The Price of Instant Playback: How Spotify Sacrificed Music for Speed (Turntable Tips series)

Browse Pro Tools courses...
Browse Logic Pro courses...
Browse Cubase courses...
Get the most from your studio with the Audio Masterclass Music Production and Sound Engineering Course.
Learn more...
@Gjoa1906: I suppose worrying about crap like this is a way to ignore one’s real issues..
@AudioMasterclass replies to @Gjoa1906: If you think this is crap then clearly you don't care about details. If no-one cared about details, what kind of a mess would we be in?
@Gjoa1906 replies to @Gjoa1906: @@AudioMasterclass Yes, I understand, but realistically seems unlikely people could actually perceive these ‘details’. The alternative being a mess (?) is a bit strong, don’t you think..
I enjoy your channel and these chats, and understand this is your livelyhood (audio, not the YouTube thing). I’m not trying to denigrate that. It’s about the fixation, whether this or something else, that some people have and use to ignore issues that truly affect their lives.
Keep at it and thanks.
@Gjoa1906 replies to @Gjoa1906: @@AudioMasterclass yeah, given it some thought and.., the subject of this talk is indeed crap..
@lxduc: I look at streaming services like the traditional radio station, to hear and discover music, and when I really like a album or a song, I will by it, on CD or vinyl and keep it for future days, or careful listening
@coffeeandshots: Most of that concern is misplaced.
Firstly, in a home setup, with a stable internet, slower internet speeds leading up to the streaming service streaming more compressed audio doesn't happen anymore except maybe in some parts of rural America served with the worst ISPs. But it's not a thing anywhere in Europe or UK. Furthermore, Spotify caches aggressively often, when you're streaming, a lot of times you're playing a local file cached in your device, eliminating slow internet issues.
Secondly, when this problem actually occurs, that is to say, outside, with a bad network connexion on your phone, you're less likely to hear the drop in quality due to listening conditions. And even if you were able to hear it, everyone would prefer a temporary loss of quality rather than outright pauses in the music. If the choice is between no music at all, interruptions or more compressed music, I'll take the latter everytime.
Thirdly, the "only keeping 1/10 of the music" line is bull and you should know it. 16/44 is 1411 kbps of which you can most often losslessly compress (like a zipped file) to 850 kbps most of the time. Most streaming services offer 256 or 320 kbps streaming, so you're keeping 1/3 of the music even in pure data size so to speak of. However this fails to take into account the human hearing system, at 256 or 320 kbps, with modern codecs, even trained listeners routinely fail ABX blind testing compared to the original except on very specific details on very specific parts of specific tracks, lossy compression at those levels is essentially transparent to the listener. More accurately, it could be said that you're throwing out data that you cannot hear and keeping 99.9999% of the music.
To make an analogy, no one is complaining about throwing out the yellow wavelength at 580 nm during movie capture and display and making due with a mix of red, green and blue wavelengths instead because our vision only sees red, green and blue and recreates yellow in our brains.
If anything, the biggest scam with streaming is "studio masters", there is no reason to pay extra for high resolution studio masters when most of them are actually the exact same mastering as the 16/44 version, but at 24/94 or higher bit depth and sample rates. Same compressed dynamic range, same tonal balance, same everything except disk space.
@Cypeq: Youtube as a music streaming service often does not dissapoint to help me find new good music to listen to, algorithmic suggestions are invaluable.
@AudioMasterclass replies to @Cypeq: The algorithm does sometimes have flashes of inspiration, which is probably why I found myself listening to Steps all afternoon https://youtu.be/OiwDHHcHPh0?si=MYHFtK0T1SWgm30q
@tomstickland: Have you ever made a video about MQA, the rather dubious concept that some DAC players support?
Someone did some trials of uploading some test tracks and running a spectral analysis on MQA and non MQA versions.
@tomstickland: I can't abide Spotify. I use Amazon hi def which is CD quality or above. CD quality is as good as I need, I can hear how much better that is compared with Youtube or mp3 format. Anything above that then I can't discern any difference. This is Amazon to a DAC Magic unit to Sennheiser hifi headphones.
