Is high-fidelity audio a genuine product or unnecessary overkill?
Note: if this topic is better served in ~music than ~tech feel free to move it!
If I wanted to buy Linkin Park's A Thousand Suns, I have the following options:
From Amazon
- 256 kbps VBR MP3 ($11.49)
From 7digital
- 320 kbps MP3 + 256 kbps MP3 ($12.99) (I'm assuming it's 320 CBR/256 VBR)
- 16-bit/44.1kHz FLAC ($16.49)
From HDTracks
- 24-bit/48kHz FLAC ($19.98)
From Qobuz, which appears to be a different mastering of the album:
- "CD Quality" FLAC ($14.49)
- 24-bit/48kHz FLAC ($16.49)
- 24-bit/48kHz FLAC ($10.99 with subscription to their $250/year service)
Does paying more for the higher fidelity actually matter? I suspect that this is just a form of price discrimination preying on my want to have an "objectively" better product, because I'm assuming there's a ceiling for audio quality that I can actually notice and the lowest encoding available here probably hits that. I also don't have any special listening hardware.
I understand the value of FLAC as a lossless archival encoding (I used to rip all my CDs to FLAC for this purpose, and I've been downloading my Bandcamp purchases in FLAC all the same), but for albums I can't get through that service it appears that the format has a high premium put on it. Bandcamp lets me pay the same price no matter the format, but every other store seems to stratify out their offerings based on encoding alone. A Thousand Suns costs nearly double on HDTracks what it does on Amazon's MP3 store, for example, despite the fact that I'm getting the exact same music, just compressed in a different way.
As such, is paying more for FLAC unnecessary? Is high-fidelity FLAC in particular (the 24-bit/48kHz options) snake oil?
Furthermore, Qobuz seems to offer a different mastering of the album, which seems like it actually could be significant, but it's hard to know. Is this (and the various other "remasters" out there) a valid thing, or is it just a way to try to get me to pay more unnecessarily?
(Note: I'm using this specific album simply because it was a good example I could find with lots of different stratified options -- I'm not interested in the particulars of this album specifically but more in the general idea of audio compression across all music).
Most people can't tell the difference, and as you get older that becomes more and more likely - good hearing doesn't last forever. The simple truth is you are better off with a pair of high quality headphones than a hundred thousand dollar audio system - as awesome as those can be, the headphones will get you 90% of the way there and shave two digits off the price tag. If you are the kind of person who can tell the difference easily, odds are pretty good you have a shot at a career as a musician or sound technician. The techs make damn good money setting up venues for live shows or producing masters in the studio.
Everyone is always comparing formats and talking about the loudness wars. I think that kinda misses the point. In truth it's more about the mastering process than the physical or digital format. The sound engineer who produces that final master has far more to do with the sound than any media format ever will. The reason so many say vinyl 'sounds' better is because you have to master it differently (and not for volume) just to encode the music on the physical media. Buying vinyl is buying a better mastering process in most cases.
The only time I can notice any real sound quality difference with digital formats is when there is a very busy piece of a track that has a lot of higher notes, and it's usually easiest to notice on instruments that produce a pure, clear sound such as violins and horns, rather than fuzzy guitars or thumping bass drums. Sometimes the encoding gets a bit confused with those parts and fails to capture the layered sound, it ends up washed out or warbling a bit.
I always go for the FLAC just because I can then encode it however I like. If I bought an MP3 and re-encoded it in another format, that'd be the kiss of death for the audio quality. Always get a lossless copy, storage space is cheap and it only gets cheaper every day. As for the various FLAC bitrates, I've never been able to spot a whiff of difference there and I'd be a bit skeptical of someone who says they could. It's not impossible, it's just someone with a one in a million pair of ears.
If I were buying physical copies I'd always prefer the vinyl to the CD just because vinyl is gorgeous. Larger art, more liner notes and trivia, generally better sound due to the more restrictive mastering process, and they still appreciate in value more than any other physical media format made today. The CD jewel cases are ugly, cramped, and annoyingly fragile by comparison. Vinyl is a hell of a lot tougher than those flimsy CDs, too.
