5 ms·

> However, with a sampling rate of 44,000 samples per second, there were overtones of sound now missing. Human hearing only goes up to 20 kHz when you are youn

> However, with a sampling rate of 44,000 samples per second, there were overtones of sound now missing. Human hearing only goes up to 20 kHz when you are young, dropping to about 12 kHz at the age of 50. 44 kHz sampling gives a Nyquist frequency of 22 kHz, easily encompassing the full range of human hearing. Any overtones that are missing in the recording are overtones that didn't make it through the human ear in the first place. >I have long since abandoned arguments about the quality of sound. The analog sound was better. It was fuller and warmer, and it held all the sonic information. 16-bit samples on a CD can represent about 96 dB of dynamic range. The cassette tapes that he is arguing for can only represent 60-70 dB of dynamic range. It certainly doesn't hold "all the sonic information". I really appreciate the historical view that the author brings. In terms of the technology, though, he sounds like one of the foolish audiophiles buying the $1,500 HDMI cables.

mrob 8y ago
Also: >The sample rate of an MP3 is 23,000 compressed samples per second, half the sample rate of a commercial CD and a quarter of the sample rate of a studio quality digital recording. The MP3 format does not support 23kHz sample rate, which is a strange non-standard rate that I've never hear of anybody using in any format.
- stephen_g 8y ago
  Must be confusing it with the 32kHz option, which I don't believe was that widely used. As far as I recall, most MP3s were 44.1kHz (same as CDs) but 48kHz was an option too.
ms013 8y ago
Yup. I dug deep into the technology recently when I taught a class on the science of sound at a local college. When you look at media not only from the mathematical perspective (Nyquist, etc), but the physical (grain density and size on tape), it’s pretty clear that past technologies did have fundamental limits. It’s hard to disentangle the technical appeal of certain recording media from the nostalgia for a rosier looking past when reading some of these articles.
TheRealDunkirk 8y ago
This stuff again. I have several friends who are convinced that a $2,000 HDMI cable can produce "better" audio and video. I've grown so tired of these debates, that I just nod my head and move on when they start down these roads. Because of these arguments, I've been saving articles about them when I run across them. When I retire, I may make a web site devoted to debunking this nonsense forever. One good example was a guy who used "cheap" and "expensive" speaker cables, and COAT HANGERS, and challenged people to reliably tell the difference in A/B testing. No one could. It seems to me that we have the cheap and highly-available ability, these days, to 1) lift audio and video from a source directly, and 2) pass it through various stages in the reproduction chain, and 3) PROVE that this-or-that piece of kit ACTUALLY reproduces the ORIGINAL recording any better than another. (My favorite piece of gear, in an audiophile catalog, from about 30 years ago, was a replacement knob for your amplifier. Apparently, the bakelite one that came with it introduced unfavorable harmonics into your listening experience. The replacement fixed all that for you. It was a wooden puck, with a hole drilled in the back. They wanted $500 for it. .... I'm in the wrong business.) All this boils down to is various levels of equalization. You can say that a turntable and an analog amp is "better," and that's fine. You may really prefer it. But don't tell me it's "more accurate." I can demonstrate that - at the end of the day - it's just an EQ on the original recording, and does not reproduce the original more accurately. (With math and graphs and everything!) Unfortunately, I know what I'm up against. The people who believe this audiophile nonsense will very likely NOT understand the math and the graphs, and I'll be wasting my time regardless. The entire market is fed by -- and I hate to say it, because these kinds of folks are GOOD friends of mine -- people with literally more money than sense. Guess that came out a lot more rant-y than I was shooting for.
- insickness 8y ago
  > it's just an EQ on the original recording EQ along with a bit of harmonic distortion and record player noise.
- InitialLastName 8y ago
  > it's just an EQ on the original recording I'll be pedantic and say that it's actually not just an EQ, but a non-linear compression function. The needle has a different frequency response at different amplitudes, due to its mechanical properties. They used to master music specifically for that, but now they just cut the Spotify master into the record and hope everyone attributes the sound differences to "analog warmth".
saidajigumi 8y ago
There were also serious problems in many early digital (and hybrid) audio production chains. Some productions had great engineers on board who'd taken the time to learn the new tools. Others ... were varying degrees from "meh" to "disaster". For example, I have a CD that's unlistenable if you've still got your high-end hearing – there's major high-frequency artifacting. I'll bet that the mastering engineer couldn't even hear the problems. Many people formed their critical listening opinions during this time, and came to the at-the-time logically supportable conclusion that "analog sounds better". In fact, that was just an unfortunate period of industry-wide learning curve. Much of what followed was rationalization of that early opinion, generally not backed up by the science of audio recording and production.
- criddell 8y ago
  I always wonder how some bands (like Dire Straits) seemed to figure this stuff out so early on while a decade later others struggled to do the same.
  - eatmyshorts 8y ago
