Decibels, distortion, amplifiers and golden ears
Jul 29, 2007 at 9:34 PM Post #31 of 790
At least in Germany everything is regulated.
So is everything related to Hifi and it's measurements.

So if I read "THD <0.01%" then this value is -at least in Germany- correlated to the the value at 1kHz. No statement is made about the harmonics or their distribution over the spektrum.

m00h
 
Jul 29, 2007 at 9:46 PM Post #32 of 790
Quote:

Originally Posted by elrod-tom /img/forum/go_quote.gif
EDIT: We ought also be mindful of the ways in which our own argumentative style might be needlessly irritating or provocative as well...


What is it that is so irritating about my style?

I hear that a good bit, but when I ask for specifics no one seems to be able to tell me anything.

I don't throw insults around and I try to stick to the subject at hand.

After this I probably won't post here any more. Truthfully, I'm tired of being insulted when I have offered no insults myself.

As I wrote in my OP, I know my opinions irritate many here, but do I not have a right to those opinions and to state them as I wish as long as I stay within the rules?

Why I mentioned my insurance and financial situation?

To point out that we never know what's coming over the horizon at us, hobbies are all well and good and audio is one of my hobbies. But I try to keep it within the realm of rationality.

Lately I have bought considerable audio gear, but I have sold off other things to do so.

There is a tremendous amount of high quality audio gear on the used market. I found out long ago that buying high tech equipment new is not a good deal, it goes obsolete too fast as technology progresses.

I paid almost $700 for a digital camera a few years ago which I just sold in order to buy audio gear. I got $87 for it despite the fact that is was cosmetically and operationally perfect and had many accessories which did not come with the camera new.

I won't make that mistake again.
 
Jul 29, 2007 at 9:50 PM Post #33 of 790
Quote:

Originally Posted by guttorm /img/forum/go_quote.gif
this is what i was thinking. also i would expect even a good cartridge to have more distortion than headphones because of its greater amount of fine movements, making precision more difficult.


And yet there are some on this very site who think vinyl sounds better than digital.

I convert vinyl to digital and I can almost always make it sound better than the original, at least in the sense of removing the clicks and pops and taking the background noise down to about -60 dB.
 
Jul 29, 2007 at 9:55 PM Post #34 of 790
Mr ripper I welcome your views.

I agree that mechanical transducers are the "weakest" part of the
playback chain due to the vast differences in frequency response,
distortion, resonance of the materials and interaction with differing
ear & head shapes. The differences between various headphones,
speakers or cartridges will probably always be more pronounced that
the differences between different amps.

This is not to say that I believe that all amplifiers or DACs or even
cables sound the same.

I have no problem being informed by both sides objective or
subjective and keep an open mind and my eyes on my wallet.

But anyone dogmatically on either side is suspect in my mind.
 
Jul 29, 2007 at 10:07 PM Post #35 of 790
Quote:

Originally Posted by JadeEast /img/forum/go_quote.gif
I have no problem being informed by both sides objective or
subjective and keep an open mind and my eyes on my wallet.

But anyone dogmatically on either side is suspect in my mind.



I suspect that most people would agree with this position. It's also the reason that we don't to DBT discussions in the cables forum. It ends up being a circular argument with no resolution point...and people tend to get irritated at that.
 
Jul 29, 2007 at 10:08 PM Post #36 of 790
Quote:

Originally Posted by JadeEast /img/forum/go_quote.gif

I have no problem being informed by both sides objective or
subjective and keep an open mind and my eyes on my wallet.

But anyone dogmatically on either side is suspect in my mind.



Thank you.

I do try not to be dogmatic, in fact if you look at the thread I linked to I had some of the most golden ears on the thread, despite the fact that my hearing above 8 kHz is mostly gone unless I turn the volume up almost to the point of pain.

Granted, I may have just guessed the last stage or two in the test, but the same could be said of anyone taking the test.

Personally, I think my success with that test was because I have a long history of listening for distortion in recordings.

Back in the days of cassettes I had a fairly high quality cassette recorder and everyone I knew came to me to record their vinyl to cassette. I knew what brand and type of cassette which gave the most bang for the buck. I knew just how high into the red I could push the VU meters without creating obvious distortion. I kept my heads clean and demagnetized them regularly.

As with many things, give an incompetent person the best tools in the world and he cannot do as well as a competent person with far lesser means at his disposal.
 
Jul 29, 2007 at 10:14 PM Post #37 of 790
Quote:

Originally Posted by TheVinylRipper /img/forum/go_quote.gif
There is a tremendous amount of high quality audio gear on the used market. I found out long ago that buying high tech equipment new is not a good deal, it goes obsolete too fast as technology progresses.


