You are trying to debunk a fact, not a "belief". It's like trying to debunk the existence of gravity, or claiming that the earth is flat, not spherical.
I have heard this problem (unwanted clipping in high-NFB SS guitar input stages) for decades, without fully understanding why it existed, because, like you, I was fooled by the standard methods of testing and analysis; a sine-wave input signal and a 'scope or other device to monitor the output of the amplifier stage.
Those methods suggest that an op-amp (or similar high-NFB amplification stage) will work just fine with a guitar signal as input. After all, such stages work very well in Hi-Fi applications - so why wouldn't they work with e-guitars?
Well, it took a long time, but I figured it out. It turns out that the real problem is that guitar signals inherently have a high dynamic range, high enough to necessarily cause clipping during the initial portion of the waveform. The problem doesn't appear when you consider steady-state sine wave test signals. This is probably what has fooled engineers for decades.
That, and insufficiently good ears. Not everyone can hear this problem, though quite evidently tens of thousands of musicians can; which is why the majority of good guitar players stayed away from solid-state analogue tube amps for decades. In the 1980s many guitarists moved to SS amps as they became fashionable; a decade later most moved back to "obsolete" tube amps, because their ears told them that the old unreliable noisy tube amps somehow sounded better and more musical than the new and shiny "improved" SS amps.
(Listen to the nasty lead guitar tone in Toto's "Love Isn't Always On Time", for example. Ugh! Steve Lukather's guitar technique is impeccable, as always - but his tone? Absolutely horrible!)
So: I could hear the nasty sounds made by every SS guitar amp I built, bought, or listened to in the store. But I couldn't explain why it sounded like that. And not everybody could hear it.
That posed a problem for me; I could hear the problem, but couldn't prove its existence objectively. A bit like saying "I saw the Loch Ness monster". People who could not hear it for themselves would, like you, think this was merely a deluded "belief" of mine, not a fact. And that is an entirely reasonable position to take; if someone told me they saw the Loch Ness monster, I wouldn't believe them, either.
But that has all changed now. I figured the details out. Now I know what I was hearing, and why I was hearing it. I've put all that information in quite objective form in this very thread. In the opening posts I showed quite clearly that if you put a typical guitar signal into a typical op-amp input stage (set for quite conservative gain), you will, indeed, produce brief and abrupt clipping events. If you have good ears, you will hear these as a harsh noise underneath the clean tone.
And if you can't hear it? Well, you are in good company. Many people can't hear it. But that doesn't mean the distortion doesn't exist. Many, many, many musicians can and do hear it - and my LTSpice simulation shows exactly what it is that they are hearing.
The problem exists. It's real. It's engineering fact, not some musician fantasy. Just because you can't hear it, doesn't mean it's not there. The stark objective reality of the LTSpice simulation proves otherwise.
There are similar situations in other areas of sensory perception. My wife is a trained visual artist who has exceptional visual abilities. She can see fine details of colour that I cannot. She can distinguish between similar shades of, say, yellow, that look identical to me. That doesn't mean she's lying, or that she just "believes" she sees these differences. I would be an idiot if I tried to deny her "belief" that she can see finer colour detail than I can. It would be arrogant and narcissistic. And, crucially, it would be just plain wrong.
I have normal colour vision, by the way - I've been tested more than once (usually as part of a sample population in science surveys done by students studying on my campus). But my wife has better-than-normal colour vision. Many people do, and many of them, it turns out, are women. There are even women who are believed to have four different colour receptors in their eyes - not just red, green, and blue, but also cone cells that respond to yellow. Hawks and some other birds of prey also have four different colour receptors in their eyes, which is why they can spot a dusty brown mouse on a dusty brown field from a hundred metres away.
How do we know my wife isn't just deluded? For starters, most of the women in the bead store can also see the colour differences that I cannot (women often have better colour vision than men). The bead manufacturer has given these different shades of yellow different names, because the manufacturer also knows that they are similar, but not the same, colour. And if we want objective proof, a spectrometer would easily provide it. Just as my LTSpice simulation provides objective proof of the op-amp clipping problem. (You could build the actual clipping detector circuit I designed if you want to - I didn't bother, since the simulation proves the point quite clearly.)