    You think they figured it out early? Go listen to some Miles Davis recordings from the late 50s. Analog tape, two mics, and that was it. And it still sounds better than almost everything made since. Studio-quality analog has had better than CD-quality audio since the late 50s (lower noise, more headroom). With "modern" studio techniques we've adopted multitrack recording, close mic'ing of instruments, and what not. While these may make the recording process easier, they don't make recordings of a higher quality. Give me a pair of high-quality mics in ORTF configuration and natural, unamplified instruments any day.
    - criddell 8y ago
      Is it fair to compare studio-quality analog with a consumer format for digital? You would have to compare studio-analog with studio-digital and I'd be surprised if the analog comes out on top in any way. Techniques like close mic'ing aren't really about ease, it's about capturing a different sound and it predates the digital era. Your preference for particular mics and instruments is kind of irrelevant to digital vs analog as a storage format, isn't it?
pier25 8y ago
While I agree with you, it's also true that aliasing occurs below the Nyquist frequency and some people can hear those artifacts in ideal listening conditions (acoustic environment, DA conversion, monitors/headphones, etc). I seriously doubt the author can hear the differences considering his age. Another point is maybe the author actually meant he preferred analog recording and mixing vs doing it all in digital, and his point was not really about the reproduction method. It's common knowledge in the industry that good analog mixing desks (eg: Neve) are still better than any plugin at summing and EQing. Deadmau5 recently switched his studio to a Neve desk: https://www.youtube.com/watch?v=TWZdPQAXxW4 https://www.youtube.com/watch?v=TWZdPQAXxW4
- ska 8y ago
  It's true that you can alias into these frequencies, but that's not really an issue of reproduction but of capture and the filter chain. This is often confounded in discussions of "CD quality"
- stephen_g 8y ago
  Aliasing is a problem coming from insufficient filtering. That is actually one reason that 48kHz still makes sense to upgrade to from 44.1kHz - it makes the antialiasing filter design a bit easier because the rolloff doesn't have to be as sharp to stay above the range of hearing and below the Nyquist frequency. Neve make beautiful mixers, but it's not necessarily "better' - it just has a nice sound because the analog components colour the sound in certain ways. There are other ways to do the same thing though - as long as you have good enough preamps and digitisation, you can download characterisations of certian amplifiers/mixers to apply to your digital signal. This is massive in guitar effects - you can download basically any vintage amp and sound basically identical without having to ship around heavy amps with delicate tubes in them!
- ovao 8y ago
  > It's common knowledge in the industry that good analog mixing desks (eg: Neve) are still better than any plugin at summing and EQing. At certain points in history, it was common knowledge that the Earth was flat and that a reliable method to determine whether someone was a witch was to see if she would float.
Mediterraneo10 8y ago
> Any overtones that are missing in the recording are overtones that didn't make it through the human ear in the first place. Even if human beings cannot hear higher frequencies directly, they can hear the beats produced when higher frequencies sound at the same time as lower, directly audible ones. This has sometimes been exploited for musical purposes and I do wonder if this plays some role in some listeners’ feeling that low-bitrate encoding or speaker playback lacks oomph.
- jerf 8y ago
  If the beats are audible, they're "real" and recordings will pick them up just fine. You don't need to record the original ultrasound to hear them. You still can't hear the inaudible frequencies, at the risk of stating a tautology. You can make some small case for lower frequencies, because even if they are not audible they get to the point where other senses can pick them up too. A sufficiently loud 1Hz wave is simply something shaking. But you can't hear ultrasound.
  - Mediterraneo10 8y ago
    They are an artifact, a side effect, of the way human hearing works, they are not actually present on a recording medium (only the frequencies that produce the effect in the human ear are).
    - 6nf 8y ago
      Yes so the EFFECT is recorded cause the microphone hears it just the same as a human ear.
    - hannasanarion 8y ago
      And they weren't actually present at the time of recording either, they're a product of waves adding together. Those same waves are still added together when it plays from your speakers.
      - Mediterraneo10 8y ago
derefr 8y ago
I think people's experience with digital audio being "worse" than analogue audio, mostly comes down to the average audio reproduction equipment getting worse over the years. (Worse and cheaper. Like flights.) Everyone used to have a pair of gigantic bookshelf speakers, where you didn't even need a subwoofer because the woofer covered the full low end of the range. They were also shitty, but in a particular analogue way that made them "warmer" and introduced a bit of hiss (which essentially was doing a cheap antialias of the source, according to our ears.) When was the last time you listened to a good set of speakers, outside of a movie theatre? (Good headphones don't count—they can reproduce audio with high fidelity, but they can't produce enough sound to vibrate your body from the direction they're coming from—let alone enough to make the acoustics of the room you're listening in important.)
Hasz 8y ago
Analog introduces it's own set of distortions into the final sound -- it's simply a preference for that style of distortion. Totally duplicate on modern equipment though, no $1000 cables needed.
musicale 8y ago
Well-mastered CDs sound fantastic - no hiss and amazing dynamic range. I like tape artifacts such as saturation, hiss, wow/flutter, distortion, etc. as much as the next person. I even like the "sound" of cassettes. But CDs are vastly superior in terms of accurately reproducing audio signals.

