How many parameters would need to be tested so that two or more amplifiers with all results matching could be considered to be identical soundwise?
Looking at various audio test equipment manufacturers' websites, I come up with this quick list.
Gain
Sensitivity
Power
Frequency response
Bandwidth
Phase Shift
Input Impedance
Output impedance
Slew Rate
Distortion (THD,IMD,TID)
Distortion vs. Output
Signal to Noise Ratio
Crosstalk
Anything else?
and if an extensive data base of all measured amplifiers existed, would it even be possible to find two designs that match on all the tested parameters?
If so, Would those amps sound identical?
Looking at various audio test equipment manufacturers' websites, I come up with this quick list.
Gain
Sensitivity
Power
Frequency response
Bandwidth
Phase Shift
Input Impedance
Output impedance
Slew Rate
Distortion (THD,IMD,TID)
Distortion vs. Output
Signal to Noise Ratio
Crosstalk
Anything else?
and if an extensive data base of all measured amplifiers existed, would it even be possible to find two designs that match on all the tested parameters?
If so, Would those amps sound identical?
Rayma, Well, of all the possible respondents out there, I'm certainly glad the first one is from you ! I was a little worried the thread might need some flameproofing right from the get-go. I guess it may well yet.
My intent in posting comes after thinking about some of the less friendly threads I've seen here and I found myself wondering how it might be possible to bridge between techies and feelies (making it black and white) .
Leaving out those who just want to joust, riding a horse they've been out of control on for years, the chasm separating the the two camps is, in my view, largely comprised of misunderstanding by the more extreme positions in each camp, of the other's use of language.
I thought that if what could be agreed upon as every parameter could be tested and results matched, in all likelihood some differences in sound would be perceivable. In my optimistic imagination, the technically inclined would get interested in what new measurements could be devised and the sensory based participants would get the feeling the engineers were more interested in the truth than defending their knowledge pool and maybe take the trouble to study a little themselves.
My intent in posting comes after thinking about some of the less friendly threads I've seen here and I found myself wondering how it might be possible to bridge between techies and feelies (making it black and white) .
Leaving out those who just want to joust, riding a horse they've been out of control on for years, the chasm separating the the two camps is, in my view, largely comprised of misunderstanding by the more extreme positions in each camp, of the other's use of language.
I thought that if what could be agreed upon as every parameter could be tested and results matched, in all likelihood some differences in sound would be perceivable. In my optimistic imagination, the technically inclined would get interested in what new measurements could be devised and the sensory based participants would get the feeling the engineers were more interested in the truth than defending their knowledge pool and maybe take the trouble to study a little themselves.
Thomas C. Just to be clear. First of all, I don't consider your posts to be interfering, and secondly, just in case you think I was referring to you in the posts above somehow, I assure you I wasn't.
Rayma, Re: Production units. When I got my headphones (wow, 16?) years ago, after listening to some AKG's on the demo wall and deciding on the model, the sales staff went and got some unopened ones from the back.
Knowing the guys in the store, on impulse I opened the box and tried them on the set-up that had run the demos. Noticeably different. "Cuz the demos are broken in." So then "Humor me, you don't have any other customers in the store, let's try another new pair" Different again from the other new ones. In all I listened to three pairs and each was distinguishable. The sales guy though reluctant to even try at first, listened and agreed.
Since then, whatever the object, I always take what I try and like even if it's the demo. You can't assume that another one from the stock room will be exactly the same. or to put it another way, won't be different in ways that matter to you.
Rayma, Re: Production units. When I got my headphones (wow, 16?) years ago, after listening to some AKG's on the demo wall and deciding on the model, the sales staff went and got some unopened ones from the back.
Knowing the guys in the store, on impulse I opened the box and tried them on the set-up that had run the demos. Noticeably different. "Cuz the demos are broken in." So then "Humor me, you don't have any other customers in the store, let's try another new pair" Different again from the other new ones. In all I listened to three pairs and each was distinguishable. The sales guy though reluctant to even try at first, listened and agreed.
Since then, whatever the object, I always take what I try and like even if it's the demo. You can't assume that another one from the stock room will be exactly the same. or to put it another way, won't be different in ways that matter to you.
Industry people have talked about this for many years. Even Bill Johnson brought this up unprompted
while we were privately talking back in the 70s.