The harsh clipping I - and tens of thousands of other guitarists - heard from high-NFB SS guitar amps is real. It has nothing to do with imaginary "golden ears". You just need better-than-average hearing to hear it. Not everybody has that hearing ability, but lots and lots of people do.
It may be true that a person with statistically average hearing can't hear this particular type of distortion. But there is no doubt that thousands and thousands of musicans can and do hear it - which is why analog solid-state guitar amps developed such a poor reputation for sound quality, a reputation that lasted for many decades, starting somewhere around the 1960s or so - some sixty years ago.
Until recently, there hasn't been any publicly available objective proof that this phenomenon existed, or, at least, I've never come across it. But now we all have the proof. It's right there at the beginning of this thread.
More than half a century after those nasty-sounding early SS analogue guitar amps were first marketed, designers are finally figuring out how to make affordable solid-state guitar amplifiers that don't make harsh gritty sounds when they're set for clean tone. I'm starting to hear reviews some of them on You Tube.
Nowadays, apparently for the first time, there are both analogue and DSP-based solid-state guitar amps that manage to avoid making these nasty noises. And a few of them are affordably priced.
And that's a good thing for anyone interested in the electric guitar.
-Gnobuddy
After 400 posts, the mystery of the unwanted clipping is still not solved but gets more complicated.
Regards.
Regards.
A slomo video of a guitarstring being plucked hard will ilustrate what we are up against.
Cheers!
Cheers!
So, let's take the infamous Line 6 for example. Its founder Marcus Ryle is a professional session musician, acoustic engineer and a trained classical pianist
Michael Doidic is a guitarist and an acoustic engineer
Susan Wolf is a professional singer, songwriter and a pianist who started gigging in the age of 15
Jeff Slingluff (RIP), their sound designer (e.g. POD series, later Roland BOSS Katana series), was a professional guitarist
Yeah, what do these "engineers" know about music or tone. Why would those arrogant engineers hire "sound designers" and musicians to audition, test and develop their products when they, according to you, already know everything.
And this approach is different from practically what every company does in what way?
So have practically every other company, despite that their founders and engineers actually often were real musicians, and were even able to play guitar or design amps, unlike Leo.
Sigh.
Your personal bias on the matter is pretty obvious. Fact is that professional and extremely famous musicians like Dave Mustaine, Kirk Hammett, Steve Howe or Billy Sheehan would disagree with you. Who exactly is being arrogant here?
Last edited:
All I'm saying is that input stages of typical modeling amps have very high headrooms and are designed to tolerate dynamics of a guitar and bass signal. I demonstrated why and it well complies with Rod Elliott's measurements and worst-case scenarios assuming, say, even 2V peak inputs. We have anecdotal evidence that Bass POD does not clip even with a very hot pickup.
Why do you assume I was "fooled"? I am perfectly aware of sine wave's limits and merits as a test signal, and that a guitar DOES NOT output sine waves!
If their signal characteristics were new information to YOU don't draw a conclusion that many of us weren't familiar with them. I can imagine that amp designers have scoped guitar signals as long as there have been oscilloscopes. Their transient nature isn't exactly "news".
No. It hasn't. Engineers have been well aware of this for decades.
Ah, the good old tin ears and talent / professionality arguments.
I would rather say that guitarists are just fascinated by new gadgeds. I suppose many just experimented with different amps for whatever reason. There are very few guitarist who keep using the gear they started with. Any gear.
If you drive a tube input to grid clipping you get abrupt clipping. Why is this not a problem?
Yet that reputation hasn't stopped professional musicians from using those amps or the products still becoming extremely popular. I think the poor reputation is more a sekä sustaining myth than something based in reality.
The right answer would be you never came across it.