How can a piece of equipment that reproduces sound perfectly go obsolete? Or how can a piece of equipment that has a distortion below hearing level get any better than below hearing and therefore go obsolete?

Just a set of rhetorical questions.

This begs questions from me which is; is your position that sound reproduction is a moving target? Or is it that if distortion is below the hearing level that everything is equal? Or what are you saying exactly?

BTW, I don't want you to go away either. I think you are raising some interesting issues if we can discuss them. This is how I learn most of what I truly learn. Your return on my harmonic issue made me think about which harmonics I was addressing in my first post.
 
Jul 29, 2007 at 10:16 PM Post #38 of 790
Quote:

Originally Posted by TheVinylRipper /img/forum/go_quote.gif
This is a graph of the frequency response of my AT440MLa phono cartridge playing a 300 Hz tone from my quite expensive test record which I bought specifically for this purpose. Note the broadness of the response. Note also that the THD is nearly two percent and that the harmonics are -50 dB for the strongest one.

4pk0j2s.gif



One thing that irks me is when people publish plots made using fourier transforms without specifying the window function. The window function can have a huge impact on how things appear. More specifically, it can have a huge impact on how sharp the peaks appear. Your screenshot shows that a Blackman Harris window is being used, which is good, but comparing it to a graph made with an unknown window function is going to be difficult at best, especially when it comes to a subjective comparison of how sharp the peaks are.


Quote:

Originally Posted by Roam /img/forum/go_quote.gif
On why THD numbers do not correlate well with perceived sound quality (PDF file) This is a graduate level thesis which was passed and approved by Electrical Engineering professors. It's based on the prior art of tests conducted by the BBC labs decades ago, the results & conclusions of which have unfortunately become lost for the most part in modern hi-fi. To briefly summarize, it's not the THD number which is important, rather, it's the distribution of the harmonic spectra.



From the linked paper:
Quote:

The output impedance of these amplifiers is in the ohms range, hundreds of times worse than solid-state push-pull amplifiers. This requires careful mating with loudspeakers that do not have great impedance variation with frequency. For example, a set of 1960’s Radio Shack 16 ohm PA monitors was necessarily chosen in the listening tests because it had a very little impedance variation over the range of frequencies used in the listening tests- a grand piano.


I am willing to bet that not every set of cans out there has the kind of flat impedance curve necessary for this fellow's findings to apply. In fact, I'm willing to bet that this is one reason why people choose certain can/amp combinations. They're looking for something that sounds different. They obtain and carefully bias a tube amp that promises very low distortion, and then they mate it with cans that have an all-over-the-map impedance curve and bask in the warm euphonic distortion that results. There is nothing wrong with this, but one shouldn't fool oneself about the accuracy of it.

Quote:

Originally Posted by slwiser /img/forum/go_quote.gif
We have a memory of what a true note sounds like and what true music sounds like. The difference between when that music is first made and what comes into our ears has many significant things that happens to it. First it has to be miked then it goes into a long train of devices to be pressed into a form to be reproduced by our home equipment which does its thing. What we finally want is a sound which matches what our memory has of what that sound should be. All our trials of which components we like best comes to which set makes the best reproduction of that memory.


It might have been F. Alton Everest's book that I read, or it might have been another source, but I do recall reading a study on human memory and how it relates to sound. It's actually piss poor. We remember a lot of qualitative things about a sound, but when it comes to subtle differences most people can't even walk from one room to the next without forgetting. If you're chasing some golden sound from your memories, where you find it will have very little to do with what the original sound.


-----------------------

My own take on this subject is that the better an amp is, the less it should alter the original waveform other than to make it bigger. This means that the best amps *should* sound remarkably similar because they're coming so close to perfect reproduction. However, this is often not the case. You often hear expensive amps that "wow" you with some kind of difference. Why is this so? If you're selling an outrageously expensive amp you need to justify that expense. Any kind of "wow" you can produce therefore helps, even if it means deliberately introducing coloration.

Now, I should note here that I'm coming from the speaker side of things and am basically a neophyte when it comes to cans. However, high end speakers and amps are rife with "wow" factor, and I haven't seen anything that makes me think cans and can-amps are any different. People are still using the same ancient tube designs that, if carefully biased, pampered, and paired can *approach* the accuracy of a solid state system, but typically do not. Typically, they have at least a little euphonic distortion that gives them their distinctive sound. i.e. Their "wow". Again, I say there's nothing wrong with wanting a little "wow". Just don't try to delude yourself or others into thinking that you can have accuracy at the same time as your "wow".

