If you are going to design an amplifier then you need some numbers to put into the specification, and to test against. If you try to ignore numbers and instead just play around with circuits (possibly poorly understood circuits) until it sounds nice to you then you are definitely not pursuing hi-fi. You may be pursuing 'high end audio' but that has little to do with hi-fi. You may have a successful business, and may try to claim that as some sort of endorsement for your wisdom - but it isn't hi-fi (unless you get there by a happy accident - which is unlikely).
Nobody is imposing a definition of anything on anyone. You just have to accept that audio is a branch of engineering, and engineering deals in numbers. These numbers are not dreamt up by evil engineers, but are the result of careful listening tests. The entirely sensible aim of the tests was to determine which numbers matter, and what their values should be. Subsequent tests all seem to confirm that we have roughly the right numbers and roughly the right values, as equipment which the numbers say should be indistinguishable from each other (because they are all indistinguishable from the original) turn out to be indistinguishable when peeking is not allowed.
Nobody is imposing a definition of anything on anyone. You just have to accept that audio is a branch of engineering, and engineering deals in numbers. These numbers are not dreamt up by evil engineers, but are the result of careful listening tests. The entirely sensible aim of the tests was to determine which numbers matter, and what their values should be. Subsequent tests all seem to confirm that we have roughly the right numbers and roughly the right values, as equipment which the numbers say should be indistinguishable from each other (because they are all indistinguishable from the original) turn out to be indistinguishable when peeking is not allowed.
Although, the exact numbers may vary somewhat if we are talking about specially trained listeners, or statistical outliers.
Heh, still need some numbers though. All the right numbers, but not necessarily in the right order. (With thanks to Eric Morecambe)
And I understand the the need for numbers in design and calculation, but I thought we were including WHAT high definition is, not just how to quantify it. And my interest in it goes beyond counting.
I thought what it is is plain: a high degree of fidelity, so it sounds like the original. How difficult is that to comprehend? As soon as you start asking what that means you unavoidably encounter numbers.
When I designed my first BJT class AB power amps, my measuring equipment was quite restricted. To get an idea of crossover distortion without elaborated equipment, I fed the amp with a 10Hz sinewave, RC-filtered it to reduce signal distortion and connected the amp output to a dummy load resistor parallelled by a piezo tweeter quite common these days.
The piezo had an extreme sharp low frequency rolloff, i.e. worked as a perfect low frequency filter and what you heard was noise and distortion. Training the ears to hear that squizzling overtones was was simple by comparing the sound with and w/o the dummy load. I was able to hear the reduced distortion with increasing bias current - and it never totally vanished.
I presume THD would have measured in the range of 0.1% - not too bad. On the other hand I am convinced with a "normal" speaker and "normal" Signal I never could distinguish different bias settings at all - simply by the psycho-acoustic masking effect.
So I think that THD is a useful number to describe linearity of an amp, but this aspect often is vastly over-estimated imho.
The piezo had an extreme sharp low frequency rolloff, i.e. worked as a perfect low frequency filter and what you heard was noise and distortion. Training the ears to hear that squizzling overtones was was simple by comparing the sound with and w/o the dummy load. I was able to hear the reduced distortion with increasing bias current - and it never totally vanished.
I presume THD would have measured in the range of 0.1% - not too bad. On the other hand I am convinced with a "normal" speaker and "normal" Signal I never could distinguish different bias settings at all - simply by the psycho-acoustic masking effect.
So I think that THD is a useful number to describe linearity of an amp, but this aspect often is vastly over-estimated imho.
Last edited:
Nope. Just look at some of the measurements that stereophile publishes for ultraexpensive equipment. It's clear what few numbers these manufacturers put on their marketing are plucked from the air rather than some careful design goal. But a pet theory combined with some mumbojumbo always gets a great write up even if JA scratches his head over how something can intentionally be so bad.
The classic, the multiple $100,000 price tag only makes it more interesting.
Wavac SH-833 monoblock power amplifier Measurements | Stereophile.com
Last edited:
That's pretty impressive all right, but I'm not sure the numbers and graphs were plucked from the air, exactly. I'm afraid they might be telling the truth.
Taking your words literally that would mean the specs and graphs are fake or made up. Rather I see they simply ignore common wisdom and leave out or ignore the telling graphs/data and substitute their "story". Terms like apodizing sound sexey, but when it simply means you allow deliberate image leakage which is easily measurable, let's just stick with the words.
How about Audio Note CDT One/II CD transport & DAC 2.1x Signature D/A processor Measurements | Stereophile.com . A $300 marantz does textbook redbook.
I don't disagree about the words. It's just that Bill's comment intrigued me, and I didn't know if the part about the numbers was a little bit of entertaining embellishment or a real issue with audiophile equipment. I still don't know.
Bill, accidental cross post. Let me take a look at that.
Bill, accidental cross post. Let me take a look at that.
How many folks were fleeced to the tune of $10,000? We just can't smarten up the chumps.
There is a real issue with boutique equipment that it is designed by ear, often by people with no actual engineering or technical training (ever heard of David Manley?). Roberts lying on the floor glomming bits in until it sounds right is an example of boutique design, but he doesn't charge a silly price and has gold plated customer service.
In extremis there is no spec, and no real measurement going on, just markeing. YBA is an interesting one as the designer is an EE prof, but some of his ideas are right in the batsh*t insane camp.
The real skill in high end is the sales patter to precondition the reviewing chimp to like the idea of it before he plugs it in!
In extremis there is no spec, and no real measurement going on, just markeing. YBA is an interesting one as the designer is an EE prof, but some of his ideas are right in the batsh*t insane camp.
The real skill in high end is the sales patter to precondition the reviewing chimp to like the idea of it before he plugs it in!
Member
Joined 2009
Paid Member
Okay, that looks pretty bad in total.
However, people have legally been fleeced out of much, much more. Including by some entities which may be off topic for the forum.
Anyway, in the case cited, it looked like the measurement guy was trying to clue people in to be leery. This is really no different from many niche technology magazines trying to stay afloat, and which need advertiser dollars to do it. Usually, there are lightly veiled clues somewhere near the end of an article hinting at what they really think, but not quite so overt as to kill off the magazine business entirely. I don't think its unique to Stereophile or audiophile publications in particular.
Is there anything you would like to see done about it, or do you just want to express disgust?
Mark: People have been showing disgust about poorly measuring high end equipment with silly price tags since subjective review mags started. JA for all his faults measures everything and posts the measurements for us to see and compare. Despite being the editor he never goes back to the reviewer to make them re-write when something is truly dire.
Scott has been doing his own thing about it for (mumble) years in the circuits he has posted on his own and with others like Walt Jung as well as designing some awesome opamps that just happen to be usable in audio as well as mind boggling physics. You should try one of his mic preamps one day.
Scott has been doing his own thing about it for (mumble) years in the circuits he has posted on his own and with others like Walt Jung as well as designing some awesome opamps that just happen to be usable in audio as well as mind boggling physics. You should try one of his mic preamps one day.
- Status
- Not open for further replies.