Please help me understand


There are two concepts I've never been able to reconcile. How is it possible for different amps to have similar distortion levels (inaudible), yet still have what audiophiles might describe as a "tonal signature". In other words, how is it possible for an amp to have a perceived "warm" or "bright" sound, if it's accurately reproducing the input signal? It seems to me that all high quality, properly functioning amps should sound the same. If they don't, them some of them aren't doing their job very well.
danoroo
Dynamics.slew rates. current vs.voltage are some things that come to mind that will affect sound and not distortion figures.
As I understand, the way THD is measured involves measurement with a simple resistive load -- for example, an 8 ohm resistor. But speakers are not resistors. At the very minimum, the measurement doesn't fully capture what's going on.

Also, this is an interesting article:

http://www.passdiy.com/pdf/distortion_feedback.pdf
Two cars both have the same tires, but one is noisier on the road... different type outputs have their own character, different capacitors have the own, different resistors, the amount of and type of wire in a circuit sound different. Not to mention circuit layout. It would not be unusual for two identically spec'd amps to sound quite different.
Two cars both with 250 horsepower, but one clearly is faster than the other, one handles better, brakes better etc. Alot goes into engineering.
The measurements we can make are obviously not complete. Science is not perfect or tells the whole story. Everybody thought all cd players should sound alike because after all bits are bits. Then someone discovered jitter which can alter digital sound. If you can hear a difference but can't measure it it does not mean the difference doesn't exist.
Danoroo,

Music is just much more complex than any test while warm or bright is just different types of distortion. Why do you think amps should sound the same? What is to my liking is not necessarily your cup of tea. I like neutral sound while you may prefer warm sound. There is also synergy with the speakers and the rest of the system including listening room. In addition, since there is no perfect amp, some amps are better than others for given type of music. It would be nice to know ahead, performing some type of complex test, but as it stands now even known types of distortions (like TIM responsible for brightness) are not listed, being often too cumbersome to measure.

"some of them aren't doing their job very well"

Yes some of them are poor, but often you face compromise for the type of music you like - unless you have a lot of money (there are amps up to $1,000,000) and it can still have poor synergy with your system. Trust your ears and quit reading technical specifications because often amps with the best specs have the worst sound (and there are technical reasons for that).
The same way it is possible for two people to have the same height, weight and close size but look completely different. THD is a small part of the story.
"have what audiophiles might describe as a "tonal signature". In other words, how is it possible for an amp to have a perceived "warm" or "bright" sound, if it's accurately reproducing the input signal?"

for amps it is easier to understand - different components/circuit/design to PROCESS the signal...

but what i dont get it is how power cables, fuses, IC have "tonal signature" also ??? thats a lot bigger mystery...
wire is wire, it should all sound the same. bits are bits, all cdp's should sound the same. distortion is distortion, all amps should sound the same. this is what scientists would have you believe.

in the real world, distortion numbers are a very poor measurement for how an amp sounds since you can lower distortion numbers by increasing feedback. Most amp designers today realize that eliminating feedback is more important than lowering distortion numbers for an amps sound. That's why when you look at the price tag of many amps, the price goes up as the distortion levels go up. Scientists that buy amps by distortion levels can do so on the cheap.

whatever floats your boat.
This is why measurements have 'limited' use when it comes to audio components. The ears can simply hear far more than any current test bench equipment can hope to measure. Maybe one day, but certainly not now.I`ll choose what sounds good vs impressive numbers and graphs on a piece of paper any day.
I understand what you're saying Hakuchosan. One piece of 10 foot, 12 gauge copper (Cu) wire sounds "warm", yet another piece of 10 foot, 12 gauge copper (Cu) wire sounds "harsh". Go figure. And, thank you guys for your replies to my question.
Some good comments above. And wonder of wonders, not one comment claiming that they DO sound the same if they measure the same. I would just like to add that not only don't we yet know how to measure everything that might affect the signal passing through an amplifier, but we also underestimate how little the deviation from perfect has to be for the human ear to hear it. The human ear is more sensitive than any measuring tool.
PRetty much everything in the signal path makes a difference to some extent, some more audible than others. Add it all up and there is good chance two amps will not sound exactly the same.

