That was a long, slow, error-prone process of discovery for me as well. I found it helpful to realise that electric guitar amps - at least in their more distorted incarnations - aren't intended to reproduce the sound of an instrument, but rather, to create it. They are sound generators as much as they are sound amplifiers.
I think the word "clipping" is ripe for misunderstanding. An op-amp is designed to operate with very large amounts of negative feedback, which makes it extremely precise and accurate - until it clips. When it does clip, the effect of all that negative feedback goes awry, and as a result, the clipping is harsh, abrupt, and very unpleasant to listen to. Well, unless you grew up listening to death-metal.
The "harsh abrupt clipping" part is not subjective; it is mathematical fact, and the inevitable result of clipping a high negative-feedback amplifier.
The abrupt clipping of high-feedback audio amps is simply not a problem for Hi-Fi, despite thoglette's misgivings; you simply ensure your amplifier has enough power to never clip at the loudest SPL you can actually use in your home, which is quite easy to do. 100 watts RMS will get you to 105 dB with typical Hi-Fi speakers, and 105 dB will (a) make you deaf, (b) result in an immediate visit by the police, and (c) drive your loved ones out of the house.
In more precise engineering terms, the distortion from this type of (op-amp or similar high-feedback) gain stage is inaudibly low right up until the threshold of clipping. As soon as clipping starts, distortion is severe and takes a very unpleasant and extreme form.
To a guitarist, this amp transitions abruptly from too-clean (sterile, cold, thin) to too-harsh (buzzy, fizzy), with nothing in between. This is okay for the death-metal crowd who will simply turn up the gain until deep into too-harsh (for my ears) territory, and stay there. This type of amp is, however, useless for genres such as blues or pop, where we want small, but progressively increasing, amounts of distortion as the guitar is played progressively harder and harder.
Meantime, a triode (valve) gain stage will distort a few percent at relatively low volume - enough to hover at the threshold of audibility, not being heard as actual distortion, but rather, as a warmer and more lush clean tone. Turn up the input some more, and the output will distort a little more, still without ever actually clipping (clipping being defined as limited to a maximum peak voltage which cannot increase with further increases in input amplitude.)
You can keep this up over a range of input signals, which means the guitarist can, to some degree, vary the timbre of his instrument simply by varying playing dynamics. The amp need never enter actual clipping, just progressively increasing the level of nonlinearity.
Using valves is not the only way to generate progressive, subtle distortion, though. Like many others, I have been tinkering with ideas to generate progressive distortion from solid-state circuits for many years. The attached images show one example, a JFET circuit I came up with. The JFET is wired in common-gate mode, which is not typical in audio circuits; a BJT emitter-follower drives the low-impedance JFET source directly, eliminating the linearising effect of a source resistance.
The images show this circuit generating a significant (and certainly audible) amount of distortion, which would probably be heard as a clean tone, but a richer clean than the thin cold sound of an unclipped op-amp stage.
The second image shows an undistorted sine wave (purple trace) overlaid on the output of the circuit (green), so you can see just how much distortion is actually being generated. Note that there is no actual clipping whatsoever, though - neither half-cycle of the output is flat-topped!
An FFT of the output waveform shows a large amount of second-harmonic distortion, and inaudibly low levels of all other harmonics. This is no accident, as the goal was to use the square-law characteristic of the JFET to extract musical-sounding second harmonic distortion.
For my musical tastes, this is where good clean tone comes from - progressive soft distortion, long before any actual clipping occurs. Op-amps configured as the usual textbook inverting or non-inverting gain stages fail miserably at this.
-Gnobuddy
Attachments
I am not quite sure that a 100-watt amplifier will get you a CLEAN 105db without pretty efficient speakers. Let's say 100 watts per channel, 200 watts peak. Into a typical 86 db-sensitive HiFi speaker, that gives you 109db peak. Allowing for 14 db of rms-to-peak headroom leaves you with 95 db SPL; stereo would make it 98 db SPL.
Exactly. And 109 dB peaks will damage your hearing permanently within minutes, infuriate your neighbours, and probably cause your family to flee to preserve their sanity (and hearing.)
A typical home vacuum cleaner is loud enough to be annoying, and unacceptable at certain times of day - and that's only about 70 dB SPL at 1-metre distance. By contrast, 109 dB is the equivalent of nearly eight thousand vacuum cleaners. In your home. In one room. Three feet from your ears.
