Bob Carver was the real deal back in the day, his designs were based on no-bullshit performance objectives.
Audio Science Review (audiosciencereview.com) is an interesting counterpoint to all this huffing and puffing about what sounds "good" that has historically held the audiophile industry hostage. The guy has some state of the art test equipment and thoroughly tests equipment that people send in to be tested. Some of the results are really suprising, like a $14K DAC that was, from a performance standpoint, just garbage. He'll test almost anything, including the "audiophile network switch" that (surprise!) didn't do anything. If you're into this space his tests are almost always interesting.
What's so funny is that Carver came out with the compatible amplifier as a product. But it didn't sell.
So then he build the Carver Silver 7, as something of a joke.[1] It has fourteen tubes per channel. The power supply is on a separate chassis. All wire is silver, soldered with silver solder. It costs $17,500. Per channel. Here's a picture of a pair of them.[2]
It sold quite well. Versions of it are still selling.
Stereophile magazine has never been known for science, mostly pseudo-scientific, unsubstantiable claims. They still insist that double-blind testing is not suitable for audio. I used to subscribe to the magazine, but would get infuriated at the "reviews," especially at the unabashed glowing reports of what is obvious snake oil.
OTOH, I have owned a Carver magnetic field power amplifier, and I liked it very much. Bob Carver is a audio genius.
That's at least a step forward from 1985, it would seem:
> We made no effort to do A/B testing, since we feel it does not replicate normal listening conditions, and there is still insubstantial evidence that A/B testing reveals small differences as well as does prolonged listening to each unit under test.
So, in 1985, not even non-blind A/B testing! These people are delusional.
Bob is completely an audio genius: I have his Sunfire HRS monitors and one of the super cool tiny Carver 400 cubes and I can literally peel the paint off my walls with stuff smaller than your average boombox. #gameon
My fraternity in college (an engineering school) had pretty much all Carver amps. The folks that picked them were electrical engineers focusing on audio, and beside Marshall Leach, Carver was considered a god. I vividly remember the Carver Powercube could rattle windows halfway across campus.
They should re-do the test with recent PA amps from e.g. Crown or QSC. Those have now reached ridiculous levels of technological sophistication. They're light, powerful, indestructible, and can carry any load you plug into them. And in a blind test with good speakers they will sound better than anything "high end". Oh and they're almost always under $1K, with brownout inducing 2KW+ units below $2K.
If it's high fidelity sound reproduction you're after (rather than just "pleasant" coloration), today's upscale studio and live gear can't be beat IMO.
I believe, as the article is implying, that much of what stereophiles perceive as quality is merely signal color of one sort or another, once you reach beyond the basics of clean signal paths and input sources.
I worked my way though college by sound engineering/DJing in a major european city. The signal paths in studios at that time were complex meandering things all adding minor or not-so-minor levels of color, and now can all be substantially replaced by something like a "Universal Audio" DSP box, which is capable of reproducing 99% of the sonic/tonal qualities of the classic equipment. I can't, however, shake the assertion that old, analog, recordings have a certain quality and warmth to them that to my ears sounds amazing, and we have moved so far from this with crappy compression algorithms and equipment which mostly does not match 90's HiFi, unless you spend a lot. Particularly, anything processed at CD quality, seems tinny to my ears. Recording at 192KHz seems to be set to overcome this however.
1) 192kHz is almost entirely snake oil. It uses 4x the CPU cycles and 4x the storage capacity for no measurable difference when it comes to human hearing.
2) 96kHz vs 48kHz does allow for some small but probably worthwhile improvement in the behavior of the anti-aliasing filters used in the analog->digital converters.
3) I have a theory that has been shot down by some highly respected DSP people but I retain it nevertheless. In Ye Olde Analogue domain, most of the distortion that occurs will be somehow related to harmonics, because of the nature of the analog electrical signal and the components it passes through. In the New Digital Goodness, there's no inherent reason why any of the distortion will be fundamentally harmonic. Consequently, if either type of signal pathway introduces any distortion at all, the analog one will tend to be "musical" because of the harmonic aspects at its core, whereas the digital one will be something else.
I'm not arguing about the amount of distortion etc. Anybody who knows anything about this stuff knows that Ye Olde Analogue pathways were full of distortion, but that either we grew to love it or it was just inherently musical enough that we loved it from the beginning. The sound of that old analog gear you love? Yes, that's right, you love the distortion it adds.
