I do not consider myself an audiophile (too much BS in this field) but I do buy lots of physical music releases (vinyl and cds, preferably cds). Beside the obvious audio quality, there are other reasons to do so, all of them as a whole making for a very different experience from streaming. Downloads don't cut it either. Alas, physical releases are not always available and that's a shame because there is no shortage of good new music these days. I found myself down the rabbit hole of Bandcamp very often only to find that some great albums don't have any physical release. That's how it is, I guess.
I'm not an audiophile but I have a pair of speakers that is probably 70 or so laptops of volume and something nice to hook them to. I stream lots of music through the YouTube app in my smart TV, and last night some of that was written by Grieg[0]. Remember this Qwest commercial [1]? Streaming music has its pleasures and undistorted sound should be one of them.
I'm not sure if it's the source material or Youtube's compression, but music on Youtube often sounds terrible compared to e.g. Spotify. There are obvious artifacts and dynamic range issues.
Music on YouTube is resampled to 48kHz for Opus. YouTube's resampler is probably optimised for speed over precision. Resampling to 48kHz before uploading can yield better results.
I watched a video recently where a Mastering Engineer said that they will cut different releases for different streaming services. There is even a tool they recommended to compare the masters to different platforms formatting: https://nugenaudio.com/mastercheck/
I don't feel like streaming quality is the biggest issue. Spotify's highest quality rate (for paid users) is pretty good, but the audio compression for streaming to a bluetooth device kills the quality. I've stuck with wired headphones because the quality is so much better than bluetooth.
I have Sennheiser Momentum True Wireless with aptX HD support and the sound is marvelous. No compression artifacts at all.
AptX HD implementation on Pixel 2 XL, on the other side, is subpar. I sometimes get nasty clicks on music with high dynamic range and there's no way to avoid them other than to set output codec to anything but aptX HD, which tremendously affects the quality.
I disagree, I have switched from Spotify premium (HQ) to Tidal uncompressed and have noticed a huge difference. I usually listen on a wired pair of Senheisser HD-25 ii’s. Its very frustrating for me that Spotify doesn’t support anything better than 320 MP3, it truly sounds muddy.
If music needs saving from something is not streaming quality, but the increasingly marginal and irrelevant role it has (even in teens) compared to the cultural power it had in the 60s up to the 90s...
If it is a recording -- electric samples from one or a few microphones -- it is Low Quality.
But it's usually good enough. Ultimately, your brain has to reconstruct something from what your ears give it. If the experience is close enough to what it would have been if you were there, then you are getting some value. The experience might even be better, if you would have been distracted, there, but now can give it your full attention. But it is never "the same".
It's just a question of whether the replication could have been better than it is, given reasonable care. It always could have been, and you might have paid enough to get that, but nobody is selling. People who claim to be selling it are peddling snake-oil: vinyl this, tube-amp that, FLAC t'other, none are the real thing.
If there was demand, Spotify would just add FLAC. They already added 320kbps compressed streams, and I doubt anyone can hear the difference between that and FLAC.
The vast majority of the audio perception people are going to pick up on happens during production. And https://en.wikipedia.org/wiki/Loudness_war is why everything sounds worse than it used to. When records were being made for vinyl, the physical material wouldn't allow for the level of "loudness" that removes dynamic range from the track. That changed when all music started being produced digitally for digital distribution. Ironically, digital audio actually has a much larger dynamic range and should be better, but because of the way it's mastered it comes out louder but with less dynamic range.
Audio compression is the least of the problems for people interested in high quality recordings.
I backed the Pono Kickstarter. I got a player that had some cool features, but I ultimately stopped using it after a few months because it was slow and battery hungry compared to my old Sansa Clip with Rockbox.
Exactly. Audio is mastered for the (generally) cheap equipment it is played back on. The master tape might have awesome dynamic range but a cheap FM radio in the eighties definitely did not. So, producers optimized for that first using analog filters and later digital ones. Not much has changed with this practice despite lots of technical progress. People still listen to music on car radios, on cheap phones with lousy speakers or some cheap head set or a pair of 20$ bluetooth speakers, or worse, in hotel lobbies. Everything is optimized for equipment with poor dynamic range. Worse, many phones add their own filters to "improve" the sound.
It's sort of the the audio equivalent of increasing contrast, sharpness, and saturation in a photo. These are variations of compressing dynamic range. Instagram took this to the extreme. And phone cameras exploit this by simply pairing mediocre lenses and sensors with software that blurs out the noise and yanks up the contrast and saturation. It works. Consumers love it.
Audio filters do the same with sound. What goes on the master already lost most of the information during the editing process and this is on purpose.