This, naturally, brings us full circle to the OP's contention that the amp doesn't matter. He's right in that amps designed for accuracy tend not to matter. However, amps designed to add a little "wow" do make a difference. Some spend a long time looking for the "wow" that suits them. Those that have found it will never be convinced that amps make no difference precisely because they found amps that *do* make a difference. Convincing them that difference is actually a form of distortion is yet another uphill battle, but it's at least one you have a hope of winning, if only a very slim one.
 
Jul 29, 2007 at 10:23 PM Post #39 of 790
Quote:

Originally Posted by TheVinylRipper /img/forum/go_quote.gif
What is it that is so irritating about my style?


Well, a couple of things...

Quote:

Originally Posted by TheVinylRipper /img/forum/go_quote.gif
I do try not to be dogmatic


Quote:

Originally Posted by TheVinylRipper /img/forum/go_quote.gif
in fact if you look at the thread I linked to I had some of the most golden ears on the thread


This is interesting...first, we're not dogmatic. Then, we refer to those with whom we apparently disagree "golden ears". This is the sort of thing that irritates people.

Another example...

Quote:

Originally Posted by TheVinylRipper /img/forum/go_quote.gif
As with many things, give an incompetent person the best tools in the world and he cannot do as well as a competent person with far lesser means at his disposal.


The assumption, apparently, is that you are the competent person, and others (the "golden ears" perhaps?) are incompetent. Again, this is the sort of thing that tends to piss people off.
 
Jul 29, 2007 at 10:29 PM Post #40 of 790
Quote:

Originally Posted by Cor /img/forum/go_quote.gif
This, naturally, brings us full circle to the OP's contention that the amp doesn't matter. He's right in that amps designed for accuracy tend not to matter. However, amps designed to add a little "wow" do make a difference. Some spend a long time looking for the "wow" that suits them. Those that have found it will never be convinced that amps make no difference precisely because they found amps that *do* make a difference. Convincing them that difference is actually a form of distortion is yet another uphill battle, but it's at least one you have a hope of winning, if only a very slim one.


Yes, but the point that seems to be lost in this quote is that I and many others frankly don't care how the difference is characterised. If it's a more musical sounding (because of what you call distortion, or for whatever reason) amp, then that's what I want. If it sounds better to my ears, then I'm willing to pay more for it. You're trying to win an argument about a point of detail that means nothing to me.
 
Jul 29, 2007 at 10:30 PM Post #41 of 790
Quote:

Originally Posted by Cor /img/forum/go_quote.gif

It might have been F. Alton Everest's book that I read, or it might have been another source, but I do recall reading a study on human memory and how it relates to sound. It's actually piss poor. We remember a lot of qualitative things about a sound, but when it comes to subtle differences most people can't even walk from one room to the next without forgetting. If you're chasing some golden sound from your memories, where you find it will have very little to do with what the original sound.



Cor, I think I will let that book and what the author said speak for itself if you desire to read it. He uses double blind testing in a lab environment to come to his positions about memory and how it relates to our perception of music. His discussions about memory is specifically addressing our memory of musical qualities rather than memories in general.
 
Jul 29, 2007 at 10:40 PM Post #42 of 790
Quote:

Originally Posted by JadeEast /img/forum/go_quote.gif
I agree that mechanical transducers are the "weakest" part of the
playback chain due to the vast differences in frequency response,
distortion, resonance of the materials and interaction with differing
ear & head shapes. The differences between various headphones,
speakers or cartridges will probably always be more pronounced that
the differences between different amps.





Thanks, this is almost exactly what I wanted to post, so DITTO!

I heard long ago from some audiophile a maxim and have always pretty much followed it: spend most of your money in an audio system at the points where mechanical energy is transferred to an electrical signal, and then back again (at that time, the turntable and phono cartridge and the speakers; now, it's a bit more complicated, but the DAC is obviously important).

My first quality speakers were Klipsch Heresy models, which I lovingly hand-finished to save the cost of the factory finishes (they offered unfinished birch those days.) I drove them with a fairly inexpensive Yamaha integrate amp, and they sounded great! (I still have them). They represented about 50 per cent of the value of that whole audio system.

Yet, some people actually use some percentage like spend 20 per cent of their total budget for cables. I can't imagine spending hundreds of dollars for an interconnect (or more) or some speaker cable. That's just plain nuts, to me. And, I do think there are differences in amplifiers, especially for high end headphones, but it's clearly less than the difference between average and excellent headphones.

Just my two cents.
biggrin.gif
 
Jul 29, 2007 at 10:40 PM Post #43 of 790
Quote:

Originally Posted by slwiser /img/forum/go_quote.gif
How can a piece of equipment that reproduces sound perfectly go obsolete? Or how can a piece of equipment that has a distortion below hearing level get any better than below hearing and therefore go obsolete?