Well up until Fender actually started with DSP based amps (a bit late to the game on it), they had their Frontman solid state series. They did sell lots of those..were they that well-reputed? Peavey did pretty well on their solid state tube emulation, and there have been a few others that have achieved good performance from solid state, but in sales and volume probably pale in comparison to what Marshall and Fender sell in the solid state market.
The best seller is usually the cheapest one (like the Line 6 Spider about 20 years ago) while the expensive "pro" stuff generates less interest, if at all. But I think that's just standard way how the market functions.
My point was more in the vein that when people supposedly discovered the awfullness of SS amps did they vanish from the markets? No, they steadily became more and more popular. And did people return to those nice tube amps that they formerly had? No. They bought new and improved tube amps with different sound. Even the once supposedly loathed "digital" amps are all the time steadily increasing in popularity. Maybe the stuff just wasn't as bad as claimed. At least historical evidence doesn't strongly support that it was.
The trends of these come and go and only moderately lately has, say, "vintage" become a fashion. A Vox AC30 or Fender tweed Deluxe were loathed in the 1970's, the newer tube amps sounded and worked better. Fender was seriously struggling in the late 1970's to make any sort of fashionable amp and today Marshall's state is pretty much a ruin compared to its days of glory. New products just come to surpass the old stuff. That's how it goes. Bunch of designs that stood the time merely means a good 9/10ths didn't and we have forgotten That bunch. Does it mean the technology itself is bad or poor sounding? Of course not. And by the way, there is a trending interest towards those 1960's - 1970's horrible SS amps that are now "vintage".
And back to the topic, it's a pretty far fetch to assume guitar amp designers weren't familiar with rudimentary concepts like timbre and envelope and a 2021 Rod Elliott webpage (though fantastic work) would have been a major reveal in that regard. Just epic.
My point was more in the vein that when people supposedly discovered the awfullness of SS amps did they vanish from the markets? No, they steadily became more and more popular. And did people return to those nice tube amps that they formerly had? No. They bought new and improved tube amps with different sound. Even the once supposedly loathed "digital" amps are all the time steadily increasing in popularity. Maybe the stuff just wasn't as bad as claimed. At least historical evidence doesn't strongly support that it was.
The trends of these come and go and only moderately lately has, say, "vintage" become a fashion. A Vox AC30 or Fender tweed Deluxe were loathed in the 1970's, the newer tube amps sounded and worked better. Fender was seriously struggling in the late 1970's to make any sort of fashionable amp and today Marshall's state is pretty much a ruin compared to its days of glory. New products just come to surpass the old stuff. That's how it goes. Bunch of designs that stood the time merely means a good 9/10ths didn't and we have forgotten That bunch. Does it mean the technology itself is bad or poor sounding? Of course not. And by the way, there is a trending interest towards those 1960's - 1970's horrible SS amps that are now "vintage".
And back to the topic, it's a pretty far fetch to assume guitar amp designers weren't familiar with rudimentary concepts like timbre and envelope and a 2021 Rod Elliott webpage (though fantastic work) would have been a major reveal in that regard. Just epic.
Mandu, there is no mystery, the problem is entirely solved, and the explanation is not complicated at all.
Now that I understand the whole thing clearly, I can put it in very simple terms.
Here we go:
1) A Strat-type single-coil pickup puts out up to 1 volt peak-to-peak at the start of a strongly picked note on a single string. (See Rod Elliott's 'scope captures at the start of this thread, for example. Or plug your guitar into your own 'scope and see for yourself.)
2) Consider a 15-watt (max output power) guitar amplifier with an 8 ohm speaker. This means maximum unclipped output signal is about 30 volts peak-to-peak. (31 volts if you're really picky.)
3) Therefore, if we want to avoid clipping the guitar signal, we have to limit maximum voltage gain to 30 times.
(Why? A voltage gain of 30x will take a 1 Vpp input signal up to 30 Vpp, which is the maximum this amplifier can handle without clipping.)