From those you list above, the one that catches my eye immediately is the Quobuz offering just because it's mastered differently. I'd want to hear both versions and then decide which engineer was phoning it in, and which one was bringing the soundscape to life.
The higher quality FLAC are mostly only useful if intending to edit it in some fashion, the rest is snake oil 99.99% of the time. I'd probably choose the highest quality available if cost is negligible anyway.
100 on board the vinyl/FLAC train. Usually the physical copy comes with a download copy too now, but if it doesn't, I'll pirate digital copy shamelessly if I want a portable copy. Media industries try to play both sides of the 'you own this' and 'you license this for personal use' coin, and that is a moral workaround IMO.
I agree FLAC is good if you want to future-proof for transcoding in the future.
I think basically no-one can hear the difference between CD quality and 24-bit/48khz (especially not without very expensive equipment) and 24-bit is only useful if you're planning to edit the audio yourself.
So that CD quality FLAC option from Qobuz is looking pretty good to me.
Qobuz actually has the original release as well as the remaster if you want to compare. I'm assuming the samples for the song on the site aren't streaming losslessly but it seems like even a compressed stream could still be used to identify differences in mastering. I compared a few samples for both and couldn't distinguish anything.
As for getting FLAC, with everything I buy from Bandcamp it's easy, and that's where I plan to spend most of my money moving forward. I mostly asked this because I'm in the process of "buying back" some of the stuff I liked from streaming but don't own (now that I'm moving away from Google Play Music), and I wasn't sure whether it was worth paying the premium for high-fidelity audio. Based on this thread, I'm more than likely fine with the lower quality stuff.
I can't tell the difference there, but in those samples there's not much going on - the last part is the place where one might be able to notice something as the layers of sound start to stack up, but the clips end right there. :P
Just to add on here, HDTracks releases are also typically remastered (though I am not sure about this exact release), so they are another option to choose if mastering quality is important to you.
I can’t seem to find the article I once read that confirms and explains it, but my ability to appreciate aural nuance and complexity increases as I age (but i still crank the Metallica now and then). It’s also why us olds tend to prefer jazz and classical music.
While I don’t think cd quality can be much improved upon, compressed formats pretty quickly reveal their limitations to my ears using headphones and an average stereo. Note audio quality on the low end is way better now than it was 20 years ago. I use sond md-7605 (I think?) studio headphones. Also airpod pros which are damn good. Also the mbp speakers are really amazing for a given volume.
I rip cds using aac vbr highest quality, which i can’t distinguish from lossless aac or flac, but i can, at least with high quality source material, hear the difference highest quality vbr and 320 cbr. And 128 mp3 blech. Standard rock and pop is usually fine at 192 tho.
If you got ~25 minutes of time, I HIGHLY recommend watching this video from Xiph.org's Mont Montgomery (also hosted here, on Youtube)for a very good introduction on digital audio. Basically goes to show why the 16-bit, 44.1kHz standard was chosen for listening...and why you really don't need more than that (for listening).
As far as compression goes, I prefer having my "archived" audio available in a lossless format whose source can be transcoded to another format if need be. That "archived" format is available on my home system, which is what I'll listen to when at my desktop. Whether that's worth the premium is up to you. The websites dedicated to audiophiles definitely gouge, but if they're using "better" (in the ear of the beholder, of course) masters, they could be worth it.
As far as whether or not you can hear the different between a lossless file and a 256kbps MP3... probably not. But the differences are there. I've done comparisons where I line up the wave-forms of a lossless audio track and it's lossy counterpart, inverted the waves, and was left with the difference which constitutes what's missing from the lossy file and the artifacts that result from the lossy compression. Of course, the lower the bit-rate, the more that's missing (original audio content) and the more that's added (compression artifacts). What's mostly lost is higher-frequency content from cymbals. Not a lot is lost from the classical music I've done this test with, besides the occasional cymbal hit and some noise from fingers moving on strings. Anyway, thought that was an amusing result as I often think of hardcore audiophiles as also being classical music snobs. But it's ~rock~ music that seems to be most hampered by lossy compression!