It's been a couple of decades, but I've tried to measure average power to my speakers using an RMS (analogue) voltmeter while listening to loud music, and the average was always well below 1 watt. That might have translated to ten times as much power on peaks, but I don't know, as I didn't have the equipment to measure that.
-Gnobuddy
a speaker may well be sensitive at a single frequency (which is the typically reported value for sensitivity) but if you don't specify a bandwidth and a tolerance where are you?
i feel the same goes for hearing damage without specifying frequency and duration (weighting) a peak of 109db doesn't necessarily constitute damage. in my arena touring days i've seen 110-114 at the mix position 200 yards out from the stage (not me at the helm, i agree concert sound can be ludicrously loud)
constant magnitude broad band noise is worse (i.e. industrial noise)...oh wait i forgot this thread was about distorted guitar!!!
i feel the same goes for hearing damage without specifying frequency and duration (weighting) a peak of 109db doesn't necessarily constitute damage. in my arena touring days i've seen 110-114 at the mix position 200 yards out from the stage (not me at the helm, i agree concert sound can be ludicrously loud)
constant magnitude broad band noise is worse (i.e. industrial noise)...oh wait i forgot this thread was about distorted guitar!!!
Last edited:
Since I have no medical credentials whatsoever, I will keep quiet on this one.
I will say, however, never allow myself to be exposed to 109 dB peak levels without serious hearing protection. Nor to 85 dB for 8 hours in a row.
Having read some of the medical research that led to the OSHA recommendation that eight hours a day of 85 dB SPL is safe, I will note that the researchers concluded it is only "safe" if you are exposed to no noise at all - zero dB - for the remaining 16 hours of each day, so that your traumatized ears can recover. Work for 8 hours, then jump into your sensory isolation tank, where you remain until you start work the next day. Sure. That will work!
Speaking of concert levels, if you Google "loudest concert", you will find reports of SPLs over 140 dB. Complete insanity. There are few things more miserable than a deaf musician, except a deaf musician tortured by tinnitus.
-Gnobuddy
...except it doesn't include distance from the sound source. e.g. that's the only way an orchestra can be less loud than an individual instrument.
Still, it does highlight that instruments get very loud.
I was playing with race cars when the governing body started to get interested in noise management. I recall that for some cars (mostly rotaries and high revving fours) we were measuring the proposed "1m limit" levels at the pit wall. I only discovered those little yellow ear plugs afterwards. :-(
Not surprisingly I blame my hearing loss on fast cars, fast music and some (self declared) loud women. Oh, and flyback transformers.
p.s. if you're a golfer beware of the current crop of metal "woods". They're a known cause of hearing loss
Last edited:
I didn't know about golf club induced hearing loss. But I do know that musical instruments - classical ones, not just loud electric guitars - have a long history of causing hearing loss. Apparently piccolos, flutes, and violins are among the worst offenders. All three can not only be very loud, they also produce that loud sound very near to your ears.
avtech23, a few years ago, I used an online chart similar to the one you posted to estimate how much power a truly apartment-friendly electric guitar amp would need. With a typical guitar speaker (say 95 dB@1W@1m), I calculated only a few milliwatts would be permissible at night, and 300 mW would almost certainly be too loud for most apartments (at any time of day or night). When I say "most apartments", I'm assuming normal people (i.e. not twenty-something metalheads) living on the other side of walls, ceilings, maybe floors.
My most recent DIY guitar amp project puts out an estimated 250 mW. It's discussed here, in a thread started by Printer2: Mini-amp for Output Tube Distortion . Plugged into the stock speaker in my Fender 65 Princeton reissue, it is in fact too loud to overdrive without the likelihood of annoying the neighbours. I don't want to annoy my neighbours, so I need to use a speaker attenuator with my quarter-watt amp...no kidding!
-Gnobuddy
Generally the measurements refer to the person's physical location, and are weighted in the time or frequency domain. Obviously 135dB 5 miles away isn't going to mean much.
I've attached a rather dry OHS guide for anybody interested in the 'official' blurb.
I've been lucky to avoid too much other than slight reduction in high frequency on one side. I'm lucky to work in an environment where those little yellow earplugs are plentiful and Peltor hearing defenders are standard issue.
Me neither. I wonder if that has to do with the high frequency nature of the noise. I know the high frequency component of jet engine noise is particularly damaging.
I found an article here: Analysis of impact noise induced by hitting of titanium head golf driver. - PubMed - NCBI
I never really understood how little power is actually required until recently when playing around with the almost-5-watt princeton I built.
Many dubious entries.
110dB power mower is highly unlikely. (Maybe when my neighbor mows rocks...)