There's a lot more to it than the minimal measurements and specs that are typically made available if you are lucky.
Hi all ! I agree with everyone above except the 1st response . The way an amp is made (talking about the chassis and covers ) has alot to do with the way they sound also . Think a big , heavy amp does bass so well because of the sheer output power ? That is only part of it , the amps weight and thickness of its panels also add bass . Dont think so ? Take a 10 lb weight from your weight bench , wrap it in a towel ( to keep it from ringing ) and set it on your amp . Lower and tighter bass .
What is really amazing is the human ear. The differences being discussed here are, electrically, pretty minor. But our ear hears them. Awesome engineering.
"The human ear is more sensitive than any measuring tool."

You sure you don't want to rethink this statement?
Why not just believe what you wish, but when you try to justify your beliefs it gets kind of silly. It's obvious that all the so-called analogies are not applicable. To try and discredit science in order to say one amp sounds BETTER than another, is just plain silly. The point is conceded, You can hear it all.
I hate to be the boring techno person here, but Bob Carver was right. all electronic circuits have what is called a transfer function that determine what the circuit will do over frequency with specific voltage/current variations. Unless the amps in questions employ exactly the same circuitry, there is no way on this planet that they will operate or sound the same. If you look at the results of the circuit's transfer function you will see exactly what I am talking about. Bob Carver did something very cool and tricky years ago. He got a really high end tube amp (don't remember the manufacturer). reversed engineered it and actually determined the amps transfer function (this is literally an equation showing the voltage in/out characteristics showing resistance, capacitance, inductance impedances for the circuit) then he managed to actually design an amp with the same (or close to it) transfer circuit) to show that 1) he could do it and 2) he could do it to sound just as good or better than the other amp for less cost) and guess what? he did it. really good amp. My point is that each and every electronic circuit when completed has a vo/vin transfer function. So if you computerize the transfer function and run signals in over frequency, you will see the amps characteristics are totally different over frequency and phase. They couldn't possibly be the same. not a chance. Also, I disagree that measurements have limited use. If you don't use measurements, there is no way on this planet you could actually design any electronic equipment. You design amps based on detailed specifications, as I have mentioned several times previously, such as phase, frequency response, gain, input/output impedance, load handing (what loads are it supposed to handle over the power band width and frequency), etc. once designed and built, guess what? you have to put it on a test bench and ....... measure the responses to see if the amp performs based on your actual design specifications. Any yes, we can measure for sound of particular amps. This is definitely where transfer function responses come in. but, this is not cost effective and unless it is a military device, no way on Earth will this type of design be done for amps. But make it worth someone's while and specifications and measurements will be developed (already are) to tell you how it will sound. Right now it is simply not worth it to spend the time and extreme money to do this. Because audiophile and music reproduction is not large enough for scientist to invest that amount of time and expense to do this. But, let me back up here, there are some audio manufactures that know what particular circuitry does sound like, Nelson Pass, and many others, and what they do is incorporate that particular circuitry in their amps all the time. The only time they change is when they develop new circuitry and amps and listen for some time and like the sound over the previous circuitry. this is physical measurements using human ears, not paper and pen or computer measurements. But don't think for a moment it can't be done. This is a physical universe, of course it can be measured. We determine many things and science by reverse engineering. I thought Bob Carver's approach for that particular tube amp was great.

enjoy
I do not think it is true that the human ear is more sensitive than any measuring tool. But what is often claimed is that while some levels of distortion are measurable, it is provably inaudible.

My argument would not be that such a claimed level of distortion is in fact audible, it would be that we are not appropriately measuring distortion through the entire system, and that the THD specs often listed do not accurately represent real world system wide distortion, not least of which being the interaction with a real world speaker.