Note: the method I used may suck.
Noticing compression also depends a ton on the types of music, the mastering, and the hardware played on. Classical music tends to auffer worst IMO.
Good headphones are a great starting point for value quality listening. Big stereo/home theater sets can and do scale dramatically up in cost.
In terms of plain listening, with no analysis, I agree.
However, if you sum two audio files as I described, you don't even have to listen to understand how they're different. You'll see in the resultant file what's missing/added. And in my analysis
this generally doesn't follow. It's mostly high frequency content and sibilance that seems to be removed, which isn't too common in classical music. Or, I should say, orchestral music. At least, the orchestral music that I've tested this with (Berlioz, Stravinsky, Debussy, Beethoven, Bartók, maybe a couple others... been a while).
You can do an ABX test [1] to check your ears and equipment to see if it makes sense for you. It is a combination of subjective experience / what you want from your music and your equipment (meaning every component in your signal chain including your ears ;) ).
[1] https://abx.digitalfeed.net/
This was great! Thank you.
I tried the test twice, 5 trials each, one with the pair of earbuds that came with my phone, and one with a nice pair of IEMs. I got 60% on the first trial with my earbuds and 64% on the second trial with my IEMs. If I'm being honest, most of the time I was outright guessing. With some of them I could convince myself that there was some tiny perceptible difference by listening really hard, but with the James Blake and Chicks samples I couldn't distinguish anything at all and just went with a "hunch" (funny enough, I actually got 80% correct on the James Blake one both times, higher than some of the other ones where I convinced myself I could hear a difference).
I think this was a good way of pointing out that even if there is a difference, it's beyond the scope of my hardware and hearing.
The headphones also save you the hassle of trying fifty brands of speakers to find the right ones for your ears. Those discussions are longer than war and peace when they crop up because it's almost entirely subjective taste. The material in the cones makes a difference, and not in the good/bad way, it's more a personal preference. For example, ceramic speaker cones (rather than paper) tend to last longer and bring out more of the brass sound if you like to listen to a lot of horns.
The materials used make a difference, the shape of the speakers makes a difference, even the material used in the speaker case can make a difference. The truth is most speaker brands kick ass, the discussion of 'best' speaker is a truly pointless one. High quality audio is about having a lot of speakers that each specialize in certain areas of sound reproduction, and then sending the right parts of your audio to the specific speaker best suited to bring it to life. That's part of why it costs more - you just need more speakers and smarter electronic components.
Odds are good there's at least one major speaker shop in your closest major city. Places like that are good if you want to try out a lot of brands in person, and there are usually some techs there who really know their audio. If you buy from them, typically as part of the package they'll let you take home the purchase to try it out. If you aren't satisfied, within a month or two there's a window where you can swap to different brands - the shop doesn't care if you change up your purchases as long as your money stays with them. Just beware of any shop that has their folks on a sales commission rather than a flat rate because that muddles their incentive to sell you exactly what you want rather than what makes them money.
I'd just say beware of Bose. Not for the quality - it's the price to quality ratio that tends to suck there. If you think you want Bose, I'd say check out Klipsch instead. Nearly the same price, and vastly better quality. Don't just pay for the name. I'm also partial to Polk (reliable as fuck), Rocket (their subs will crack your foundation), and Phase Tech (their tower speakers are on another level). That's just my preference from the last time I was in the market, and that was many years back. Vendors change hands, quality changes, just like with any industry.
The go-to place online for real home entertainment gear heads has been the AVScience forum. You can discuss brands there with an obsessive audience until the stars burn out if that's your thing. Not many specialty forums like that online that are 20 years old and still bumping. That company will fly your product out to you, install it, and guarantee it like no other, as long as you are in the continental USA.
Your expertise on this and music in general is incredible! Do you have experience working in audio, or is this a hobby of yours (if you're comfortable sharing that information)?
