I think with piano tuning, or anything that is historical and physical and analogue and has status, you're going to get lost in human psychology and so miss the real issue.
A similar field that has always been digital is audio and video compression codecs.
You see similar issues with say Apple advertising their compression as perfect, and devoted hobbyists slaving away to meet goals that the average person cannot even discern.
The capitalist, rather than technical, reality is that advances will be used to deliver the same (or lower) quality at a cheaper rate. And if you measure progress by the number of people who can watch Fast and the Furious on their Android phone on a capped data plan in a third world country that is perhaps the best thing.
And even if you care only for perfection, despite all the genius and effort applied, you could argue that most of the progress over decades has been the ability to throw more CPU at the problem.
All of this makes me think. I recently bought a car, and there seems to be quite a bit of audio difference. Or how pleasurable the audio sounds, how immersive, how the bass sounds.
1) Using Android Auto
2) Using Apple CarPlay
3) Which USB cables are used. Some seem to be better than others.
4) If it's over bluetooth - the worst experience with Android for me.
5) Car radio vs the above
6) Am I using YouTube Music or Spotify on Android Auto/Apple CarPlay
It's almost like with some of these configuration it's frustrating, while it's really pleasurable with others, but I can't tell if I'm imagining it.
From my tests so far, it feels like I'm only happy with the wired Apple CarPlay, and everything else doesn't seem satisfying. I'm not an audiophile or anything. And it has a real effect on my mood. But I have Android myself.
If anyone has any suggestions to try out here... I'll try to maybe see if all the music is downloaded with offline support beforehand and maybe check the settings in Spotify/YouTube Music apps. It seems to be related mostly to how smooth or immersive bass is to me. Because bass can be annoying or it can be pleasurable.
The majority of music apps on devices have their own "custom" EQ settings. It's probably not surprising that they all sound a little different, even if everything around them is digital and "exactly the same signal".
It very much gets into matters of preference and taste - people love to abuse the word "objective" when there's not really any objective measure. Sure, you can say the dominant frequency of a note on a piano is correct or not, but the rest? More due to taste. The preferences can often be "shared" by culture associating specific quirks with "better" or "worse", but that's still just preference.
How many musicians intentionally use older techniques or "worse" equipment specifically because of the sound they create? Look at the culture around Synthesizers, for example. Is it really crazy to believe that the "perfect" concert piano tuner example in the article is just a preference in a shared culture of "Concert Pianists"? And are we then really losing something that's "better"? Or just another cultural artifact that may fade with time?
We're not "losing" things any more than any cultural shift is a "loss".
I like one forum user's suggestion: build an audio file where the left channel is from one recording and the right channel is from another, but in perfect time sync. For example, the left may be a recording of an Apple CarPlay rendition, while the right is a recording of an Android Auto rendition. Then it should be easy for your ears to discern the difference.
The Bluetooth issue could be real. I have a Bluetooth-enabled car radio that switches between full duplex (two-way audio) and half duplex (one-way audio) mode. It sounds normal in half duplex mode, but in full duplex mode, it sounds a lot like an old telephone.
Are you sure about that? Because there are some extreme examples, for instance some USB cables won't work at all for Android Auto. Then some USB cables I had music skipping, and with at least 1 cable the quality seemed much worse for Apple CarPlay than for the best one.
It's a digital data stream. The bits either make it or they dot. Unless the cable is actually faulty (i.e., the error rate exceeds what error correction and recovery can handle), one cable cannot "sound different" to another.
The ones I tried were new, of course still could be faulty. I went to the store and asked for one that would divide to both Apple lightning cable and USB-C, so my partner and I could both use it, since she has iPhone.
The store owner said that he's not sure which ones will work, so he gave many different ones which we tried in the parking lot and then brought back. He gave us only one that spreads into 3 different connectors, USB-C, Lightning and the old Micro-USB. He said that this won't likely work. Rest of the cables were all either USB-C or lightning. However when trying, contrary to what he said, the 3 different connectors one worked, and it seemed to also have best quality, e.g. compared to one lightning which definitely had much worse performance.