On your third point. I remember a very good article on control theory which had a hard and often missed point. Which is digital sampling creates far more noise in the phase domain than the amplitude one. For control loops this is important because stability depends on phase. The upshot is you need a higher sampling rate than you would naively think.
Bob Pease also had some articles on audio. I think he would agree that open loop analog circuits often produce harmonics that people like. And digital and feedback controlled ones tend to produce harmonics people HATE.
The thing is feedback control systems have specs that fall apart well before the loop becomes grossly unstable. Phase noise due to sampling pushes you closer to that than you would think. Which explains the need for somewhat higher sampling rates.
Sampling doesn't really introduce significant phase noise though. A single analog filter will introduce frequency-dependent phase shift and that will dominate any tiny phase noise that might happen from sampling.
We've sacrificed quality for convenience in a lot of domains.
MP3s are an obvious example. (Although many of us realistically can't tell the difference.) And many/most listen to music through cheap earbuds rather than expensive speakers/headphones.
Or take phone calls. Some may remember Sprint's hearing a pin drop commercials to advertise the quality of its network. Now, many of us don't even have landlines and just deal with the drops and other problems of cellular phones and IP video calls.
I can easily tell the difference using good headphones. You just need to know what to look for. Once you know, you can't help but notice. The difference is the transients, especially the cymbals and the snare. If it's muddy, you're listening to compressed audio.
When it comes to phone calls, I vastly prefer calling with Facetime Audio/Whatsapp/Messenger instead of the cellular network. The difference in audio quality is enormous.
Totally depends for me. There are days when my Internet is awful, days when my cellular is awful, and days when both are. Both are less consistent than when I had a landline.
New truth: at sensible bit rates, most people cannot differentiate mp3s and wav in double blind tests. New truth: at very high bit rates, almost nobody can.
And frankly, if you replace MP3 with a newer decent codec like Vorbis or AAC that doesn't have obvious glaring issues, "very high" isn't all that crazy high either.
That said, disk is cheap, and FLAC is easy piece of mind.
Yeah. I have a big bin of CDs in my attic, many/most of which I ripped as medium-quality MP3s. Realistically, given my so so hearing, I'm not going to go redo all those at this point--especially given that so much of my music listening is streaming at this point anyway.
That's true, but psychoacoustic compression models like MP3 work precisely because they're throwing out what most people can't hear. That said, people are different, and some people can hear the difference. My ears are probably only average, but my eyes are (or were) significantly "faster" than average: Until I was well into my 30s (fortunately, this has faded with age!), fluorescent lights were pretty much just like a disco strobe to me - as a teen I could literally see them flashing, even though the phosphors moderated it a bit. (The 2000's flirtation with those wretched CF twisty bulbs brought this roaring back, as almost all of them flashed enough to be really annoying to me.) Just like there are some (very rare) women who are tetrachromes, it's reasonable to expect that there are some with "golden ears".
The cool thing about Bob Carver was that he didn't just rely on his (obviously very good) ears, but also on his knowledge, intuition/insight, and equipment. Just amazing. Thanks to OP for this.
While he didn't rely on them that much other than as a confirmation, it clearly surprised the folks at Stereophile that he was instantly able to pick up on the subtleties they noticed.
As a high schooler and Stereophile subscriber at the time this article was published I was a big fan of Mr. Carver's work, though I never got to own one of the products. A nearby hi-fi store would indulge me and let me listen to the Carver Amazing Loudspeaker, which I coveted ever since.
Anyone here remember the Tech Hi-Fi chain? Amazing gear they sold, they were a chain that started in the MIT neighborhood as I recall. Advent, Crown, all the good stuff way back in the day.
Audio Science Review (audiosciencereview.com) is an interesting counterpoint to all this huffing and puffing about what sounds "good" that has historically held the audiophile industry hostage. The guy has some state of the art test equipment and thoroughly tests equipment that people send in to be tested. Some of the results are really suprising, like a $14K DAC that was, from a performance standpoint, just garbage. He'll test almost anything, including the "audiophile network switch" that (surprise!) didn't do anything. If you're into this space his tests are almost always interesting.