Disco/LoudBar is rarely 120dB.
140dB Rock Concert is absurd, even in the front row.
150dB "Chest begins to vibrate" is just so wrong. Chest is a linear system. It vibrates at low intensity, just not much. Chest effect is very obvious at guitar amplifier levels: little from a 6W 6" Champ, more from a eight-10" 100W, so in the 90dB-100dB zone. Profoundly deaf people "hear" through chest at medium music levels.
There are better charts.
One of the striking things about charts like the one you posted, is how few natural sounds are listed. In the chart you posted, everything above 65 dB comes from an artificial (human-created) source - there were no "noisy restaurants" or "busy hotel lobbies" on the African savannah a quarter of a million years ago, when evolution was fine-tuning the hearing mechanisms of our nearest ancestral species.
Our hearing mechanisms, like our hands and brains and the rest of our bodies, still date from a couple of hundred thousand years ago. Our ears were designed to maximise survival on our ancestral grassland habitat.
The loudest sounds heard with any significant frequency in nature are probably waterfalls and large animals roaring. Maybe the sound of a too-near lightning bolt.
A lion roaring three metres away is a heck of a lot louder than 65 dB, but if an angry lion gets that close, you're dead anyway. So evolution is not going to modify human hearing to tolerate nearby roaring lions without damage. Ditto for a lightning bolt close enough to scorch your eyebrows.
Waterfalls? They are very loud, but the noise is pink noise, and the nature of a waterfall tends to keep humans at some distance from the loudest parts of it.
Tropical forests can be loud, with hundreds of noisy animal species all yelling at once. Many of those species have average lifespans of just a few years, though; their hearing didn't have to last long, because they'd be dead anyway.
By comparison, the ability to hear very quiet noises was extremely important for survival. If you didn't hear the little brook gurgling 500 metres away, you might die of thirst. If you didn't hear the rustle of a poisonous snake in the grass, you might die of snake-bite. If you didn't hear the pack of lions roaring a few kilometres away, you might not hightail it out of there, and they might eat you in your sleep.
So it doesn't take much thought to realize that evolution refined our hearing mechanisms to be acutely sensitive to quiet sounds. It also seems clear that there was no reason whatsoever for evolution to have evolved our hearing in such a way as to be able to tolerate sounds much louder than typically found in the pre-industrial era. Without such fine-tuning, any ability we have to tolerate SPLs higher than those found in pre-industrial human habitats would be purely accidents of nature.
For these reasons, I think it is only our near-complete disconnect from nature and our evolutionary origins that lets us believe that we can somehow tolerate, say, sustained 110 dB sound pressure levels without suffering hearing damage. To me, that seems like the equivalent of believing that our knees and backs should be able to cope with carrying a five-tonne weight, or our skin should be able to survive contact with boiling water unscathed.
It doesn't help that there are powerful vested interests who didn't want the expense of having to install sound proofing to quiet noisy industrial environments, so they managed to lobby successfully to have their lawmakers set a "safety standard" that is quite unsafe for your hearing. (But that useless standard puts more money in the hands of the factory owners. Who cares about a few thousand assembly-line workers suffering from tinnitus and hearing loss as they age?
)Incidentally, since there are several guitar speakers with sensitivities of 100dB@1W@1m or more, it only takes a ten-watt guitar amp to generate 110 dB SPL a metre away from the speaker. No wonder so many famous guitarists (and other musicians) now suffer from varying degrees of hearing loss and / or tinnitus. Here are just a few of them:
- 37 Musicians With Hearing Loss - The Cavan Project
- Rock '''n''' Roll'''s Not-So-Silent History With Hearing Loss
-Gnobuddy
Gnobuddy I really agree about all your thougts about guitar sound and hearing. The whole line from the guitarpickups to your ears or the mixer acts like filter ading limiter, eq, compression, distortion etc. to make bearable to listen to. -So far from what we want from hifi. A clean guitar can sound good through a stereo, but the attack peaks from stromming are teribly loud for the ears. Then again OPs choice of components may sound great although I doubt it a bit.
If using ampsimulator - To me it would be better to use hifi speakers or ad a tweeter.
If using ampsimulator - To me it would be better to use hifi speakers or ad a tweeter.
Wait.. I had an epiphany...
Those of an American persuasion call trousers = 'pants'...
Urban Dictionary: pants Might help
Nicely put!
Exactly, especially if you turn up the volume to anywhere near performance levels. The sound may be bearable if it's kept to very low volume.
-Gnobuddy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.