What I think is happening is that like someone said somewhere below as well, there's some difference in how fast or stable the data goes through the wire and something will change how they encode/decode it for those reasons. I don't know if it's the apps or the car software. I have to assume there's some sort of quality check and something will change depending on that.
It doesn't help that packaging seems to list out random details and it's hard to compare cables with each other. Some USB cables have listed the data transfer rate, some have not, but they do allow data transfer. Then others don't.
The gold plate-ended USB cables that come in a wooden case with velvet finishing, make the reproduced sound so much smoother, the way yhey were intended by the composer. The glossiness of the gold combined with the subtle soft touch of the velvet will seemlessly flow over to the analog side of the 192kHz DAC, polishing these once harshly digitized signals up to their finest quality.
You are joking, but all I want is the bass to feel really good and immersive in EDM songs. It somehow really makes me happy and the car ride so enjoyable. Like when the bass kind of smooths around you and it feels it's coming immersively from everywhere at once. With some setups the bass kind of does a rough "boom" and goes away quickly, while with others it kind of simmers and tickles you.
I think we are on the same page. The bass is the part that makes a lot of difference. From working in automotive amplifiers i remember that there are tricks to make the bass sound louder and deeper with a limited amount of voltage difference. Usually the mac diff is +14 to -14 volts due to the battery limitation, which seems plentiful but for low tones, the energy efficiency as well as human perception is very low with audio. One of the tricks is introducing first and second harmonics. I've experienced plenty of in-car demos that we would present to customers e.g. in Munich.
Other tricks are speed adaptive eq and volume.
But in the end most quality yield boiled down to tuning. Cars full of, rather low quality speakers (BOM cost a few dollars per channel) could create a quite immersive sound.
That's very interesting to me. Because of this specifically being so weirdly pleasurable. I have now driven a lot of different cars because where I'm locally, you can use an app to rent cars for very temporary times and they are everywhere, within walking distance. Maybe it's in other countries as well, I don't know. But in any case, this has given me so much opportunities to try out various cars from 2021-2023 year release. I only recently got a license, and the car I bought is my first, so I don't have much historical experience, or experience in general, but to me when driving a car this particular audio quality is the main thing to affect my mood and perception. I either feel hyped for the rest of the day or frustrated and disappointed. And I usually don't think of myself as a music fan or anything like that. Also don't think I'm very young, I'm just a late bloomer in terms of getting a car.
Because of that impact, and because of my lack of knowledge, it was very difficult and concerning to find a car, because we are kind of trying out Android Auto, Apple CarPlay, bluetooth without anyone really giving advice on the topic. And car sellers don't seem to be mindful about the topic, neither any relatives that we tried to consult on which car to buy. The best advice car sellers have had on the topic is to "just always use the original cable" which I don't even know what it means. Well I know what it means, manufacturer likely put the proper USB with optimal transfer rates and all, but it doesn't really answer to what the factors are or anything like that.
Are we using wrong settings somewhere? Are we using optimal settings in all the apps, the car audio tuner, equalizer? I don't know.
So we have learned a lot and we decided on a car, which we bought, but presently I still feel there's something different about Android Auto, vs Apple CarPlay, that doesn't make it quite as pleasurable. Like I'm not getting this immersive hype feeling, but more of a rigid "boom". Apple CarPlay showing that the car should be capable of it, but I can't really reproduce it with my Android phone.
But surely I can't be the only one in the World, so it seems like a thing that can be this pleasurable would have been obviously solved.
And it's not about the loudness, it's about how it smooths out or spreads, or feels like it's all around you at any volume. Like you are inside the music and the bass.
When I tried to Google for solving this problem, I didn't really find any answers, so I'm sharing this here on hackernews, since frequently interesting perspectives appear here.