Two cars may have an identical 0-60 mph number in a straight line, but once you get on the track, things like wind resistance, weight distribution, front versus rear versus all wheel drive, gear ratios, coefficient of friction, etc all come into play.

THD as a standard measurement does an ostensibly admirable job of presenting a metric given a very simple and constrained model. But it does not account for the bends on the road.
"If i take a vision test and have 20-20 vision does that mean I can see colors?"

no, but you could take this test and check if you do:
http://www.opticien-lentilles.com/daltonien_beta/new_test_daltonien.php

each manufacture when designing amps/speakers/preamps etc - are largely basing their testing on measurments and actual listening - now, depends on the equipment they have available
(the rest of) - it might sound perfect with what they have.
But it does not mean that it will sound same with other combination (for example Accuphase preamp with McIntosh power amp and reverse - just does not sound as good as manufacture match - if manufacture match, then sound is good on both but different). Simply speaking - each equipment has some kind of "equalization/sound proceesing" circuitry that will effect actual output. After that is up to the buyer to decide if they like it or not...
There are sound correction device (like McIntosh MEN220) that tries to compensate for all the equipment/room and make the output more listener "friendly" - but in some cases it can not.
As mentioned before about Bob Carver - yes, it is possible to make 2 different amps sound exactly same using different components - but why? if it was the case, then we would not need all the audiophile forums and discussions.....
so trying to achieve perfect sound by swapping components - it is a endless job, every time we change some component, we gain somewhere and loose somewhere else - but all this keeps audiophiles busy and generates nice profits for manufactures
Rok2id,
Who`s discounting science?it can explain alot but not 'every thing'. Certainly we`re able to detect audible characteristics with our ears that can`t be verified by measurements(yet), Why is that so hard to except. We can measure many things we could`nt 30 years ago and I`m sure in the future we`ll measure things we can`t today. In the mean while we have ears to rely on. Rok2id I`ll assume when you audition and decide to purchase your'audio' components you listened to them rather then just read the measurements and buy based on that criteria alone. Science is great but it does`nt at this stage have all the answers.
"The human ear is more sensitive than any measuring tool."

My post was primarily in response to the statement above. It should be self-evident that human senses, among the animal kingdom, SUCK!! we build machines all the time, to see, hear, smell, and detect things that human senses cannot. We have systems on submarines that can ID other subs by the sound they make. I am sure a machine can be built to ID every amp on the planet, based of that amp's output of a known input. But, who would want to do that? My post may have been an over statement, but if people say they can hear wire and amps and all the rest, FINE! Just leave science out of it.
Just food for thought, your family pet 'hears' more of your high-end system than you do. :) Maybe we need a K-9 audiogon hahahahahah those would be some interesting threads!!
Peace.
Another point not really mentioned much here is that equipment designers do not just use measurements/science, they also use their ears, and they design something that sounds good to them. The vast majority of them understand that it is all about the music and how it sounds, which is ultimately more of an aesthetic choice rather than a scientific one - a choice made with their ears, not their numbers.
"In the mean while we have ears to rely on. Rok2id I`ll assume when you audition and decide to purchase your'audio' components you listened to them rather then just read the measurements and buy based on that criteria alone. "

actually, that is exactly what I do. I have never auditioned a piece of gear, and that includes speakers. I bought my JBL 150s because they were 'known' to be good party speakers. I got my polk Ls90s, because AUDIO magazine said they sounded just as good as the high end stuff, and FT Hood got a truckload of them and sold them for 600 a pair. MSRP was 1600. I got my Polk Lsi 15s because they were said to be better than the Ls90s.

