Well, well, well ...
Have just found this: Just How "Absolute" Is Recorded Sound? | Stereophile.com. This tells me a lot, and much of it has little to do with the intention of the presentation. Listened on normal PC speakers, on the HD setting - I'll be interested in what other people think ...
Have just found this: Just How "Absolute" Is Recorded Sound? | Stereophile.com. This tells me a lot, and much of it has little to do with the intention of the presentation. Listened on normal PC speakers, on the HD setting - I'll be interested in what other people think ...
While nicely thought out, no news in what JA says, at least to me. Audio is an illusion, always has been, no matter how enjoyable it is - and it is!
Our proverbial quest for the absolute sound is nonsensical in absolute terms, the best we can hope for is a life-LIKE sound, something fairly close to the absolute.
Our proverbial quest for the absolute sound is nonsensical in absolute terms, the best we can hope for is a life-LIKE sound, something fairly close to the absolute.
The existence of an original reference isn't dependent on imperfect elements in the record/playback chain. JA expresses simple audio defeatism. I stopped reading those magazines when his perspective became dominant. Pearson wasn't the post card picture of stability but at least he pursued an ideal.
'They're all professional, high quality mics'. High comedy. I have that recording. The RE20 and SM58 are designed with colourations as a feature. They're brushes, not cameras.
Thanks for the thoughts ... what struck me was the poor quality of the playback of the system used for the demonstration - all the symptoms of defective audio playback were so obvious in this, came through loud and clear even though "only" a YouTube clip. Considering who the demonstrator was, the supposed quality of the equipment being used, and the efforts made to tweak the sound in the usual audiophile way - note the supports under the laptop! - it still stunk of all the "badnesses" of conventional audio ...
First the cowbell: the real thing sounded real, the copy was badly stunted, the dynamics of the strike were severely compressed.
The voice playback was riddled with silibance, and chestiness, depending on the mic; the demos of enhanced airiness through manipulation all sounded as bad as each other, none worth listening to unless one was paid to, the repeated drum sound in the large acoustic was scarily bad, it was all uniformly dreadful ...
The one piece at the end, showing good signs of quality at low volumes, collapsed as soon as any real dynamics were called for, compressed and blurred badly ...
If this is the best that people in the business use for showing the "subtleties" of sound reproduction, in 2013, heaven help us ...
First the cowbell: the real thing sounded real, the copy was badly stunted, the dynamics of the strike were severely compressed.
The voice playback was riddled with silibance, and chestiness, depending on the mic; the demos of enhanced airiness through manipulation all sounded as bad as each other, none worth listening to unless one was paid to, the repeated drum sound in the large acoustic was scarily bad, it was all uniformly dreadful ...
The one piece at the end, showing good signs of quality at low volumes, collapsed as soon as any real dynamics were called for, compressed and blurred badly ...
If this is the best that people in the business use for showing the "subtleties" of sound reproduction, in 2013, heaven help us ...
Exactly what I have been saying for about two decades now.
While I understand and accept that any replay chain, no matter how good it is, is no more than an illusion of the live sound, I have heard a few in my day which were truly excellent. The one on the clip, even takinh into account the inevitable losses from the clip itself, is below par, to my ears. The humbler version of my system will outshine it easily ("humbler" meaning a much more down to earth priced amp rather than the very costly Karan).
While I do have respect for JA, that particular setup wasn't much to speak of, and I don't give a damn how much it cost, or what exalted components might have been used. My ears don't take into account brand names and/or prices.
This is consistent with my own comperatively humble experience of the last 20 years or so. I find that only CD players have made some progress, while overall, I feel the audio as a field is actually declining in quality. I appreciate the problems of selling one's products in a field which is, at best, No.3 on many top lists, computers and video (DVD and BD) take the first two places, and the third is shared between stereo and HT and that classic stereo is very possibly heading towards place No.4 - but I feel a declining sound quality is the wrong response.
I realize some concessions to the new times must be made, such as adding a remote control, but I cannot shake the feeling that classic stereo gear is generally overpriced.
On the other hand, the overall quality of stereo buff has also changed, and for the much worse. Today, people will believe anything they read on the Internet. In our days, we had to wait a month for the next issue of whatever, so we had a moth to digest and mull over the info supplied. Above all, to talk it over with friends. No Internet also meant incomparably more persoanl contact with other audiophiles, more visiting, more listening to other people's systems and consequently much more personal experience.