If there's a traffic jam or rush hour, but the music is so great and immersive, it wouldn't matter and you never want the music to end, it seems odd that there's not that much focus on it from sellers point of view or people I know in real life.
Or do I just have a weird physiology/psychology that responds to some odd things that no one else really cares about, or is some small setting somewhere?
That was my thought about part of the process too, like does YT and Spotify use a lower bitrate if you're using usb out vs analog out vs bluetooth? Plus if the data connection is 4g vs 5g or sitting in the driveway using wifi, how does that affect the streaming?
I'm not an audiophile by any means, but streaming almost always sounds worse than CD or even sometimes radio.
Yeah - with bluetooth, definitely there's much, much lower quality in some cases. Initially when trying cars I always used Android and bluetooth, and it gave me very bad impression of some of their audio systems. Then when finally bringing a cable for Apple Car play and using my partner's iPhone, the sound was totally different.
But it does also seem to depend on a car. Because I'm pretty sure I've had some very good experiences with bluetooth as well. And a case where bluetooth Youtube Music sounded good, and bluetooth Spotify sounded bad.
In addition both of these apps have different volume levels despite car having the same volume level, and Android has different volume levels compared to iOS, so all of it combined, testing around made for an annoying experience.
Audio doesn't use very much data. CD bitrate is 1.4Mbps while USB2 is 480Mbps. It would have to super bad to for audio to have problems, and USB2 is super reliable.
Or the car has a fair amount of EMR noise and one cable - being shielded - experiences less loss as a result. This could also explain why bluetooth was particularly bad.
When I asked ChatGPT about it, it also brought up interference as a potential culprit, but I'm not an expert on the topic to consider it or be confident enough to bring it up by myself. I'm not sure how well different materials are shielded from interference or how much effect can interference actually have on it.
What it specifically said is:
> High-quality USB cables often have better shielding, which helps to reduce electromagnetic interference (EMI). EMI can introduce noise or degrade audio quality, particularly in cars where multiple electronic devices are in close proximity.
How the music was produced/mastered makes a huge difference too.
There is a lot of stuff from the late 70's and early 80's where I think a combination of new technologies and cocaine made for producers creating masters that sound awful to my ears.
(disclaimer: not a professional audio engineer, just an instrument player that has some production experience).
There is also a lot of bullshit about “quality” — like the idea that you could get a significant quality improvement sampling audio at 96k as opposed to 44.1k or 48k.
For that matter I can’t believe people’s eyes don’t glaze over when they see TV ads that say, for instance, that Dawn dish detergent is better than other brands even though Dawn really has a better package of surfactants than most competitors and, over time, P&G has invested a lot of research into improving it. (E.g. ultraconcentrated soaps were an advance in practical chemistry and you really can clean more with less soap)
Although their ads must be a reason that Dawn is market leading I’m pretty sure the ads, through their style, make a withdrawal from the legitimacy bank account.
> like the idea that you could get a significant quality improvement sampling audio at 96k as opposed to 44.1k or 48k.
There can be real quality improvements, if you’re editing, recording, mastering, producing audio. If the thing you sample isn’t the same as the thing that’s going to be played back, then a higher sampling rate is sometimes called for.
This is true in imaging too. You don’t need HDR or raw formats or excessive resolution if you only show the same thing you captured, but sometimes you do if you’re going to modify it.
Some of the lore around high sampling rates is when people make a rule of thumb for recording and apply it always without thinking about the specifics of a situation. That’s easy to do, but you also have to examine the cost of using a high sampling rate. Does it cost a lot or hurt? If you’re running short on storage space, maybe, but usually no, it’s usually a tradeoff with very low downsides.
Sales people and corporate marketing absolutely do take advantage of this, for sure, and turn what are legitimate tradeoffs for some professionals into specious needs for mass market consumers.
If you want to do physics-based sound transformations which amount to applying partial or ordinary differential equations you have to sample at a rate 3x or more than the Shannon rate because otherwise you encounter instabilities.