Also, if I can put you on the hook for another question, what would you recommend in the way of headphones? My IEMs are nice but I dislike the feeling of my ear canal being sealed off, as well as the exterior sound isolation that comes with that (I'd like to be able to listen to music and still hear when my dog is barking at something, for example). I'm considering getting a nice pair of open-back over-ear headphones, but don't necessarily know where to start, as looking at headphone recommendations tends to pull me down a very deep rabbit hole.
The simplest way to put it is that I'm a music addict, and everything else is just a hazard of that affliction. :P I've never worked in it professionally (tinnitus) but I do have nearly perfect pitch. I also have more than a few friends who do work on this stuff, one of my old college buds built his own audiophile rig - and by built, I mean right down to designing it and cutting the wood for the speaker cases himself. He went with cherry, it really lights up the room.
Cans are hard to shop for because there are a million brands out there - but there's a trick to this. That ocean of brand names is just a relabel on a rather small pool of manufacturers, not unlike hard drives and computer memory, though not quite that small a pool. That's what you want to know - who made the speaker, not the wrapper. Headphones are also like speakers in that most brands kick ass and the endless arguments about them are pretty pointless.
Beware of gaming headsets, though - the ones marked for gaming are usually a lot of flash with zero substance behind it and a sickening price tag. Lot of scam-like behavior there, it's easy to get burned. If you stick with 'pro audio' brands you'll have less cruft to sift through, as krg linked in his comment. That's exactly how I shop too.
Picking a set of cans comes down to two things. The first is comfort - I have big ears, so I need big cans that can go all the way around and not press on them. I hate having sore ears after listening or gaming. You'll also have to decide if you care to have a gaming mic attached or not. It's been my experience that the cord will fray or the mic jack will fail long before the speakers do. If you listen on the move (jogging for example) you might want good quality earbuds instead. They've long been second class citizens for audio fidelity but I hear that's changing now.
The second is the speaker manufacturer. Sennheiser is highly lauded and for good reason, but you can't go wrong with brands like Beyerdynamic or Audio-Technica either. You can find their parts in a lot of other people's headphones - my gaming set is a HyperX Cloud that I only bought because it's got quality Sennheiser components, big cans, and a detachable mic that's built well.
I will say that there is a big difference between the ~$100 and the ~$300 variety when it comes to audio quality. The technical differences lie in the precision of the audio reproduction, and that does cost more, and most people can hear the difference. The gold standard for the last decade or two is the Sennheiser HD650, I have a set myself and that's my main when I'm listening to music. Must be fifteen years old at this point, definitely the longest lasting brand I've ever owned myself.
The best way to be sure is to listen in person - high end audio shops absolutely do have headphone test/listening areas, and you get to try them on too which is a bonus over ordering from Amazon just to find out they don't work on your head. Either way though I wouldn't obsess over it because at the $300 price point just about any brand kicks ass. I can't vouch for the $1000+ variety - those are intended for professional audio use in the studio, and that's not something your average person is going to need for recreational listening.
Not Amarok, but I would suggest checking out some pro-audio headphone options. (I'm personally checking out these). I'm generally a proponent for pro-audio gear when it comes to general music listening as there tends to be a bit less snake-oilsmanship there as the equipment is actually used in a professional environment.
If it'll work for your listening setup, I'd recommend a pair of high-impedance (250+ ohm) headphones, plus a headphone amp.
An ELI5 of what "high impedance" is and why it's better - all speakers and headphones are essentially an electromagnet attached to a membrane. If you use an electric current to make the electromagnet vibrate, the membrane vibrates with it, and sound waves are produced. Impedance is a measure of how much the magnet and membrane resist that current and "spring back" to normal.
So as an oversimplification, higher-impedance means higher potential for sound quality. Obviously there's a million other factors, if your sound source is a pirated 64kbps MP3 nothing will magically make it sound better. But in general high-impedance headphones, if driven by a high-quality amp, will sound better than the same source fed through otherwise equal low-impedance headphones.
The vast majority of low-end headphones, especially ones designed for mobile devices, are low impedance - 32 ohms or lower. This has an obvious benefit in requiring lower power consumption from the device, which is the right decision for the majority of users - they want longer battery life, and are probably listening on mediocre headphones in an already-noisy environment like a subway car.