I love the Amazon streaming service. I've been working through complete albums from the 70s onwards as and when I remember things or find reviews that stimulate me.
@soundssimple1: Streaming whether at home, or on the go is for convenience and for exploring and finding new music, bands, styles etc and expanding your musical horizons. Very few and I really think a very small percentage of people would sign up to a streaming service and expect an audiophile standard listening experience on their systems costing 10's of thousands of pounds. Streaming from whichever source is a product service for the masses at the monthly price of under one album or CD or whatever. My point is that streaming is a convenience and a tool, if someone wants the 'real deal' in a musical experience then pay for it and visit a live performance, concert or gig in your local town. Get the real 'sound' experience and keep musicians in work and show appreciation by supporting them directly. Stream global, support local.
@andreasboe4509: Lossless is possible with low latency. I just exchanged a number of 10Gbit per second routers with 100Gbit per second routers in an office environment and their capacity is insane. Lossless transfer of a 10Mbyte file only takes milliseconds, so a whole audiobook can be downloaded instantly to a listening device and be buffered and played without having to rely on a jitterfree network connection. Podcasts have worked like this for years. Click to download and the playback starts when the file transfer has been completed. It just goes faster now.
@AudioMasterclass replies to @andreasboe4509: The future has arrived then. It just took a long time. DM
@andymouse: Please do your test as I love experiments....cheers !
@therealwolfspidertoo: The so called lossless audio still sounds different. Not necessarily bad, just different, but yes, it does sound bad sometimes too.
@adamgarlow5347: I like streaming for discovery and qobuz download store for things I know I like. Tidal is the streamer I went with because subjective tests I've seen on YouTube have rated it well, it has direct output to my DAC from PC which spotify refuses to do, and the library seems sufficiently large if a bit less than spotify on some tracks I want.
@OrangeMicMusic: The reality is:
Without performing a double blind test no one can confidently say that they "know" from the beginning that 16Bit 44.1 KHz sounds better than 320 MBps mp3.
It's science and it's proven already.
@AudioMasterclass replies to @OrangeMicMusic: Yours is clearly an alternate reality to mine, and I don't begrudge you it. However, there is no way I can see degraded audio as anything more than a sometimes useful convenience, saving bandwidth and storage. And of course the streaming services can serve degraded audio at one price point and lossless at another, of course higher. DM
@OrangeMicMusic replies to @OrangeMicMusic: @@AudioMasterclass You are correct, no doubt.
I'm a musician and music producer (I can't compare to your level of work in any way) but my mind can't comprehend how a "diluted" version of something can be the same as the original. At the same time I can't hear the difference because I'm too old (I can hear max 13KHz) 😱
But I'm also a science enthusiast, and from a scientific point of view, there is a "different reality" for everyone since our auditory mechanism is very imprecise. Most of the people with untrained ear, or hearing depreciation hear no difference. A small percentage would definitely hear the difference.
Because of this, some of the streaming services are offering hi-resolution at a different price. Sadly some others fell for the idea that if the majority hear no difference then they will never offer this service, like Spotify.
@paulroundandroundandround: I never use music streaming services simply because of the deal the musicians get financially before even consideration of the actual audio quality of the streaming. I think the Bandcamp model is the way forward for musicians and listeners - artists properly rewarded and music given to the listener in any format they could digitally desire!
@AudioMasterclass replies to @paulroundandroundandround: I'm on streaming services myself and my royalty payments usually start with a decimal point then three zeroes. DM
@sbbinahee: How 'bad' is youtube sound? To my ears it sounds great, depending on the source recording.
@nsfeliz7825: the average joe dont care about lossless or lossy. dats da problem for music lovers.
@coraliefontesdeaguiar2333: Hey ! We just mastered a song with the band. I can send it to you so you can compare it with the version that is on streaming service. A good old A/B tells a lot. Great videos, big fan !
@SO_DIGITAL: Would it be revealing to take a WAV and an MP3 and subtract one from the other to see what the difference is that was lost? Perhaps make a video about Psychoacoustics and how lossy encoders take advantage of it.