4) If you actually try this - plug a guitar with single-coil pickups into a 15-watt power amplifier with a total voltage gain of only 30x from input jack to speaker terminals, you will find that the guitar volume is very low - far, far less than the amplifier is capable of. There isn't enough gain to make this amp usable.
Even for clean tone only, you need much more voltage gain than 30 times. (A single half-12AX7 gain stage in one of Leo Fender's classic tube amps usually has a voltage gain of more than 50 times, all by itself. And there are multiple gain stages in the signal chain.)
5) The only way to bring the guitar volume up is to increase the voltage gain. If you try the experiment you'll find that you need a voltage gain far, far more than 30 times. My experience was that ten times as much gain - 300 times - is still far too low to get any reasonable output from the amplifier. I found I wanted a gain well over 1000 just for clean tones.
6) If you raise the gain to only 300 times (which is too low), simple math shows us that any input signal bigger than 0.1 Vpp will clip the output. Remember, max output voltage is 30 Vpp; divide that by the gain of 300, and you have a maximum unclipped input signal of 0.1 Vpp.
7) Look back at item 1 - the guitar signal starts out at 1 Vpp, and decays from there. At some point it will drop below 0.1 Vpp and clipping will stop. But the entire portion of the signal before that - from 1 Vpp down to 0.1 Vpp - will necessarily be clipped.
Let's look back at what we've just shown: if a fifteen-watt guitar amplifier has an end-to-end voltage gain of over 30x, then the initial part of the guitar signal will clip badly every time you pick hard.
The problem only gets worse if you strum chords, if your guitar has humbucker pickups, or if the amplifier has more than 30x voltage gain. And remember, EVERY practical 15-watt guitar amp has much more voltage gain than 30x.
So: The clipping problem is real. The cause is simple. If you have enough voltage gain to make reasonable use of the amplifier's output power, then you also have enough gain to clip the initial part of any strongly played note or chord.
If the amplifier uses high NFB gain blocks (like op-amps and contemporary high-NFB solid-state power amps), then this clipping will be hard, harsh sounding, and abrupt. The results are audible, and unpleasant.
What about classic 15-watt tube amps, say a blackface Fender Princeton Reverb? The argument above says that these have to clip, too, as gain is far more than 30x.
However, these amps don't sound harsh in the same way.
We can only conclude that rather than harsh and abrupt clipping, the gentle curvature of tube characteristics does something much more audibly subtle. The big initial part of the signal is compressed with much less audible harshness.
-Gnobuddy
The whole 30X thing is bs, need to make up for the volume knob, tone control losses. Then a tube amp takes all this voltage gain and then trades it off for current gain. All those SS dirt pedals must be harsh, running at 9V no less. Wonder why they sell?
A no fancy high wattage bass amp head (around 200 watts) with guitar speakers sounds more clean to me when using guitar.
Just run at lower power, set amp tone controls to center and use external effects for the tone.
Somehow, by magic the preamp distortion becomes insignificant.
Regards.
Just run at lower power, set amp tone controls to center and use external effects for the tone.
Somehow, by magic the preamp distortion becomes insignificant.
Regards.
Indeed. Already very early posts of this thread there was discussion how the input stages powered by +/-15V rails are adequate to handle typical input signal ranges from guitar pickups, even hot ones. E.g. I'm looking at Marshall 8100 and the input stage gain is only 5.5x in clean mode. That is totally sufficient to tolerate 1V peak transients, heck even 2.5V peak transients, just like the earlier example of a Line 6 digital amp.
And then, a lot of gain is indeed often sacrificed to compensate losses, like say a logarithmic volume pot that drops the signal amplitude to 10 - 30% even at half of its rotation, or passive tone controls with -20 dB insertion losses.
And then, higher gain designs are supposed to clip. Yes, they are supposed to clipping distort.
And no, the tube distortion generally isn't any softer. On the contrary. They do clip pretty hard and if softness was a virtue the tube circuits would lose 10 - 0 to likes of CMOS linear amps, open loop differential amps like OTAs, or to even generic early 1970's distortion pedals like the Tube Screamer. All these, may I say, circuit elements employed in several solid-state amps.