My point is, the changing social circumstances have also changed the users, and both together have made it possible to slowly reduce the quality of audio gear and shift it much more into percieved quality waters.
This in turn caused the vast majority of what were once medium priced companies to either lower their prices (very few), or to make a move towards higher price tiers (the vast majority), with inevitable casulaties along the way dropping out altogether (Sansui, Akai, JVC, Technics, etc). Those who survived responded by having at least two product lines, one for the masses, and another for the rich (e.g. Marantz).
Frank, it's a very complex topic.
High levels of distortion are apparently inaudible or tolerable for most people, but what if the character of the distortion changes with signal e.g. as it gets louder? I presume that in many systems, it's the speakers that perhaps 'give the game away' as the cones break up or reach their displacement limits. I guess that full range and two-way are worse offenders than three. Passive worse than active.
So is one important element of sound quality the ability of the system to maintain the same characteristic regardless of the input i.e. as the signal gets louder, the system maintains constant gain, and distortion stays low? Or should that merely be that the nature of the distortion doesn't change appreciably as the signal gets louder?
Listening to my car stereo last night (Radio 3 - classical), with 70mph background noise and a dodgy FM signal, I was still struck by how good it sounded. It occurred to me that if the listener simply can't or won't use genuinely 'hi fi' speakers, for whatever reason, could there be something to be said for 'pre-distortion' i.e. not the usual meaning, but simply adding a level of background noise and distortion to the signal that masks the inevitable changes in characteristic as the system is pushed harder? Could that person do this subliminally through his choice of source and amplification perhaps?
So is one important element of sound quality the ability of the system to maintain the same characteristic regardless of the input i.e. as the signal gets louder, the system maintains constant gain, and distortion stays low? Or should that merely be that the nature of the distortion doesn't change appreciably as the signal gets louder?
Listening to my car stereo last night (Radio 3 - classical), with 70mph background noise and a dodgy FM signal, I was still struck by how good it sounded. It occurred to me that if the listener simply can't or won't use genuinely 'hi fi' speakers, for whatever reason, could there be something to be said for 'pre-distortion' i.e. not the usual meaning, but simply adding a level of background noise and distortion to the signal that masks the inevitable changes in characteristic as the system is pushed harder? Could that person do this subliminally through his choice of source and amplification perhaps?
Interesting thoughts ...
IME, it's the amps that give way, not the speakers. This is counter-intuitive, but I've experienced this over and over again: if I improve the ability of the amplifier to go loud without stress then the speakers are always able to keep up. I've been able to get relatively mediocre speakers to run at subjectively very loud levels without audible problems, or marked deterioration of quality.
Subjectively, the distortion should always remain low, and will not appear to vary either as you up the volume, or the musical signal itself increases in volume. In the aforementioned clip, it's obvious, at least to me, how the quality degrades badly as intensity of the music increases -- this should not happen ...
Simple sources of music can work remarkably well, because they are a unified design. So, yes, a car radio at times can sound very together -- now, imagine that you can push that car radio harder and harder, increase the SPLs without any apparent change of characteristic, to the point of clipping: this is what one should be after in a system. It is achievable, and doesn't require any masking to happen - it's all about engineering a truly cleanly working system.
IME, it's the amps that give way, not the speakers. This is counter-intuitive, but I've experienced this over and over again: if I improve the ability of the amplifier to go loud without stress then the speakers are always able to keep up. I've been able to get relatively mediocre speakers to run at subjectively very loud levels without audible problems, or marked deterioration of quality.
Subjectively, the distortion should always remain low, and will not appear to vary either as you up the volume, or the musical signal itself increases in volume. In the aforementioned clip, it's obvious, at least to me, how the quality degrades badly as intensity of the music increases -- this should not happen ...
Simple sources of music can work remarkably well, because they are a unified design. So, yes, a car radio at times can sound very together -- now, imagine that you can push that car radio harder and harder, increase the SPLs without any apparent change of characteristic, to the point of clipping: this is what one should be after in a system. It is achievable, and doesn't require any masking to happen - it's all about engineering a truly cleanly working system.
Yes, it could also be the amps that 'give the game away'. In my passive days I found that some pretty decent Tannoy 2.5-way speakers which I had, sounded good at low volumes, but distinctly different at higher levels which, at the time, I perceived as 'stress' which was not very relaxing to listen to. I perceived the sound being forced out under pressure. I experimentally "activated" them as 3-ways, and they sounded much easier on the ear. Maybe it was going from 2.5 to 3 way, the lack of passive crossovers in the way, steeper DSP filters, or the fact that the (now multiple) amplifiers were working less hard. The upshot being, however, that they didn't exhibit audible 'stress' at higher levels. However, if the source had had higher levels of distortion to begin with, might I have perceived them differently?