Most people in the USA today are obsessed with the latest and greatest, thanks in large measure to advertising and marketing which is designed to appeal on an emotional rather than an intellectual basis.

You make a good point ,perhaps I shouldn't have used the term obsolete but rather unpopular.

It wasn't too long ago that 5.1 was all the rage, now a lot of people are upgrading to 7.1 and 5.1 gear is starting to show up on the used market more and more.

I have both a TB Santa Cruz sound card and an Audiophile 2496 card, I honestly can't tell the difference between them reproduction wise when playing back WAV files ripped from my CD's.

Another point I would like to make about the HD650 spectrum analyzer plot is that not only are you dealing with one transducer in that plot but rather there are two transducers. There is the headphone which is a transducer and also the microphone which is a transducer. I honestly find it hard to believe that the combination of the two transducers had a total distortion level as low as was shown on that graph.

I rather suspect that the plot was taken at a very low , but not too low, volume, where both transducers are likely to be more linear than at high volumes. At levels which are too low, noise would become an issue, particularly with the microphone.
 
Jul 29, 2007 at 11:02 PM Post #44 of 790
Quote:

Originally Posted by Cor /img/forum/go_quote.gif
One thing that irks me is when people publish plots made using fourier transforms without specifying the window function. The window function can have a huge impact on how things appear. More specifically, it can have a huge impact on how sharp the peaks appear. Your screenshot shows that a Blackman Harris window is being used, which is good, but comparing it to a graph made with an unknown window function is going to be difficult at best, especially when it comes to a subjective comparison of how sharp the peaks are.


I have no control over what others may publish, I only know that which I do.

It was the best data I could provide and I tried to include all the relevant details.





Quote:

It might have been F. Alton Everest's book that I read, or it might have been another source, but I do recall reading a study on human memory and how it relates to sound. It's actually piss poor. We remember a lot of qualitative things about a sound, but when it comes to subtle differences most people can't even walk from one room to the next without forgetting. If you're chasing some golden sound from your memories, where you find it will have very little to do with what the original sound.


I have read this also, but didn't want to bring it because, like you I do not recall the source and knew that I would be challenged and found wanting.


Quote:

My own take on this subject is that the better an amp is, the less it should alter the original waveform other than to make it bigger. This means that the best amps *should* sound remarkably similar because they're coming so close to perfect reproduction. However, this is often not the case. You often hear expensive amps that "wow" you with some kind of difference. Why is this so? If you're selling an outrageously expensive amp you need to justify that expense. Any kind of "wow" you can produce therefore helps, even if it means deliberately introducing coloration.


I noticed a difference when I went from a basic simple Pioneer receiver in 1983 to a Harman Kardon unit that had a high current capability although its power rating was only a little higher than the Pioneer. The Pioneer unit had decent specs but it just didn't interface with the speakers as well as the HK unit.

FWIW, I watched that HK unit for months get advertised at $199 in a certain hifi store's ad, one week they screwed up and switched prices with something else that was $129, I looked carefully at the ad for disclaimers, found none, and the store sold it to me for the advertised price.

Quote:

Now, I should note here that I'm coming from the speaker side of things and am basically a neophyte when it comes to cans. However, high end speakers and amps are rife with "wow" factor, and I haven't seen anything that makes me think cans and can-amps are any different. People are still using the same ancient tube designs that, if carefully biased, pampered, and paired can *approach* the accuracy of a solid state system, but typically do not. Typically, they have at least a little euphonic distortion that gives them their distinctive sound. i.e. Their "wow". Again, I say there's nothing wrong with wanting a little "wow". Just don't try to delude yourself or others into thinking that you can have accuracy at the same time as your "wow".


Certain forms of distortion are more audible than others, I suspect that some are even pleasant for some people. That may be what is driving the claims here of one amp being more enjoyable than another.

Sound is a very subjective thing.. And certainly it is everyone's right to purchase whatever they want or feel the need to have.

All I'm really trying to do is throw a cautionary note, particularly to n00bs to audio, to take various claims with a hefty helping of salt. The n00b in particular is unlikely to be able to discern the difference in different amps or even DAC's and I think that to tell them to invest a lot of money right at the beginning is doing them a disservice.

It takes training and experience to critically listen, and even then, as you so rightly have pointed out, our memory for sounds is quite poor.