All the painstaking equipment reviews are kind of useless if only one unit is tested. In fact, it is well known
that some mfrs cherry pick an especially good sounding production unit to send to reviewers.
This is why more savvy reviewers buy equipment to be reviewed from dealers, and under an alias,
so the dealer won't cherry pick a unit on behalf of the mfr.
while we were privately talking back in the 70s.
All the painstaking equipment reviews are kind of useless if only one unit is tested. In fact, it is well known
that some mfrs cherry pick an especially good sounding production unit to send to reviewers.
This is why more savvy reviewers buy equipment to be reviewed from dealers, and under an alias,
so the dealer won't cherry pick a unit on behalf of the mfr.
Last edited:
As far as transducers go, I tried hundreds of units (of about ten brands) of LS3/5A speakers. They all varied
quite noticeably, particularly from certain mfrs. I own the best pair that I've ever tested, and they happened
to be serial numbers 7 and 8. I won't say which mfr though.
Last edited:
Do you know this for a fact or are you speculating? The amount of effort required to find a golden unit would be significant to say the least. I also highly doubt you'll find any difference in sound between two identical pieces of gear - at least not it the solid state world. Maybe an SET tube amp would require hand-selection, but then it's just a matter of selecting a handful of golden tubes.
Another issue with subjective reviews is that it's a pay-to-play program. The manufacturer provides the equipment for review on an indefinite loan to the reviewer. No, no. It's not a gift. That would be a bribe. It's an "extended loan for as long as the reviewer would like". A friend of mine used to review for one of the British hifi magazines and has stacks of gear to show for it.
There's also the issue of manufacturers indirectly paying for good reviews. A manufacturer with a 6-figure ad spend with a HiFi magazine will get good reviews in that magazine. Period.
Now, not all reviewers are like that. Many of the influencer/youtuber types will ship the gear back if you pay the shipping cost. That's more reasonable. But I'm not clear on what kind of ROI a manufacturer can expect from a youtuber/influencer, if any.
As for which measurements are relevant, I'd say: Most of them. One of my personal favourites is the multi-tone IMD test. But I think it's a mistake to look at just one measurement in isolation. In my view, HiFi equipment should perform well across a wide range of measurements.
There's a considerable body of scientific evidence which shows that a wide range of listers (audiophiles, college students, musicians, trained listeners, etc.) prefer equipment with good measurements (plural) in blind listening trials.
Similarly, there's considerable evidence to show that the same listeners prefer equipment that looks good and that has a high price tag in sighted trials. It turns out that audiophiles are human (even though it sometimes seems hard to believe...
).See the work by Olive and Toole at Harman Kardon for example.
I think it would be naive to ignore that the listener is generally not a participant in a blind trial, but rather a human in a listening space of some sort. Thus, I think it would be wise to take look and feel into account. But similarly, I also believe we should take the measurements into account. Measurements can be a good way to separate the wheat from the chaff. If you're selling a PCM1794 DAC, I expect you to match the performance shown in the TI data sheet. If you don't, I will be disappointed. If you further claim that you don't meet the specs because "measurements aren't everything" and that your DAC sounds "extra flappy-dappy" (or whatever audiophile verbiage d'jour) I'm going to shop elsewhere. But that's me. Others may have different preferences. That's OK.
Tom
Last edited:
I don't have any trouble at all with what you're saying, nor with the evidence you mention.
What prompted my thinking was reflecting on the endless unnecessary argument in threads that leave everybody grumpy and accomplish nothing new. My hope was that just considering it might lead to little less absolute certainty and a little more interest in the other camp's perception and cognition.
It was all a mental exercise on my part. Could there be a battery of tests that those on both sides of the fence can agree is comprehensive, and then would amplifiers perfectly matched under such a regimen sound exactly the same?
What I'm trying to get at is that extra step that gets taken in debate. Measuring amplifiers and then finding that listeners prefer the ones that measured well shouldn't lead to thinking we were measuring the listening process.
I did a test on a Danish forum where people had to hear the difference between 4 to 6 pieces of music, which were all the same music, but had been through different things in the signal path.
They knew nothing and had only the opportunity to use their ears and therefore were not affected by anything else.
What was put into the signal path was:
5 inch copper tape or 4 inch aluminum strips or a capacitor for $ 7, or a capacitor for $ 2000, or 5 inch solder.