The direction I'm heading in, is to suggest that people may subconsciously choose relatively noisy and distorted sources such as vinyl in order to mask the inadequacies of the rest of the system... And that with our tolerance of constant distortion, it sounds absolutely fine. Changing to a less distorted source, however, e.g. digital, the modulating nature of the amp/speaker's distortion, although low, becomes audible and is registered as something decidedly unpleasant, meaning that a person with a passion for valve amps and passive speakers may find digital genuinely atrocious - his "reference system" that sounds great with vinyl tells him so.
Possible?
Yes, I appreciate that distortion levels of the drivers will always increase, because of the mechanical nature of them. And what typically are the distortion characteristics of decent amplifiers. However, what I believe is relevant is the type of distortion that occurs, subjectively this is critical - simply put, it's the old low order vs. high order harmonics thing; all the tests of drivers I've seen show dramatic drop offs of distortion levels beyond the 3rd's - whereas an amplifier can easily generate much higher, and audibly nastier harmonics when they misbehave.
I've always found it an iterative process, as you're suggesting: if the amp/speaker setup improves, it might then highlight more clearly that the source does have a problem. So, then I redirect my attention to that, and continue in a step-wise fashion.
All these things are possible, and I've personally gone through some of these stages of thinking, many times. What I have found, over and over again, is that if I persisted, telling myself that the system is sounding bad at the moment because there is a problem somewhere that I haven't tackled, that I don't understand well enough, that I 'don't give up', then I always manage to push forward, achieve better understanding, and get better sound.
It's been fascinating helping out a friend who has both vinyl and digital - he's an enthusiastic tweaker, especially of the vinyl - and the 'battle' between the 2 media. It's ebbed and flowed, first one is better, then the other - the distortion styles are very different: he tends to feel that vinyl distortion is 'nicer', and on his system I would agree.
I can understand that some people may prefer heavy masking by some means, of distortion elsewhere, but it's not a good direction to go towards, it just boxes you in. Not the best choice really ...
Be careful here, higher driver harmonics rapidly go beyond the useful upper limit of the driver. High order intermodulation may still exist in the driver pass band
Correct. This has all been investigated subjectively, having ready access to accurate instrumentation most likely would have helped in some areas, hindered in others. My progress forward has mostly been based on empirical findings, largely because that gave me the results. The magazines made it clear that other people were generally not impressed by components designed by people using relatively conventional approaches, and my personal experience with systems using technically "correct" items didn't push any of the right buttons. So, whatever works ...
Before delving deeper, I would like to draw your attention to the problem of amp settling times. This is often overlooked, as opposed to rise times, but I maintain that settling time is in fact more important than rise time, assuming that the rise time is say 2 uS or less.
I have made a few strictly test circuits with various version, and the best by a long shot was the one I saw and stole as an idea from the German LAS power amp.
Obviously, if the amp is not up to the job, the speakers won't be either. I find that much of the stress attributed to SS amps is precisely this problem, they have inordinately long settling times, which may also be rather messy in terms of sound (cluding, loss of details, even loss of overall focus, etc).
I do not think this is the sole important factor, of course, it's just one of many aspects which must be dealt with, but unfortunately, many manufacturers simply don't bother for whatever reason.
I have made a few strictly test circuits with various version, and the best by a long shot was the one I saw and stole as an idea from the German LAS power amp.
Obviously, if the amp is not up to the job, the speakers won't be either. I find that much of the stress attributed to SS amps is precisely this problem, they have inordinately long settling times, which may also be rather messy in terms of sound (cluding, loss of details, even loss of overall focus, etc).
I do not think this is the sole important factor, of course, it's just one of many aspects which must be dealt with, but unfortunately, many manufacturers simply don't bother for whatever reason.
Optimising "settling time" was advocated by Linsley-Hood as long ago as 1975. At the time this parameter was recognised as a defining one in the performance of op.amps - apparently unnoticed in the audio industry which had established a standard for itself of achieving distortion figures of 0.01 % at 20 kHz.
It had been supposed that reducing distortion by a factor of 10 in comparison to valve/tube designs would overcome any subjective deficiencies and objections to the so-called "transistor sound".