Since the typical headphone jack on your phone or laptop assumes low-impedance headphones, you need a separate amplifier. As I type this I'm listening through a Schiit Magni 2. This plugs into your normal headphone jack or line-out port and amplifies the sound for the headphones.
There's also devices that act as a combination DAC and amplifier, such as the Fiio E10K. These present as a USB soundcard to your computer, which means sound is sent over USB to the DAC (digital to analog converter), amplified, and then output to your headphones.
There are also portable headphone amplifiers that are essentially a headphone amp with a built-in battery pack so you can carry them around along with your phone or media player.
Since you mentioned you wanted open-back construction, I'd recommend either the Beyerdynamic DT990s or DT880s (fully-open and partially-open, respectively). My "daily driver" headphones are 250ohm DT770s, which are the fully-closed version of those.
Only addon I would say for that is frequency response range matters a lot more than most of the 'more subjective' stuff. If the output devices are incapable of producing the frequencies required, it will affect quality a ton.
You won't generally run into that unless you're buying dirt cheap sketchy brands, though. That's the whole reason to get tower speakers that have a lot of smaller speakers in them - each is specialized for its own part of the frequency spectrum, which leads to a more faithful reproduction of the original sound.
It seems like with portable gear where space is at a premium, there are more trade-offs, though? It’s apparently not that easy to make a laptop sound good.
When I get back into designing portable electronic instruments, I’m going to want to figure out how to get decent sound without much size or weight.
Yeah, smaller is a challenge, triply so if the small thing has to put sound out to a big space. If you want loud that sounds good you really need the big speaker surface area. Headphones get to cheat since they are so close to your ear, and also blocking out other sounds they might have to compete with.
You might be able to compensate for that with a lot of tiny speakers that specialize, but you'd really want to take that up with someone who knows the math. Audio production is a science unto itself, if you've ever seen a couple of audiophiles talking tech it's remarkable how something that seems so simple can become so complex.
I have no interest in Tidal, and MQA was new to me, so thanks for putting that on my radar. I've decided to move away from streaming and go back to a curated library for myself, mostly via Bandcamp. Nevertheless, there are a handful of albums and artists I like that aren't available through that platform, and when I started exploring buying them from others I ran into this quality stratification issue, which is what prompted me to ask.
I think the encoding matters in the context of the music style and how well it was mastered. If the recording or master is poor, why pay the $20 for a copy from HD Tracks? I've found recently recorded albums that are great on my headphones, but the noise floor from the recording is extremely noticeable over my loud speakers.
I did the Qobuz streaming trial. Great website and good catalog. But no native Linux client. So I'm streaming from Chrome. It's not the same sound as when I use a dedicated player with ALSA, so why would I spend that much money monthly and use so much bandwidth when streaming if Spotify sounds nearly as good?
CD quality audio is my personal sweet spot since I know my hardware can take full advantage of it. That's not the case for a higher-than-CD quality audio. It effectively serves as a psychological barrier so I know that what I'm hearing is as good as it gets on my system and I can stop chasing "the highest quality sound". It also serves as a monetary barrier in helping me decide where I'm going to spend my money, since I'll rule out Hi-Res audio.
192kbps VBR is considered by many to be "transparent," eg: free of artifacting and sound degradation. I encode everything I rip at this bitrate. 256 can be downstepped to that, so I'd assume it would be alright.
I happen to think a lot of the audiophile thing is scams and placebos, you don't need to pay more for a FLAC when you can get a 320kbps MP3 with the same quality, or a 192kbps VBR MP3 that matches both of them for less storage. The biggest difference is the context you're listening to it in: location and hardware. If you're listening to it on a car ride or ear buds walking down the street, it's not worth the difference, and at home with decent cans on transparency is key. People tend to feel validated by their decision to buy the high quality thing, and I think have a tendency to justify the expense, lest they feel they've been taken advantage of.