The irony of course is that rock musicians like Neil Young used analog audio filters for creative purposes. An electric guitar is basically a guitar that passes sound through lots of electronics to distort the sound in all sorts of pleasing ways. The distortion is intentional. I'm a big fan of Neil Young early Crazy Horse recordings. If you need a nice example, "Down by the River" exists in many versions on Spotify. One of the better ones is a live recording from 1970; four years before I was born. Sort of escalates into some nice guitar work. There's nothing subtle about that and it sounds fine to me on Spotify. Which he grudgingly unbanned a few years ago after withdrawing all his music for a year or so.
It's my understanding that digital production tools also increased the amount of dynamic range compression that could be applied to the signal.
As an aside, there was a study done where they took high-bitrate (do not remember if it was SACD or DVD Audio) audio recordings and truncated the signal to 16 bits with no dithering. Nobody could tell the difference between the truncated and the original. On the other hand the truncated signal often sounded better than the CD version, likely due to better mastering of the high definition version.
Lastly, there is a case for higher definition recordings, and that is for remixing. There is some N at which N stages of applying 320kbps compression will become audible, and that N is fairly low, so if you want to sample and remix audio, lossless audio (and even higher sampling-rate and resolution audio) becomes useful.
There was a study, but it was't obvious if the SACD and DVD-Audio sources really did have 24-bits of audio, or if they were made from for-CD 16-bit digital captures of the original sources.
And even if they were clean 24-bit masters from an original recording, it's not obvious if the recording chain was 24-bit - a lot of early digital recordings were 16-bit - or if the analog masters were compressed and limited to the point where there was 24-bits of detail left.
This sounds ridiculous, but I know for a fact that some CDs in the 80s were recorded from vinyl records instead of being recorded and mastered from the original analog master tapes.
Generally getting hold of original master tapes can be a huge pain in the ass, so it's not unusual for record companies to take shortcuts - especially for relatively low-run formats like DVD-Audio.
The way to do this test is to record a clean acoustic performance through a top-quality microphone with no compression. If you can still hear no difference, then it's fair to say that 24-bit audio is pointless.
In fact it's complete normal to hear a difference on good studio monitors under those conditions.
Edit: the point being that audiophile audio is a thing. Even if only 10% of the population cares about it, that's still a perfectly valid reason for it to exist.
Normally I don't particularly mind the meandering, personal style that typifies these magazine-type pieces but this one was fairly grating. Maybe it's the more or less immediate elimination of any suggestion that there could be a difference between Young's views, the author's views, and "the truth."
> At ground level, which is to say not the level where technologists live but the level where artists write and record songs for people who care about the human experience of listening to music, the internet was as if a meteor had wiped out the existing planet of sound.
> digital-recording technology, as opposed to low-quality streaming services, can be a gift to musicians, properly deployed
The article doesn't really have any interest in talking about streaming, or quality, or music, or the various actually realistic options people currently have... I guess that's fine. The closest this comes to having something useful to say about its ostensible topic is a small aside about how music production has changed over the years.
And who speaks about the audio aspects, noise including, but unfortunately produces his own video with very audible noise, which he could have filtered even in the post-production using the open source program (1), if he used poor microphones originally. I prefer no noise to the "fidelity" in such cases.
Hopefully the core is still here but that's indeed really unfortunate to not find some more time / interest into polishing the form of the message in this case.
It's not like Neil Young can hear it anyway. Dude has had tinnitus for decades, and there's no way he doesn't have major noise-induced hearing loss from standing in front of those cranked Fender Deluxes for all those years.
I've always thought that if you can hear the difference then you aren't listening to the music; I say that as someone who plays multiple musical instruments and is a reasonably competent audio engineer.
42 comments
[ 1.6 ms ] story [ 112 ms ] threadSo it’s not surprising that most users aren’t interested in lossless or other high-fidelity options.
Meanwhile, I assume most audiophiles with traditional hifi setups would prefer downloads or physical media over streaming.
I find that unlikely. Sure, it's preferred but that's because of the quality. Not because audiophiles enjoy fumbling with one disc at a time.
[0] https://www.youtube.com/watch?v=PNIon17o5kQ
[1] https://www.youtube.com/watch?v=xAxtxPAUcwQ
AptX HD implementation on Pixel 2 XL, on the other side, is subpar. I sometimes get nasty clicks on music with high dynamic range and there's no way to avoid them other than to set output codec to anything but aptX HD, which tremendously affects the quality.
Google might not be at fault here, I suspect this is Qualcomm code.
The difference in quality is HUGE for music, not so much for anything else.
(Currently listening to spoken words at 48kbps and my ears are by no means bleeding)
Take one part modem noise, mix thoroughly with 2 parts bass and 1/2 part melody line. Serving suggestion: crank the bitrate way up to a full 96 Kbps!
/s
Lyrics are one thing that makes some music meaningful. If there aren't any, or if they are unintelligible, then that meaning is lost.
