amplified so that SNR is increased and AD conversion exploited at full dynamics.
Bottom line: This design is very unlike those of tube amps and several SS analog amps that pay less attention to input signal's dynamics and broad amplitude range. It is practically totally opposite type of design. The headroom is also very high to amply support and tolerate even higher than common input signal ranges, at least its nearly five times as high than that of generic tube amps.
Can this thing be overdriven? Sure. But that will be very much a deliberate user error.
Consider that against typical 1 - 2 Vp input headroom limits of tube amps, of which naturally no one complains about clipping during hyperbole of praising their magnificient clean tones or overall linearity of triodes.
Note also the "DAC stacking" from a sidechain with 16x gain for lower amplitude input signals. The device basically alternates between AINR and AINL inputs and selects the proper one based on input amplitude. So lower amplitude inputs are not even attenuated but amplified.
There would be no point of amplifying had the CODEC been operating close to its headroom limits constantly. So, instead we have a very clever system with careful attention to dynamics of the input: input signal's transients are attenuated so that they do not overdrive the input while low magnitude input signals are amplified so that SNR is increased and AD conversion exploited at full dynamics.
Bottom line: This design is very unlike those of tube amps and several SS analog amps that pay less attention to input signal's dynamics and broad amplitude range. It is practically totally opposite type of design. The headroom is also very high to amply support and tolerate even higher than common input signal ranges, at least its nearly five times as high than that of generic tube amps.
Can this thing be overdriven? Sure. But that will be very much a deliberate user error.
Getting a little personal here, aren't you? Why are you feeling so hostile? Is it because, while you were were writing entire books about solid state amplifiers, you managed to missed all these audible subtleties I've pointed out in this thread? You never noticed how harsh a single op-amp input stage can sound with a guitar plugged straight into it, as I did?
I always welcome thoughtful constructive criticism, but statements like "You obviously never bothered to study..." are just idiotic ******* contests, and I want no part in them. Please refrain from these sorts of immature behaviours.
The primary and most important tool for any musician is their ears. Most people don't have musician's ears. It's a statistical rarity. Just as some small percentage of people are born with greater strength, speed, flexibility, balance, or other athletic potential, some small percentage are born with keener vision, and some small percentage with better hearing.
Not only that, research has shown that, over time, those who become good musicians literally rewire parts of their brains over the years, with more processing power being assigned to processing auditory input. So musicians typically start with better hearing than the average person, and then their hearing literally gets better with age for many years, until finally hearing damage from old age begins to rob them of their abilities.
How about engineers? Statistically, engineers have strengths in areas such as analytic thinking, mathematics, and logic. Those skills are also a statistical rarity.
Let's suppose 5% of humans have the skills to be a reasonably good engineer, and 5% of humans have the hearing to be a reasonably good musician. These being independent variables, the probability of a reasonably good engineer also being a reasonably good musician is very low. It's the product of the two probabilities listed above - 0.05 x 0.05, or 0.25 percent. That's one out of four hundred.
Let's revisit that: even with the generous assumption that one out of every twenty people has the natural gifts to be a good engineer or a good musician, only one of every four hundred engineers is likely to have a musician's hearing.
If you look at more highly gifted people, the numbers get worse. Let's say only 1% of humans can be a top-notch engineer, or a top-notch musician. The intersection of those two independent probabilities is only one part in ten thousand. Only one in ten-thousand top-notch engineers (or scientists) is going to also have top-notch musician's hearing.
This is why guitar amplifiers created by engineers working on their own often sound bad to real musicians with good ears.
The problem is typically made worse by arrogance, which unfortunately tends to go with the territory. There are a lot of arrogant engineers, mathematicians, and researchers. I discovered this the hard way, from personal experience.
Three decades ago I had worked hard at my studies and had good enough academic skills to be awarded a full scholarship to attend graduate school at one of the best technical universities in the USA. (In fact I was awarded full scholarships to six different US universities, but I had to pick one to attend. That's a whole other story - let's just say that I was young, and there was a woman whom I had feelings for in the picture, and leave it at that.)