And here we encounter an unsubstantiated premise that solid-state clipping distortion is for some reason inferior. As if we could ignore a whole 60 years of history with various solid-state distortion devices that still enjoy a huge popularity. Or that overall "softness" of distortion can't explain half of it, and was merely a similar failed attempt in it as even vs. odd harmonics.
Oh, but then there comes this "klever klipper" to the rescue... and no one even bothers to notice that it's just a rehash of a clipping limiter of Vox SS guitar amps devised already in 1968 and since then we have had about a half a century worth of development. It clips moderately softly you say? Yes, because that's like how those solid state diodes actually clip; with moderately soft knee. They have been an adequate solution to emulate clipping characteristics of a tube power amp section in who knows how many amps, including tube amps, and supposedly "all tube" amps. Yes, even that simplest of the simplest diode clipper.
Oh, it's tracking the supply voltage you say? ...just like Ampeg, Carvin or Sunn clipping limiters from the 1970's! No new concepts there.
And guitar amp designers supposedly don't know anything better than this? And don't know that guitar signal is not a sine wave, but a complex wave with envelope properties including transients?
What a load of bollocks. Sorry gnobuddy, but you should really do your homework better with this topic instead of arrogantly dismissing those that disagree with you.
And then, a lot of gain is indeed often sacrificed to compensate losses, like say a logarithmic volume pot that drops the signal amplitude to 10 - 30% even at half of its rotation, or passive tone controls with -20 dB insertion losses.
And then, higher gain designs are supposed to clip. Yes, they are supposed to clipping distort.
And no, the tube distortion generally isn't any softer. On the contrary. They do clip pretty hard and if softness was a virtue the tube circuits would lose 10 - 0 to likes of CMOS linear amps, open loop differential amps like OTAs, or to even generic early 1970's distortion pedals like the Tube Screamer. All these, may I say, circuit elements employed in several solid-state amps.
And here we encounter an unsubstantiated premise that solid-state clipping distortion is for some reason inferior. As if we could ignore a whole 60 years of history with various solid-state distortion devices that still enjoy a huge popularity. Or that overall "softness" of distortion can't explain half of it, and was merely a similar failed attempt in it as even vs. odd harmonics.
Oh, but then there comes this "klever klipper" to the rescue... and no one even bothers to notice that it's just a rehash of a clipping limiter of Vox SS guitar amps devised already in 1968 and since then we have had about a half a century worth of development. It clips moderately softly you say? Yes, because that's like how those solid state diodes actually clip; with moderately soft knee. They have been an adequate solution to emulate clipping characteristics of a tube power amp section in who knows how many amps, including tube amps, and supposedly "all tube" amps. Yes, even that simplest of the simplest diode clipper.
Oh, it's tracking the supply voltage you say? ...just like Ampeg, Carvin or Sunn clipping limiters from the 1970's! No new concepts there.
And guitar amp designers supposedly don't know anything better than this? And don't know that guitar signal is not a sine wave, but a complex wave with envelope properties including transients?
What a load of bollocks. Sorry gnobuddy, but you should really do your homework better with this topic instead of arrogantly dismissing those that disagree with you.
Oh, and while no doubt the clipping of the picking transients will be inevitable at some point of the signal path (and will surely be at any modest "warm" dial settings) I wouldn't be excessively worried about it. Why?
Because, as said, it's pretty much inevitable. Unless you invest in a 1000W high headroom super clean amplifier, which will just sound horrible "sterile" in the standard application of amplifying electric guitars then you'll learn that signal "compression" by clipping is pretty usual in guitar amps, and isn't usually even limited to that initial transient because it's clipping can be, to great extent, fairly inaudible due to psychoacoustics.