Similarly you can't really hear more than 16 bits worth of sample depth if you got your levels perfect but in the real word recording in the studio and field a few more bits of headroom make a big difference.
Thus I'd expect people to record at a high sample rate but I think there is no practical delivery rate.
I'm somewhat agnostic as to whether or not people can hear anything atr all past 44.1kHz, particularly because if you consider how accurately people can point out objects in a real physical soundfield (e.g. like a blind man shooting a gun) and you believe phase accuracy is the key, you'd make the case that people can hear timing differences that couldn't be captured at 44.1khz.
But no "surround sound" technology comes close to replicating how accurately you can spot things in real life and I don't believe practically any commercial recordings past 44.1khz are noticeably better on any real equipment.
Yep for playback, I agree, you don’t need more than 16 bits/sample or more than 44.1k playback rate. The vast majority of humans can’t hear beyond that, at least through the air… I have heard that it can go a lot higher via direct bone contact, but that’s not relevant to playing back music anyway. I just wanted to help clarify that there are real reasons to “sample” (record) something at 96khz or higher, or at 24 bits/channel/sample, but that the reasons easily get lost, and sometimes artificially or intentionally misrepresented for marketing purposes.
I kinda relate this to sporting equipment or cars or other things people buy. Like, nearly everyone on a road bike is riding and wearing race gear, which is uncomfortable, while very few people are actually training or racing. Or, people often buy cars or computers that are far faster or more powerful than they need. Those things get rationalized similarly, with claims that the extra is necessary and makes a difference, or fears that it might be needed someday, and little reflection on the downsides.
One of the angles I find interesting when it comes to "quality" is the debate between Spotify and Apple Music. You'll see takes on Threads/Twitter like "I didn't realize just how much sound I was missing until I switched to lossless audio on Apple Music; Wow!" Its pedantic to point out: The audio quality you're hearing probably isn't better on Apple Music. I mean, the stream is definitely higher quality; AM delivers ALAC lossless audio for most music, while Spotify is still on 320kbps. But, you're listening to it through Airpods, which do not support a wireless, lossless codec. [1]
However: You did notice something different. I know you're using APs, because you noticed something different. What you noticed wasn't audio "quality" per se; you noticed the APs' faux-Atmos. Its a similar vein to saying that the vogue "TAYLOR SWIFT 9 DIMENSIONAL SURROUND AUDIO" [2] Youtube videos are "higher quality"; obviously Atmos, when its at its best, is packing more data from the master, more microphones, more streams. But, you're muxing all that data down to a lossy wireless codec, and then outputting it through two speakers. At best, what you're hearing is not necessarily higher quality, but is simply "more pleasing". And, when Apple Music delivers truly-fake-Atmos tracks, meaning the Atmos was synthesized and processed from original Stereo masters, which does happen quite frequently with older music on the platform, it might even be accurate to say that what you're hearing is actually worse quality; but it might still be more pleasing.
My take is that 96k MP3 “sux” but I might tolerate listening to it on my watch in a noisy gym or running out on city streets. If you hear 128k MP3 compared to the original CD the difference should be night and day but you might think it was OK, somewhere between 192k and 320k the quality difference disappears and even the phase relationships between the channels are pretty well preserved because Dolby Pro Logic works right.
Now there is a huge difference between a good master and a mediocre master and often when people release on a fancy format they start with a better master. For instance a lot of CDs are victims of the loudness war and it isn’t hard to make a better release.
I agree; I believe it to be the case that even when provided with an immaculate audio consumption setup, fully wired, insulated, abyss headphones, some insane schiit amp and dac, a perfectly silent room, let the listener pick the songs, the full nine yards, maybe 1% of the population could correctly discern a stereo lossless track from a stereo 320kbps more than 50% of the time over many songs. Genetics plays a role in getting into this 1%, but the far, far bigger cause is really just experience: Its having listened to the same tracks over, and over, and over again, until you've literally learned where the imperfections in this specific 320kbps encoding of this specific song are, rather than imperfections natural to the codec itself.