At the school I chose to attend, every student was far above average intelligence - for instance, every student there would easily qualify to be a MENSA member. Every student there also had an outstanding academic history, easily placing them in the top 1% of students world-wide.
I was really excited about the prospect of going to this amazing school. Everyone there was very smart, everyone there was highly educated, everyone there was in a technical field (either engineering or pure science). Surely this would be the closest approach I would ever encounter to the ideal society imagined by Socrates, full of wisdom, knowledge, clever ideas, fruitful intellectual discussion, fellowship, and mature and thoughtful responses to events.
Boy, was I wrong! And I mean, horribly, completely, totally wrong!
Yes, everyone was smart. Yes, everyone was accomplished. Unfortunately, the vast majority - both students and faculty - had the emotional maturity of very small children. Arrogance, selfishness, and insecurity blended with conceit and a burning desire to always prove themselves correct, which, of course, required them to constantly try to prove that everyone else was wrong. It was a constant battleground of one-upmanship, between faculty and faculty, between faculty and students, between students and students.
As I attended research conferences and met and mingled with other scientists and engineers from other universities and other research institutions, I found out that these personality characteristics were endemic. Not universal, thankfully. But quite a large percentage of smart and accomplished people in highly technical fields turned out to have quite difficult and unpleasant personalities. (The personality of Sheldon in the TV show "The Big Bang Theory" is quite well written; the fictional Sheldon reminds me of a number of very real people I encountered in very similar settings.)
When you combine a typical engineers ears with a typical engineers arrogance, and use that as the basis for designing or studying a guitar amplifier, you have a disaster in the making. The engineer's ears probably aren't going to hear the bad sounds made by their amplifier. The engineer's arrogance won't let them even consider that the amplifier they created might not be perfect. And if any other person dares to criticize their precious creation, why, that person must be blasted with scorn and withered with contempt. How dare they question the masterwork of Me, Myself, and I, engineer/scientist supreme!
Leo Fender (and other sensible amplifier design engineers) had enough emotional maturity and social skills to avoid this trap. Leo knew how to string resistors and vacuum tubes together into an amplifier, copying circuits out of the back of tube manuals and pasting them together. But he knew that there would be musicians who could hear flaws in his amplifiers that he himself could not.
So Leo recruited expert local musicians to provide feedback during the design process. Leo was smart enough to understand that he couldn't hear what great musicians could hear. So he went out of his way to find talented local musicians, and recruit them to lend their hearing to the cause of designing better-sounding amplifiers. As a bonus, if that popular local musician liked the Fender amp, that would serve as advertising for Leo.
Let's get back to this thread.
You (teemuk) can argue all day about software details, what you've studied and what I haven't, but the fact remains that the DSP modelling amps I'm discussing - i.e. products from Line 6, Digitech, Yamaha, Fender, etc - have sounded bad to varying degrees for a long, long time. From the late 1990s till at least 2015. Line 6 takes my personal award for the worst-sounding products I've heard in this category, and they held that position for a number of years.
No guitarist with sensitive ears would touch these awful-sounding things with a ten-foot pole, though they sold very well to beginners excited to start making loud raucous noises rather than music. They also sold to people who were reeled in by advertising claims that these things sounded like actual tube amps. That claim was a complete lie, except for those with tin ears.
You say my comments here are "out of valid base". I'm certainly not the only one. Some of your comments in this thread (and elsewhere on diyAudio) are pretty far off base, too.
For instance, you keep talking about circuits clipping guitar signals into square waves, for example when you mentioned the cold-biased stage in Marshall 2204 amps earlier in this thread. This is completely wrong.
The cold clipper's nonlinearity partially rectifies the guitar signal, creating additional harmonics as well as intermodulation components at non-harmonic frequencies, but it does NOT clip the signal into a square wave. And when the guitar is played with sensitivity by an actual musician, the output of the cold-clipper stage is a lovely guitar "clean tone". It sounds nothing like a square wave.