There are several reasons why. One is that the transient is a harmonic frequency of the fundamental in the first place and probably in the 1 - 3 kHz region. When we clip it we generate higher order distortion in the 2 - 6, 3 - 9, 4 - 12 kHz etc. regions and pretty much everything except the 2nd harmonic already gets filtered out by the bandpass filter that is the typical speaker system of guitar amps.
Second, it's not a low note or broadband clipping, just a single transient, and pretty much just the highest frequency component of the signal, snipped out, so the intermodulation effect remains very limited. It's not going to "fart out" and if there is any more distortion in the signal the transient distortion surely escapes us. Guitar signal tolerates far more distortion before it becomes gruesomly audible.
And usually we just refer to it as "warmth", "punch", "aggressivity", etc. The little added "brightness" might even help to "cut through the mix" better.
In fact, guitar signal is usually deliberately distorted much much more. And by the time there's compression and sustain at those lower amplitudes the clipping of the transient is going to be extremely hard no matter how "soft" clipping you employ. Even the supposed "soft clipping" of tubes is not going to save this from happening, assuming they could even clip all that soft In the first place.
So in this regard the "clipping" is not explaining why you, gnobuddy, feel tube amps sound better than solid-state amps. I would focus on looking alternative explanations.
Also, relating to second effect, the instrument signal is in "isolation" so it again tolerates far more distortion before overall IMD betrays it. The transient distorting might be a greater issue in a, say, a hifi or PA amplifier reproducing an entire musical arrangement consisting of various instruments. ...except that even then just clipping the transients is a very common trick in mastering process to compress the sound with least audibly detrimental results.
So it's not that audible to begin with. In fact, not clipping the transient might have bigger impact in terms of added noise or audible compression from trying to tame the transient with those variable gain amps with attack and release characteristics we also refer to as compressors. The transient is often too "fast" for them in the first place, and if it wasn't what's the equivalent of instantaneous gain compression? Yes, you guessed it; "soft clipping", which is distortion. Bring a transient peak down several dB and that IS excessive distortion no matter from what angle we view it. The only saving aspect is that the distortion in that transient isn't all that audible in the first place.
But if you have a decent waveform editor* you can try this very simple experiment: 1. Record a strum of a note 2. Use the editor to attenuate the waveform lobe or lobes that holds the highest transient and 3. alternatively clipping limit the said parts 4. Evaluate the results by listening
You will come to notice that treatment "2" in excess, even though not seemingly distorting the lobe, will add similar "scratchy" distortion tone to picking as treatment "3". That's because it still IS waveform distortion though in potentially most softest form. You may also notice that treatment "4" is not disturbingly audible, unless in excess expanding to cover several transients within the pick attack. You may also use the editor to mix the signal to a song arrangement with various instruments and hear how tiny amounts of audible distortion disappear in a bandmix.
I did these sorts of experiments already in the mid 1990's, some 30 years before the appearance of that Rod Elliott article, so don't come off expecting that I think guitar pickups output sine waves or wouldn't be aware of their envelope properties. LTspice, IIRC, accepted waveform inputs already in 2003 (nearly 20 years before Rod Elliott's article) being a great design aid in comparison to old-school method of breadboarding and "play tests" with real guitar input, which I think designers have employed since the 1940's or something. So... No. Designers don't limit themselves to sine waves.
*These are extremely great tools because they let one apply specific signal processing effects without any additional side effects. They allow listening to and comparing effects of just that single effect in isolation, say, like soft vs. hard clipping, or different transfer characteristics of non-linearity, unlike in comparisons between two different amps with said effects, where the amps themselves actually add myriad of side effects due to being two very different systems.
Because, as said, it's pretty much inevitable. Unless you invest in a 1000W high headroom super clean amplifier, which will just sound horrible "sterile" in the standard application of amplifying electric guitars then you'll learn that signal "compression" by clipping is pretty usual in guitar amps, and isn't usually even limited to that initial transient because it's clipping can be, to great extent, fairly inaudible due to psychoacoustics.