mrob 8y ago

Also: >The sample rate of an MP3 is 23,000 compressed samples per second, half the sample rate of a commercial CD and a quarter of the sample rate of a studio quality digital recording. The MP3 format does not support 23kHz sample rate, which is a strange non-standard rate that I've never hear of anybody using in any format.

stephen_g 8y ago
Must be confusing it with the 32kHz option, which I don't believe was that widely used. As far as I recall, most MP3s were 44.1kHz (same as CDs) but 48kHz was an option too.

ms013 8y ago

Yup. I dug deep into the technology recently when I taught a class on the science of sound at a local college. When you look at media not only from the mathematical perspective (Nyquist, etc), but the physical (grain density and size on tape), it’s pretty clear that past technologies did have fundamental limits. It’s hard to disentangle the technical appeal of certain recording media from the nostalgia for a rosier looking past when reading some of these articles.

TheRealDunkirk 8y ago

This stuff again. I have several friends who are convinced that a $2,000 HDMI cable can produce "better" audio and video. I've grown so tired of these debates, that I just nod my head and move on when they start down these roads. Because of these arguments, I've been saving articles about them when I run across them. When I retire, I may make a web site devoted to debunking this nonsense forever. One good example was a guy who used "cheap" and "expensive" speaker cables, and COAT HANGERS, and challenged people to reliably tell the difference in A/B testing. No one could. It seems to me that we have the cheap and highly-available ability, these days, to 1) lift audio and video from a source directly, and 2) pass it through various stages in the reproduction chain, and 3) PROVE that this-or-that piece of kit ACTUALLY reproduces the ORIGINAL recording any better than another. (My favorite piece of gear, in an audiophile catalog, from about 30 years ago, was a replacement knob for your amplifier. Apparently, the bakelite one that came with it introduced unfavorable harmonics into your listening experience. The replacement fixed all that for you. It was a wooden puck, with a hole drilled in the back. They wanted $500 for it. .... I'm in the wrong business.) All this boils down to is various levels of equalization. You can say that a turntable and an analog amp is "better," and that's fine. You may really prefer it. But don't tell me it's "more accurate." I can demonstrate that - at the end of the day - it's just an EQ on the original recording, and does not reproduce the original more accurately. (With math and graphs and everything!) Unfortunately, I know what I'm up against. The people who believe this audiophile nonsense will very likely NOT understand the math and the graphs, and I'll be wasting my time regardless. The entire market is fed by -- and I hate to say it, because these kinds of folks are GOOD friends of mine -- people with literally more money than sense. Guess that came out a lot more rant-y than I was shooting for.

insickness 8y ago
> it's just an EQ on the original recording EQ along with a bit of harmonic distortion and record player noise.
InitialLastName 8y ago
> it's just an EQ on the original recording I'll be pedantic and say that it's actually not just an EQ, but a non-linear compression function. The needle has a different frequency response at different amplitudes, due to its mechanical properties. They used to master music specifically for that, but now they just cut the Spotify master into the record and hope everyone attributes the sound differences to "analog warmth".

saidajigumi 8y ago

There were also serious problems in many early digital (and hybrid) audio production chains. Some productions had great engineers on board who'd taken the time to learn the new tools. Others ... were varying degrees from "meh" to "disaster". For example, I have a CD that's unlistenable if you've still got your high-end hearing – there's major high-frequency artifacting. I'll bet that the mastering engineer couldn't even hear the problems. Many people formed their critical listening opinions during this time, and came to the at-the-time logically supportable conclusion that "analog sounds better". In fact, that was just an unfortunate period of industry-wide learning curve. Much of what followed was rationalization of that early opinion, generally not backed up by the science of audio recording and production.

criddell 8y ago
I always wonder how some bands (like Dire Straits) seemed to figure this stuff out so early on while a decade later others struggled to do the same.
- eatmyshorts 8y ago
  You think they figured it out early? Go listen to some Miles Davis recordings from the late 50s. Analog tape, two mics, and that was it. And it still sounds better than almost everything made since. Studio-quality analog has had better than CD-quality audio since the late 50s (lower noise, more headroom). With "modern" studio techniques we've adopted multitrack recording, close mic'ing of instruments, and what not. While these may make the recording process easier, they don't make recordings of a higher quality. Give me a pair of high-quality mics in ORTF configuration and natural, unamplified instruments any day.
  - criddell 8y ago
    Is it fair to compare studio-quality analog with a consumer format for digital? You would have to compare studio-analog with studio-digital and I'd be surprised if the analog comes out on top in any way. Techniques like close mic'ing aren't really about ease, it's about capturing a different sound and it predates the digital era. Your preference for particular mics and instruments is kind of irrelevant to digital vs analog as a storage format, isn't it?