And since the participants could not identify which was the original file between the others and especially since no difference between 7 $ and 2000 $ capacitors could be heard or see the difference in audacity, there was a minor war. The blood rose directly to the heads of some of the participants who came up with the worst apologies I have ever heard.
The point of this "story" is that it is extremely limited what you are capable of when you are only allowed to use your ears and are not affected by knowing what change has been made, what device is playing and it is absolutely crazy what comes of excuses as to why the differences are not heard.
They knew nothing and had only the opportunity to use their ears and therefore were not affected by anything else.
What was put into the signal path was:
5 inch copper tape or 4 inch aluminum strips or a capacitor for $ 7, or a capacitor for $ 2000, or 5 inch solder.
And since the participants could not identify which was the original file between the others and especially since no difference between 7 $ and 2000 $ capacitors could be heard or see the difference in audacity, there was a minor war. The blood rose directly to the heads of some of the participants who came up with the worst apologies I have ever heard.
The point of this "story" is that it is extremely limited what you are capable of when you are only allowed to use your ears and are not affected by knowing what change has been made, what device is playing and it is absolutely crazy what comes of excuses as to why the differences are not heard.
Hi Tom, It was Pano, The thread is linked below.
Hi Thomas C,
Yah I hear ya. It was fun to read though. (Did you see this thread?)
On the other hand, if room can't be left for the possibility of an outlier, the logic might have a problem or contradiction somewhere.
An example: Frequently in conversations about comparative listening tests it's argued that auditory memory is not good enough for someone to have clear memory of the sound they were listening to a few minutes before. Yet nobody denies the existence of absolute or perfect pitch, which is memory, and if right on pitch it's very accurate memory.
.
Hi Thomas C,
Yah I hear ya. It was fun to read though. (Did you see this thread?)
On the other hand, if room can't be left for the possibility of an outlier, the logic might have a problem or contradiction somewhere.
An example: Frequently in conversations about comparative listening tests it's argued that auditory memory is not good enough for someone to have clear memory of the sound they were listening to a few minutes before. Yet nobody denies the existence of absolute or perfect pitch, which is memory, and if right on pitch it's very accurate memory.
.
You're assuming that the outlier will be "outlying" enough to sound better. In my experience, that is exceptionally unlikely - at least for well-designed solid state equipment. I've measured 100 headphone amps under my Tom Christiansen Audio brand. They measured basically line-on-line. My designs are up against the boundaries of physics and I use components with tight tolerances where it matters. You simply won't find an amp that's 20 dB better than the rest. It doesn't exist. You may find one that's 0.1 dB better than the rest. If you can show data to indicate that humans can detect such minute differences reliably, I'm willing to believe your argument. And by "data" I mean a peer-reviewed paper, not "so-and-so-on-the-internet-said..."
Tube circuits are different. You could definitely find an outlier SET amp, for example. I've also measured the THD of a 300B when fresh out of the box and a few hours later. The THD was better after a few hours. Burn-in is a thing ... at least for those specific tubes. That doesn't mean that any and all equipment will improve after burn-in.
You're making the assumption that pitch detection relies on the same memory systems as the auditory memory. Based on my knowledge of neuroscience and psychology, I would be very surprised if that was the case.
Tom
Last edited:
@Hearinspace........
Unfortunately, only very few people (mainly cassical musicians) are able to do "perfect pitch" or "absolute pitch" (as we call it here). It´s where
you ask someone to ex. sing an A, and they just do it. Or maybe I misunderstood your definition of "perfect pitch"??
Try (if you have the possibility) the test we did some 50 years back, to determine, how good the audible memory actually was.
Sit in front of your speakers, listen to well known music at a normal level with bass and treble settings in "neutral".
Then pull up the bass and treble to max. This sounds exessive at first, but sit and listen for 5 minutes, and your ears will get used to the sound. Then.... after 5 minutes put the bass and treble back to neutral. Now you´ll find the sound really thin and anemic. Then pull the bass and treble-setting to minimum. Even more anemic, but have patience and listen for 5 minutes. Your ears will slowly get used to the sound. Then pull the bass and treble back to neutral, and you will have close to the same experience as you first had, when you went from neutral to max.
5 minutes...... that´s all it takes
Last edited:
- Status
- Not open for further replies.