There are several reasons why. One is that the transient is a harmonic frequency of the fundamental in the first place and probably in the 1 - 3 kHz region. When we clip it we generate higher order distortion in the 2 - 6, 3 - 9, 4 - 12 kHz etc. regions and pretty much everything except the 2nd harmonic already gets filtered out by the bandpass filter that is the typical speaker system of guitar amps.
Second, it's not a low note or broadband clipping, just a single transient, and pretty much just the highest frequency component of the signal, snipped out, so the intermodulation effect remains very limited. It's not going to "fart out" and if there is any more distortion in the signal the transient distortion surely escapes us. Guitar signal tolerates far more distortion before it becomes gruesomly audible.
And usually we just refer to it as "warmth", "punch", "aggressivity", etc. The little added "brightness" might even help to "cut through the mix" better.
In fact, guitar signal is usually deliberately distorted much much more. And by the time there's compression and sustain at those lower amplitudes the clipping of the transient is going to be extremely hard no matter how "soft" clipping you employ. Even the supposed "soft clipping" of tubes is not going to save this from happening, assuming they could even clip all that soft In the first place.
So in this regard the "clipping" is not explaining why you, gnobuddy, feel tube amps sound better than solid-state amps. I would focus on looking alternative explanations.
Also, relating to second effect, the instrument signal is in "isolation" so it again tolerates far more distortion before overall IMD betrays it. The transient distorting might be a greater issue in a, say, a hifi or PA amplifier reproducing an entire musical arrangement consisting of various instruments. ...except that even then just clipping the transients is a very common trick in mastering process to compress the sound with least audibly detrimental results.
So it's not that audible to begin with. In fact, not clipping the transient might have bigger impact in terms of added noise or audible compression from trying to tame the transient with those variable gain amps with attack and release characteristics we also refer to as compressors. The transient is often too "fast" for them in the first place, and if it wasn't what's the equivalent of instantaneous gain compression? Yes, you guessed it; "soft clipping", which is distortion. Bring a transient peak down several dB and that IS excessive distortion no matter from what angle we view it. The only saving aspect is that the distortion in that transient isn't all that audible in the first place.
But if you have a decent waveform editor* you can try this very simple experiment: 1. Record a strum of a note 2. Use the editor to attenuate the waveform lobe or lobes that holds the highest transient and 3. alternatively clipping limit the said parts 4. Evaluate the results by listening
You will come to notice that treatment "2" in excess, even though not seemingly distorting the lobe, will add similar "scratchy" distortion tone to picking as treatment "3". That's because it still IS waveform distortion though in potentially most softest form. You may also notice that treatment "4" is not disturbingly audible, unless in excess expanding to cover several transients within the pick attack. You may also use the editor to mix the signal to a song arrangement with various instruments and hear how tiny amounts of audible distortion disappear in a bandmix.
I did these sorts of experiments already in the mid 1990's, some 30 years before the appearance of that Rod Elliott article, so don't come off expecting that I think guitar pickups output sine waves or wouldn't be aware of their envelope properties. LTspice, IIRC, accepted waveform inputs already in 2003 (nearly 20 years before Rod Elliott's article) being a great design aid in comparison to old-school method of breadboarding and "play tests" with real guitar input, which I think designers have employed since the 1940's or something. So... No. Designers don't limit themselves to sine waves.
*These are extremely great tools because they let one apply specific signal processing effects without any additional side effects. They allow listening to and comparing effects of just that single effect in isolation, say, like soft vs. hard clipping, or different transfer characteristics of non-linearity, unlike in comparisons between two different amps with said effects, where the amps themselves actually add myriad of side effects due to being two very different systems.
Last edited:
I see Gnobuddy is in great shape again! Don't get too worked up about it all, Teemuk. 😉
I think you should consider the hypothesis that people get used over time to crappy things. Just like they are getting used now to the often horrible sound quality of smartphones and bluetooth speakers. And autotune.
I will also urge to let it be. Its gone a little to far. You are both right and in some of your arguements wrong. Both of you great assets to the community in your own insisting ways😄 Kiss kiss!