pier25 8y ago

While I agree with you, it's also true that aliasing occurs below the Nyquist frequency and some people can hear those artifacts in ideal listening conditions (acoustic environment, DA conversion, monitors/headphones, etc). I seriously doubt the author can hear the differences considering his age. Another point is maybe the author actually meant he preferred analog recording and mixing vs doing it all in digital, and his point was not really about the reproduction method. It's common knowledge in the industry that good analog mixing desks (eg: Neve) are still better than any plugin at summing and EQing. Deadmau5 recently switched his studio to a Neve desk: https://www.youtube.com/watch?v=TWZdPQAXxW4 https://www.youtube.com/watch?v=TWZdPQAXxW4

ska 8y ago
It's true that you can alias into these frequencies, but that's not really an issue of reproduction but of capture and the filter chain. This is often confounded in discussions of "CD quality"
stephen_g 8y ago
Aliasing is a problem coming from insufficient filtering. That is actually one reason that 48kHz still makes sense to upgrade to from 44.1kHz - it makes the antialiasing filter design a bit easier because the rolloff doesn't have to be as sharp to stay above the range of hearing and below the Nyquist frequency. Neve make beautiful mixers, but it's not necessarily "better' - it just has a nice sound because the analog components colour the sound in certain ways. There are other ways to do the same thing though - as long as you have good enough preamps and digitisation, you can download characterisations of certian amplifiers/mixers to apply to your digital signal. This is massive in guitar effects - you can download basically any vintage amp and sound basically identical without having to ship around heavy amps with delicate tubes in them!
ovao 8y ago
> It's common knowledge in the industry that good analog mixing desks (eg: Neve) are still better than any plugin at summing and EQing. At certain points in history, it was common knowledge that the Earth was flat and that a reliable method to determine whether someone was a witch was to see if she would float.

Mediterraneo10 8y ago

> Any overtones that are missing in the recording are overtones that didn't make it through the human ear in the first place. Even if human beings cannot hear higher frequencies directly, they can hear the beats produced when higher frequencies sound at the same time as lower, directly audible ones. This has sometimes been exploited for musical purposes and I do wonder if this plays some role in some listeners’ feeling that low-bitrate encoding or speaker playback lacks oomph.

jerf 8y ago
If the beats are audible, they're "real" and recordings will pick them up just fine. You don't need to record the original ultrasound to hear them. You still can't hear the inaudible frequencies, at the risk of stating a tautology. You can make some small case for lower frequencies, because even if they are not audible they get to the point where other senses can pick them up too. A sufficiently loud 1Hz wave is simply something shaking. But you can't hear ultrasound.
- Mediterraneo10 8y ago
  They are an artifact, a side effect, of the way human hearing works, they are not actually present on a recording medium (only the frequencies that produce the effect in the human ear are).
  - 6nf 8y ago
    Yes so the EFFECT is recorded cause the microphone hears it just the same as a human ear.
  - hannasanarion 8y ago
    And they weren't actually present at the time of recording either, they're a product of waves adding together. Those same waves are still added together when it plays from your speakers.
    - Mediterraneo10 8y ago

derefr 8y ago

I think people's experience with digital audio being "worse" than analogue audio, mostly comes down to the average audio reproduction equipment getting worse over the years. (Worse and cheaper. Like flights.) Everyone used to have a pair of gigantic bookshelf speakers, where you didn't even need a subwoofer because the woofer covered the full low end of the range. They were also shitty, but in a particular analogue way that made them "warmer" and introduced a bit of hiss (which essentially was doing a cheap antialias of the source, according to our ears.) When was the last time you listened to a good set of speakers, outside of a movie theatre? (Good headphones don't count—they can reproduce audio with high fidelity, but they can't produce enough sound to vibrate your body from the direction they're coming from—let alone enough to make the acoustics of the room you're listening in important.)

Hasz 8y ago

Analog introduces it's own set of distortions into the final sound -- it's simply a preference for that style of distortion. Totally duplicate on modern equipment though, no $1000 cables needed.

musicale 8y ago

Well-mastered CDs sound fantastic - no hiss and amazing dynamic range. I like tape artifacts such as saturation, hiss, wow/flutter, distortion, etc. as much as the next person. I even like the "sound" of cassettes. But CDs are vastly superior in terms of accurately reproducing audio signals.