As far as the vinyl vs CD thing, I side with Jethro Tull's Ian Anderson: It's a horrible format that causes a loss of a large amount of audio fidelity, quality, and frequency. With digital hardware, you'll hear what actually happened in the room, not what happened in the room lost to analog signal degradation (what is often referred to as "warmth," and nostalgic hissing and whatnot). I think the important part is accuracy to what was performed, rather than any sort of signal processing.
I’m fine with CD’s too, but it’s not because I think accuracy is important. As a musician, you’re not trying to make an accurate recording of how you usually sound when playing live. It’s like Hollywood or a magic show; illusion is everything and most kinds of “cheating” are okay to make a performance sound as good as possible. The results justify the means. A lot of sound processing has happened already before you get the recording, or even within the instruments themselves, so what’s a little more if you like the results?
Amplifying higher frequencies makes music sound harsher and rolling them off makes things sound warmer. How much you want is mostly a matter of taste.
It's funny, I have plenty of CDs from the 80s and they sound fantastic compared to any I've bought since. There was a point in time when CDs were young where the music industry hadn't fucked up the mixing process yet, and if you ever come across one of those CDs you'll be amazed. Your first clue you've stumbled on one is having to turn up the volume - a lot - just to hear them.
The CD format is just fine. Whatever madness is making them sound less good is happening before the pressing.
I would say my point is more about being able to get that result. If somebody likes the warmth and saturation of a McIntosh amp (they would if they spent $75,000 on it), more power to them. But if they say it's objectively better, or even transparent, like people will argue about analog gear, they're flat out wrong. It's simply an issue of signal integrity from the generation of it to the point the signal hits your eardrum. I will contend that I prefer perfectly flat responses so I can hear the audio as it was cut, but won't judge people who make adjustments, with with the sole exception of people who slam their highs up because they like the sound of cymbals ripping their ear drums to shreds.
Most audiophile gripes tend to be more about the mixing and mastering steps anyway, the loudness wars and whatnot, which can not be corrected by the signal chain. With digital, you can get away with louder mastering because you won't skip tracks if you make one too loud, but it's an issue that is fixed at the recording, mixing, and mastering stages. I would argue that mixing like that is also horrible (see: Death Magnetic and Californication), but it's also not a particularly hot take. This creates the illusion that vinyl is better, when the real issue is the A&R team is just trying to stand out on the radio.
I disagree. The sound that was produced for a record (prior to digital gear), on analog gear, is the intended sound. There was no 'raw perfect' cut. When the gold pressing was being made, that warmth, saturation, and timbre was a consideration on how to master the material.
I've got audiophiles of both the digital and analog side in my social circles. Dark Side of the Moon is a universal tester. And a well-cared for record with the equipment from that era sounds better than any version on digital gear.
Reality is analog. Digital, at best, is a lossy conversion from analog to digital and back. The various circuits to translate between the two can have just as much effect on the music as different analog gear can. There will never be a 'objectively correct' sound, even with digital.
This isn't quite true - the Nyquist-Shannon theorem means it is possible to digitize an analog signal without any effective loss, as long as you use a sufficiently high sample rate.
With an upper limit of human hearing at around 20khz, that means a 40khz sampling rate is sufficient. And it's no coincidence that the vast majority of audio ends up being sampled at either 44.1 or 48khz.
ya!
...not to self-aggrandize much, but @vord: check out the video in my post, if you haven't already. Really shows how a proper digital signal will correctly replicate its associated analog signal.
I think I meant that as more of a philosophical commentary. By all means it gets close enough for many/most practical applications.
Our bodies are analog. Our recreations of light/sound (especially digital) are illusions to simulate an in-person experience. Good ones, but still. The frequencies we can't hear (especially the low bands) influence how we feel the music.
Vinyl is not a superior format for audio, this is true outside of some very specific contexts (randomized alternate tracks, higher fidelity than anything pre-CD that didn't get a proper remaster). There's also something to be said for the ritual of playing a record, which is much more tactile than clicking play on an app.
But, the album art that accompanies vinyl is second to none. If only for its size.