Yes, they are. Ever try plugging the output from one of those SS dirt pedals straight into the input of a Hi-Fi system (or equivalent powered speaker)?
They are horribly harsh when you listen directly to the output of these through any kind of FRFR speaker.
If you run the harsh output into a reverb pedal and/or delay pedal and/or tube amp input, a lot of the harshness is smoothed away.
Even following it with a (soft) overdrive pedal helps reduce audible harshness.
If the room is big enough and the speaker is sufficiently far from your ears, even ambient room reverb - if there's enough of it - does a good job of smoothing out audible harshness from 9V SS "dirt" pedals.
It doesn't have to be a dirt pedal to get harsh sounds. My old Danelectro Fish-n-Chips graphic EQ has a built in adjustable clean gain boost as well as the EQ sliders. Turn the gain up, and you can quite clearly hear the output clip every time you pick a note on the guitar.
No need for the sarcasm. They sell for multiple reasons:
1) Nobody listens directly to the (harsh) output of one of these pedals. Musicians who know what they're doing will send the output through reverb, delay, maybe OD or modulation, then into a tube amp. That's how you get beautiful guitar tones out of them.
2) Musicians who don't yet know what they're doing - usually beginner guitarists - buy the pedals because their guitar heroes used the same ones. Plugged into an old-school analogue SS guitar amp, the pedal sound utterly harsh and horrible. The prototypical wasp-in-a-tin-can sound.
Beginner guitarists probably don't realize that when the the same pedal is plugged into overdrive / reverb / delay / modulation / a good tube guitar amp, it sounds very different.
So you believe SS pedals powered by 9V DC don't clip harshly. Listen to this 10-year-old video comparing a 9V SS amp emulator pedal (Joyo American Sound) with an actual Fender blackface amp:
Did they sound alike to you?
If so, go back and listen to the audio right around 1:18 minutes into the video.
Do you hear the harsh clipping of the SS pedal?
In this case the harsh clipping occurs from the guitarist using pick-hand palm muting. The impact of his palm on the strings produces a big low-pitched thump. The tube Fender amp handles it gracefully. The Joyo clips, hard, making that nasty rattle.
This is not the only place in the track where the Joyo clips harshly; but it's a particularly extreme and obvious one.
I bought that Joyo pedal before I'd ever heard this video, based on good reviews all over the Internet. I found I could get some usable sounds from it, but this sort of unpleasant harsh clipping was always lurking in the wings, and would show up every now and then. The pedal got retired for that reason.
-Gnobuddy
There are some good reasons for that!
Most bass guitar amps can be set for a pretty flat frequency response from roughly 80 Hz to 5 kHz (maybe wider). That's enough for quite decent vocals. Ma Bell phones only went out to 3 kHz, and AM radio wasn't any better. Millions of us grew up listening to music on AM radio, and enjoying it, too.
There are buskers who use a bass guitar amp for their vocals, and this is probably why. Flat to 5 kHz is adequate for acoustic/electric guitar as well as vocals.
There are audible vocal sounds above 5 kHz - sibilance, etc, which lends clarity to vocals. But in casual listening situations that's not a big issue.
I've played an electric guitar straight into a bass guitar amp, and I agree, the sound is less harsh than the same guitar into a typical analogue SS guitar amp (but also sounds duller).
That difference almost certainly comes down to the difference in loudspeaker frequency response: most guitar-specific speakers have an enormous treble peak at around 3 kHz, right where the human ear is most sensitive. That peak can give the guitar "bite" and "edge" when a good tube guitar amp is in the mix, but the same peak accentuates the harsh clipping from high-NFB solid-state gain stages in traditional analogue SS guitar amps.
For instance, here is Celestion's published frequency response for their very famous and very popular G12M Greenback guitar speaker. There is about a 10 dB (!!) peak around 3 kHz.
-Gnobuddy
Attachments
So you could have EQ'd your harshness away? Sounds like a better plan that to complain about it for ten years. Really, move on to something else.