But it's usually good enough. Ultimately, your brain has to reconstruct something from what your ears give it. If the experience is close enough to what it would have been if you were there, then you are getting some value. The experience might even be better, if you would have been distracted, there, but now can give it your full attention. But it is never "the same".
It's just a question of whether the replication could have been better than it is, given reasonable care. It always could have been, and you might have paid enough to get that, but nobody is selling. People who claim to be selling it are peddling snake-oil: vinyl this, tube-amp that, FLAC t'other, none are the real thing.
Don't get me started about speaker wire.
If there was demand, Spotify would just add FLAC. They already added 320kbps compressed streams, and I doubt anyone can hear the difference between that and FLAC.
The vast majority of the audio perception people are going to pick up on happens during production. And https://en.wikipedia.org/wiki/Loudness_war is why everything sounds worse than it used to. When records were being made for vinyl, the physical material wouldn't allow for the level of "loudness" that removes dynamic range from the track. That changed when all music started being produced digitally for digital distribution. Ironically, digital audio actually has a much larger dynamic range and should be better, but because of the way it's mastered it comes out louder but with less dynamic range.
Audio compression is the least of the problems for people interested in high quality recordings.
Also, I trust the state of Neil Young's hearing even less than I trust my own.
It's sort of the the audio equivalent of increasing contrast, sharpness, and saturation in a photo. These are variations of compressing dynamic range. Instagram took this to the extreme. And phone cameras exploit this by simply pairing mediocre lenses and sensors with software that blurs out the noise and yanks up the contrast and saturation. It works. Consumers love it.
Audio filters do the same with sound. What goes on the master already lost most of the information during the editing process and this is on purpose.
The irony of course is that rock musicians like Neil Young used analog audio filters for creative purposes. An electric guitar is basically a guitar that passes sound through lots of electronics to distort the sound in all sorts of pleasing ways. The distortion is intentional. I'm a big fan of Neil Young early Crazy Horse recordings. If you need a nice example, "Down by the River" exists in many versions on Spotify. One of the better ones is a live recording from 1970; four years before I was born. Sort of escalates into some nice guitar work. There's nothing subtle about that and it sounds fine to me on Spotify. Which he grudgingly unbanned a few years ago after withdrawing all his music for a year or so.
As an aside, there was a study done where they took high-bitrate (do not remember if it was SACD or DVD Audio) audio recordings and truncated the signal to 16 bits with no dithering. Nobody could tell the difference between the truncated and the original. On the other hand the truncated signal often sounded better than the CD version, likely due to better mastering of the high definition version.
Lastly, there is a case for higher definition recordings, and that is for remixing. There is some N at which N stages of applying 320kbps compression will become audible, and that N is fairly low, so if you want to sample and remix audio, lossless audio (and even higher sampling-rate and resolution audio) becomes useful.
And even if they were clean 24-bit masters from an original recording, it's not obvious if the recording chain was 24-bit - a lot of early digital recordings were 16-bit - or if the analog masters were compressed and limited to the point where there was 24-bits of detail left.
This sounds ridiculous, but I know for a fact that some CDs in the 80s were recorded from vinyl records instead of being recorded and mastered from the original analog master tapes.
Generally getting hold of original master tapes can be a huge pain in the ass, so it's not unusual for record companies to take shortcuts - especially for relatively low-run formats like DVD-Audio.
The way to do this test is to record a clean acoustic performance through a top-quality microphone with no compression. If you can still hear no difference, then it's fair to say that 24-bit audio is pointless.
In fact it's complete normal to hear a difference on good studio monitors under those conditions.
Edit: the point being that audiophile audio is a thing. Even if only 10% of the population cares about it, that's still a perfectly valid reason for it to exist.
> At ground level, which is to say not the level where technologists live but the level where artists write and record songs for people who care about the human experience of listening to music, the internet was as if a meteor had wiped out the existing planet of sound.
> digital-recording technology, as opposed to low-quality streaming services, can be a gift to musicians, properly deployed
The article doesn't really have any interest in talking about streaming, or quality, or music, or the various actually realistic options people currently have... I guess that's fine. The closest this comes to having something useful to say about its ostensible topic is a small aside about how music production has changed over the years.
24/192 Music Downloads ...and why they make no sense
https://people.xiph.org/~xiphmont/demo/neil-young.html
[edit, add prior discussion on HN:]
https://news.ycombinator.com/item?id=19318898
Video by the same man of the same topics but with shiny lights and graphs.
1) https://manual.audacityteam.org/man/noise_reduction.html
From https://www.audiologyonline.com/releases/by-time-we-got-to-3...: "It was guitarist Neil Young who suggested Stills try a new modern-design, behind-the-ear hearing solution called Dual."
(Obviously up to a point)
https://en.wikipedia.org/wiki/Loudness_war