That's the "dirty secret" of lossless audio: If you ask even someone in this 1% to do the same thing, but for music the experimenter picks and that they've never heard before: that 1% literally becomes 0%. Maybe there's some gigahuman audio codec engineer employed deep in a basement at Dolby HQ who knows exactly the classes of imperfections that AAC encoding imprints into tracks and also has superhuman ears and such... but its damn close to 0 humans who could do this.
What this says about why anyone cares that AM, Tidal, Qobuz, whatever offer lossless audio, or that Spotify doesn't, is certainly interesting. It does seem like a tremendous waste of bandwidth and local storage, to me. But, I totally and fully subscribe to the message of the original article: It matters to me that this quality is maintained, that it exists for those people who do care about it, and that what I am consuming is as close as possible to what the original artists intended (rather than letting a bunch of cooks into the kitchen with opinions about which hertz are more important than others).
I believe you are conflating kbits/s with kHz. The latter being the unit of the sample rate. The thing is, no human ears hear higher tones than 17kHz and Nyquist and Shannon thought us that double of that as a sample rate is sufficient to reproduce tones equal or lower.
I am using k in this particular message to mean kbps but it I think it does describe the menu of sound quality options that people will hear.
I know people can't hear sine waves above 17kHz but there are questions about transient response and particularly how accurate you would need to represent phase if you want to replicate how well people can spot the direction of sounds in the real world. (Notably no "surround sound" technology of any kind would help a blind man with a gun shoot as accurately as they can in the real world)
> The capitalist, rather than technical, reality is that advances will be used to deliver the same (or lower) quality at a cheaper rate. And if you measure progress by the number of people who can watch Fast and the Furious on their Android phone on a capped data plan in a third world country that is perhaps the best thing.
This is a really strange choice of example, since advances in video compression and streaming have obviously been used to deliver higher quality at a cheaper rate.
Quoting the article, because from your comment I'm not sure you got to the end, which is a shame as it addresses your point.
TL;DR: If we're not going to aim for better, what's the point at all?
> If it weren't for the piano soloist (the conductor probably didn't notice, he just knew to defer to the piano soloist's concerns), we would have played the concert on a very slightly out-of-tune piano, and then...
>
> What?
>
> Nobody in the audience would probably notice. Certainly not in the specific. Nobody is standing up and saying, "there, see how G above middle C has one string that is 0.2hz out of tune with the others?!" Nobody is standing up and saying "that piano is out of tune, what a travesty." Perhaps some of the more sensitive listeners would have felt some vague sense that the piano could have sounded nicer, that maybe the hall needs a better piano, or something.
>
> Did the piano sound better, after all that work? Yeah... it did, I think. Hard to say. I'd like to pretend it was some colossal difference, but that's really the point. My big stupid ears are not the best judge here. Just trust the people who have the best discernment.
> Only a very few people possess the level of discernment needed to know how bad your local concert hall's piano is, and precisely how it is bad.
> If their art dies out, maybe nobody will know how bad all the pianos are. And then we'll all have slightly worse pianos than we would otherwise have. And I mean if that's the way things are going to go, then let's just steer the Earth into the Sun, because what's the point of any of this.
A similar field that has always been digital is audio and video compression codecs.
You see similar issues with say Apple advertising their compression as perfect, and devoted hobbyists slaving away to meet goals that the average person cannot even discern.
The capitalist, rather than technical, reality is that advances will be used to deliver the same (or lower) quality at a cheaper rate. And if you measure progress by the number of people who can watch Fast and the Furious on their Android phone on a capped data plan in a third world country that is perhaps the best thing.
And even if you care only for perfection, despite all the genius and effort applied, you could argue that most of the progress over decades has been the ability to throw more CPU at the problem.