Quote:

This, naturally, brings us full circle to the OP's contention that the amp doesn't matter. He's right in that amps designed for accuracy tend not to matter. However, amps designed to add a little "wow" do make a difference. Some spend a long time looking for the "wow" that suits them. Those that have found it will never be convinced that amps make no difference precisely because they found amps that *do* make a difference. Convincing them that difference is actually a form of distortion is yet another uphill battle, but it's at least one you have a hope of winning, if only a very slim one.


I don't expect to "win", people think what they wish to think and most people are immune to change in ideas once they have them firmly set in their minds.

It's also true that I'm the same way, once I have reached a certain conclusion I'm unlikely to change it unless presented with very conclusive evidence, something which seems to be in short supply in the audio world.

Mostly I'm just trying to reach those who are new to audio and gullible enough to buy on the sole recommendation of others with whom they might not have much in common taste wise.
 
Jul 29, 2007 at 11:13 PM Post #45 of 790
Quote:

Originally Posted by Cor /img/forum/go_quote.gif
My own take on this subject is that the better an amp is, the less it should alter the original waveform other than to make it bigger. This means that the best amps *should* sound remarkably similar because they're coming so close to perfect reproduction. However, this is often not the case. You often hear expensive amps that "wow" you with some kind of difference. Why is this so? If you're selling an outrageously expensive amp you need to justify that expense. Any kind of "wow" you can produce therefore helps, even if it means deliberately introducing coloration.


No doubt about it. I look at many of the "state of the art" amps in the high-end and wonder what the designers were thinking. In quite a few cases I suspect marketing was dictating the terms to the engineers.

Quote:

Now, I should note here that I'm coming from the speaker side of things and am basically a neophyte when it comes to cans. However, high end speakers and amps are rife with "wow" factor, and I haven't seen anything that makes me think cans and can-amps are any different. People are still using the same ancient tube designs that, if carefully biased, pampered, and paired can *approach* the accuracy of a solid state system, but typically do not. Typically, they have at least a little euphonic distortion that gives them their distinctive sound. i.e. Their "wow". Again, I say there's nothing wrong with wanting a little "wow". Just don't try to delude yourself or others into thinking that you can have accuracy at the same time as your "wow".


Bingo, you are dead right with your assessment and I completely agree. Looking at many of the high-end speakers today from a design standpoint will reveal that they are full of fundamental flaws and amateur mistakes for which there are no excuses. A lot of the drivers used in these "statement products" have awful resonance peaks and frequency response anomolies which cannot be fixed with any crossover in existent. Attempting to flatten out the frequency reponse still leaves the energy storage & resonance problems which will manifest itself as glare & fatigue in longer listening sessions, in other words, anything longer than what's likely in a showroom. Of course, this neglects the sharp corners and lack of roundovers on the front baffle edges which will lead to lots of diffraction issues.

On the amplifier side, the story is much the same, especially with regards to tube amplifiers. There's literally been little if any progress since the 1950's, decades later and we're still recycling the same Williamson, Dynaco, and Acro in well over 90% of the amplifiers found in today's "high-end". We just have better "boutique parts" and maybe someone stuck a CCS or two on one of the stages, but the circuits remain essentially unchanged.

On the single-ended triode segment of the scene, it's the same story all over again. Almost every amp in existence is a 6SL7 grounded-cathode gain stage followed by a 6SN7 in a cathode-follower configuration which drives the 300B or whatever the power triode happens to be. 12AX7's and 12AU7's are often substituted for the octal tubes, but the topology remains the same. It's an inherently flawed circuit with many problems which has become ubiquitous in the high-end, and in fact many companies use such a circuit in the statement products.

Quote:

This, naturally, brings us full circle to the OP's contention that the amp doesn't matter. He's right in that amps designed for accuracy tend not to matter. However, amps designed to add a little "wow" do make a difference. Some spend a long time looking for the "wow" that suits them. Those that have found it will never be convinced that amps make no difference precisely because they found amps that *do* make a difference. Convincing them that difference is actually a form of distortion is yet another uphill battle, but it's at least one you have a hope of winning, if only a very slim one.


The problem of course is that designing an amp for "accuracy" is a lot harder than it is at first glance. Unfortunately, most audio designers aren't taking the right approach and rely far too much on massive amounts of negative feedback to achieve "accuracy". The correct approach is to aim for linearity right down at the device level; every transistor or tube should be as free from distortion as possible, then optimize the circuit to maximize the linearity of all the devices, then dial in negative feedback if needed.

Now, assuming a properly designed amp where accuracy is one of the main goals, you'd be correct to assume that amps will become fairly similar sounding as we climb the quality ladder and reach the high-end. For instance my single-ended 300B headphone amp doesn't sound all that different from my push-pull 50 amp. Different tubes, different topologies, yet the sound is remarkably similar.
 

Users who are viewing this thread

Back
Top