It would have been more revealing to display the shape of the waveform into 8R//2 u.F which where the differences between input and output would be plain enough.
Nevertheless statements invite fewer questions than pictures and a better impression is possible from a few words and figures in a product brochure.
Such would have been helping sales and everyone would have been doing it. So why change a winning strategy?
Hood devised a % corelation of input and output waveforms in 1970.
If it was ever used as a design tool - no doubt reasons of commercial sensitivity would have prevented any public acknowledgement or endorsement.
Michael J
Last edited:
Thank you for the insight, Michael, good to know, I like such tid-bits of audio history.
For years in the mid-90ies, I said and demonstrated that Analag Devices' op amps generally tened to sound better than the competiton mostly because their settling time was literally hundreds of times shorter than their competitors' (45 nS compared to typical 2,000 nS). That's not the only reason, of course, but I believe it's the key one.
In general audio, I know that Keith Johnson of Spectral is a strong advocate of settling times, or at least the loudest one who has come out of the closet.
There was a short discussion on the topic here as well some time ago, and as I recall, most agreed settling time was an important factor not to be overlooked. Personally, I believe the same, it should be as short as possible, and in my mind, it's more important than having a wild slew rate; as JC put it, make it to 40 V/uS and you're home and dry.
For years in the mid-90ies, I said and demonstrated that Analag Devices' op amps generally tened to sound better than the competiton mostly because their settling time was literally hundreds of times shorter than their competitors' (45 nS compared to typical 2,000 nS). That's not the only reason, of course, but I believe it's the key one.
In general audio, I know that Keith Johnson of Spectral is a strong advocate of settling times, or at least the loudest one who has come out of the closet.
There was a short discussion on the topic here as well some time ago, and as I recall, most agreed settling time was an important factor not to be overlooked. Personally, I believe the same, it should be as short as possible, and in my mind, it's more important than having a wild slew rate; as JC put it, make it to 40 V/uS and you're home and dry.
Rgerading the previous discussion on stress in audio systems, my feeling is that unless there was a gross mismatch in pairing the components, most of it will be found in the power amplifier section.
Consider: asking a 6 or 7 inch bass/mid driver to reproduce concert level sound pressure levels in a large room, and from a dinky little enclosure, begs the question why aren't you using stronger therapy.
Assuming no such boo-boo was made, and assuming your spekares don't present a really evil load to the amp (like dropping down to say 2 Ohms with a say -60 degree phase shift), it will be the amp which is holding back.
Exactly how and why is matter of case by case examination. In my view, it's mostly because of poor (in respect to what's being asked of it) power supplies, both in terms of regulation and absolute voltages. Thus, in my view, the key design line should be separating the VAS power supply lines from the power stages, and by all means, do let the VAS power lines run at 7 or 8 V higher than the CAS power lines, but this by no means lets you off the hook for making CAS supplies stiff.
Obviously, this becomes cumbersome and expensive very fast, but hey, if you want to drive a Rolls or a Ferrari, you have to lay out some serious cash. As the saying goes, no such thing as a free lunch.
Consider: asking a 6 or 7 inch bass/mid driver to reproduce concert level sound pressure levels in a large room, and from a dinky little enclosure, begs the question why aren't you using stronger therapy.
Assuming no such boo-boo was made, and assuming your spekares don't present a really evil load to the amp (like dropping down to say 2 Ohms with a say -60 degree phase shift), it will be the amp which is holding back.
Exactly how and why is matter of case by case examination. In my view, it's mostly because of poor (in respect to what's being asked of it) power supplies, both in terms of regulation and absolute voltages. Thus, in my view, the key design line should be separating the VAS power supply lines from the power stages, and by all means, do let the VAS power lines run at 7 or 8 V higher than the CAS power lines, but this by no means lets you off the hook for making CAS supplies stiff.
Obviously, this becomes cumbersome and expensive very fast, but hey, if you want to drive a Rolls or a Ferrari, you have to lay out some serious cash. As the saying goes, no such thing as a free lunch.
amplifierguru, who used to be a member here, understood a hell of a lot of things that really matter. He believed very strongly in effectively giving the VAS very stable rails, so that the amp had superb PSRR. His rather unappetising looking amp in the 80's in Australia was the first commercial unit I heard that made me snap to attention. The Krells of the day, and such-like were pretty dismal in comparison ...
- Status
- Not open for further replies.