For instance, listen to this clip starting at the point seven minutes and ten seconds in (7:10):
Note the settings of the amplifier at that point: it is set for clean tone, with the guitar plugged into the low-sensitivity input. As you can see on Rob Robinnette's page on the 2204 preamp ( https://robrobinette.com/images/Guitar/Marshall/JCM800_Preamp_Annotated_Schematic.gif ), when configured in this way, the guitar signal is fed straight into the "cold clipper" stage. And the result is a lovely "clean tone" - nothing even remotely like the square wave you say will be generated.
So are you, my solid-state-infatuated friend. For instance, quite wrong in your belief that everything is about "clipping".
If you want to hear what happens if you actually turn a guitar into a square-wave (rectangular wave, actually), do this experiment, which I bet you have never tried: Feed a guitar signal straight into a sufficiently sensitive Schmitt trigger circuit. Then listen to the utterly horrible noises that come out.
Try that experiment. After that, if you still think a Marshall amp - or any high-gain guitar amp - simply clips a guitar signal into a square wave, then there is no hope of useful communication between us, as your sense of hearing shares nothing with mine.
-Gnobuddy
Very true. Modern low-noise electronics probably makes it a worthwhile compromise to attenuate the guitar signal right at the input, throwing away a teensy bit of S/N ratio in exchange for a large reduction in harsh nasties from clipping op-amp stages.
I rather think the answer is "Yes, in some contemporary contexts." 🙂
The author of some music book I read made the interesting observation that people's tolerance for harsh or discordant sounds has been steadily increasing for decades (centuries, if you look back at the start of plainsong / Gregorian chants and so on.)
Back then only octaves were considered musically acceptable. Even what we call a "perfect fifth" was far too discordant to be allowed in religious song. It took centuries before the "shocking" timbre of a perfect fifth in harmony became acceptable.
Fast forward to the late 1960s, and you have the garbled, distorted racket created by, for instance, Jimi Hendrix. To my ears his vocals were usually an inarticulate, slightly-off-pitch garble, while his guitar made a wall of mostly nasty noise.
But I've never heard of a single person going to a Jimi Hendrix concert, and demanding their money back because Jimi's guitar didn't have an adequate signal-to-noise ratio...
So in that context (new and exciting rock music, deafening SPL, a charismatic drugged-out musician/hero on stage, and an audience probably high on marijuana and alcohol), trash being part of the guitar signal was evidently okay to the majority of people present.
It's taken me a very long time to like anything about Jimi's playing. The racket and the over-the-top showmanship turned me off too much, and the garbled and pitchy vocals didn't help. But eventually I started to realize that buried under all that showmanship (and all the drugs hampering his fine motor skills) there was an extremely talented musician. Unfortunately he remained pretty well hidden unless you listened very carefully to what was going on with his guitar.
Yeah. The electric guitar is very much an imperfect instrument, in many ways. But it did enough things right to become (by a vast margin) the most popular instrument in the world over the last several decades.
All things come to an end. Just as the mandolin craze that swept all of North America starting in the late 1800 eventually died, the guitar craze that has captured the world since at least the Beatles will end, too.
As a guitar player, there several other sources of noise that have frustrated and irritated me many times. For example, round-wound guitar strings tend to emit loud squeaks when you shift your left-hand position quickly (assuming a right-handed player). You can hear audible thumps and thwacks through the guitar amp every time the fingertips of your fretting hand push a string down onto the fretboard. Unless you have good muting skills, you will also hear unwanted pull-offs ringing out every time you lift a fingertip and free a formerly-fretted string, and sympathetic vibrations from unplayed strings (such as the low "E" ringing out when you strum a basic D-chord, even though your guitar pick never touches that 6th string).
I've been lucky enough to be near a skilled pianist while she played, and pianists make a lot of unwanted thumps and bumps and clatters and swooshes with their hands and fingertips (and feet on the piano pedals), too. You don't hear them when you stand a little further away, but if you amplified a piano to the same insane SPL as a typical electric guitar player, you'd probably hear a lot of this unmusical racket!
-Gnobuddy