Sadly, my current and past gear has not been worthy of an audition. One more great thing about lo-fi. plus, you save on gas. My next speakers will cost around 5000, so there is an audition in my future. :)
"Just food for thought, your family pet 'hears' more"

Man: 20Hz-20kHz
Dog: 40Hz-40kHz

Same number of octaves (10)
Man: 20Hz-20kHz
Dog: 40Hz-40kHz

Same number of octaves (10)

Based on your data, you agree with me. Since Dogs don't listen to loud rock music or use headphones, or work around loud high pitched machinery, etc.... they can probably hear from 40-40k most of their lives. I doubt if anyone on a'gon can hear 20-20k flat.
Peace.
My dog loves to listen to my system even when cranked. He sits relaxed and his ears twitch as he soaks it all in. I take that as a good sign.

HE likes to chase animals. When animal noises are played, his ears really perk and he has to think about what he hears.

The dog barking and howling in "Seamus" by Pink Floyd sends him over the top every time. Also "Dogs" from PF's "Animals"

My dog may have some issues.
"Dogs don't listen to loud rock music"

Sure they do, they listen to owners loud music. They also bark - a very loud noise.

Everybody looses hearing with age - man or a dog. Prolonged loud noise can only make it worse.
I think it's quite possible that some test equipment can be more accurate than the human ear in some specific regards. I think that on some level, it might be advantageous to be able to evaluate one set of senses with alternative ones. I think it's quite possible that the human ear might be on some level analyzing things that are either yet to be identified and labeled, or yet to have test equipment available to verify. I also think it's quite possible that some of that increased accuracy is beyond the scope of perception of the human experience and therefore superfluous on it's own. On the other hand it might be meaningful in other ways. I also think it's possible that we might not have all the test equipment yet to test all parameters that are relevant to it's intended audience. I think it's quite possible that we have yet to fully understanding as to how these different test result parameters interact for it's intended audience. I also think the intended audience, is varied and inconsistent. I think it's possible that we might be able to recognize through test equipment deficienceies in audio reproduction gear, and are challenged develop a fix. I think it's prudent to continue to use and develop new test equipment along with reference to it's intended audience; our hearing, at least for for the time being and the immediate future.
Unsound, you have spoiled everything by introducing reasoned, well thought out common sense.

Maybe the machines can 'hear' all there is to hear. They hear everything the human sense of hearing detects. maybe the key is, how the brain interpets and processes those sounds. Machine can't do that. And that could be different for different people. Thats how some people can listen to Mozart and 'hear' noise and then get all weepy eyed ove tom waits. Your post was very informative.
Peace.
"quite possible that the human ear might be on some level analyzing things that are either yet to be identified and labeled"

Transient Intermodulation, unknown until 70s, comes to mind. Early SS amps sounded very bright because of overuse of negative feedback but designers claimed that it has to sound better than tube gear since THD and IMD are way lower. TIM is very seldom tested since there is no established test method. Even the simpliest specifications like THD don't show the full picture since it is different at different frequencies and power levels.

As for being able to repeat particular amp, I'm afraid that no matter what you do SS won't sound like tube gear simply because of completely different interaction with the speaker. SS amps are voltage sources while tube amps are power sources. You can attempt to emulate sound of particular amp/speaker combination in DSP processing (many guitar amps do it) but results are rather poor. Nobody, so far, was able to make electric piano sound like real one - not even close. We are pretty much on the level of trying to understand why wires have particular sound (silver vs. copper etc.). Even same tubes made by different manufacturers sound different. It is still mix of science and art.
I going with: unless and until you hear it for yourself, you'll just have to
take it or leave it on faith. If and when you hear it -- as we appear to have
demonstrated -- I'm afraid that it doesn't get one bit easier to explain....
(And people claim this "hobby" doesn't strike as a religion.
Jeesh.)

But I definitely will agree that the divide between registering data
transmitted by sound waves -- which a machine can presumably be
created to do equally well, if not significantly better, than the human ear --
and perceiving sound in the human brain (and extrapolating from it
meaning, or emotion, or anything else) is significant and not all that well
understood. There simply a lot we can't yet explain out there in the
world....