@imqqmi: I've tried it with 'You're so cool' from Hans Zimmer to see what you're loosing vs original cd. I've chosen 5his song as I know most lossy compression schemes compress phase difference data the most, ie stereo image ie reverb.
The marimba and other percussion instruments produce lots of it.
I've captured the spotify stream on a pc with Audacitys loopback function and making sure normalization is turned off and volume all the way up.
After capture I managed to align the original and spotifys decompressed version by using a steep peak and line them up in the multi track editor, inverted the original and created a mix down version of both, which subtracts the original from the spotify version.
And I heard exactly what I was expecting (so this method really works!), distortion of the reverb of the percussion instruments and nothing else, purely the distortion of lossy compression. Training my ear on these sounds (which didn't even need boosting btw) I could hear the roughness and shortening of the reverbs after 4-5 loops of listening on a solo part. Now I can never not hear it so I basically ruined this song for myself while listening to it on spotify. Sacrifices have to be made for science lol!
@imqqmi replies to @imqqmi: On the pc and mobile you can download and so prebuffer the music. But unfortunately this isn't possible on most stand alone streaming devices like receivers.
When starting a song I can often clearly hear missing parts, skips etc.
How to test this? You'd need to monitor network traffic between the access point and receiver and check if there were dropped packets, it'll be extremely tedious and time consuming to do though.
@OrangeMicMusic replies to @imqqmi: When you flip the phase between two identical audio sources, you’ll be left with no sound.
In your example the sources are different and what’s going to be left out are digital artefacts. At this point it’s impossible to tell if what you hear is true (the distorted reverb) or the artefacts.
Also, capturing a song from Spotify which uses Ogg Vorbis encoding system, to wav format on your PC (I assume), you’re modifying the information again.
The only way to check if it’s any difference between 16bit 44.1KHz or 24bit 48KHz and an mp3 is to perform a double blind test.
Many audio studies emerging in the last decade showed that there is no significant statistical difference when listening between the formats.
@imqqmi replies to @imqqmi: @@OrangeMicMusic A codec applies a number of transformations and filtering which can be content dependent. They are all digital artifects in the broadest sense as the compression processing took place in the digital domain. Capturing on pc with a digital loopback doesn't alter the digital signal in any way unless you set volume below 100%, and have any effects turned on in the audio control panel.
Logically if you add the noise difference to the original non compressed signal you get what you hear from spotify directly. So listening to that difference alone without the distractions of the music allows you to hear what's changed with regards to the original. You can decide for yourself if it's acceptable or not.
Compression algorithms are designed to mask these imperfections as much as possible, most people will accept this as the 'truth' but that doesn't mean everyone will.
Not sure what you mean by 'no statistical difference between formats'. There's no statistical difference between a monkey and a human dna wise. Lies, damned lies and statistics, as they say. And what studies?
@OrangeMicMusic replies to @imqqmi: @@imqqmi a misprint that I corrected on my comment: “no significant statistical difference when listening between formats”. If I’ll have time I’ll find again those studies for you and post the links.
When capturing a song from Spotify and save it to your PC, the file saved is different than the original no doubt here. You’re taking a format from Spotify, pass it through sound card, and save it to your PC as a different format. Plus, two AD DA conversions are involved here, using the DAC on the sound card. So, changing the audio format plus two ADDA conversions will result in a different file from the one on Spotify.
In the end you’re comparing the song from the CD, with a “few times altered version “ of that song from Spotify.
And flipping the phase will just reveal what you already knew- the files are different. It’ll not reveal that a file is superior in quality than the other.
Thing is, I see this reluctance from people on YouTube when comes to a double blind test. Knowing what you’re listening to, will always result in a biased conclusion.
Better, take 30 songs from various CDs, encode them as MP3’s and then load them on a plug-in for blind tests, for example 4U+ BlindTest from Hoffa Plugins. This plugin anonymize and randomize the files and you can have your own conclusion.
Cheers Mike
@imqqmi replies to @imqqmi: @@OrangeMicMusic No that's not how a digital loopback works, it doesn't pass through a dac and adc chain, that was the point of a digital loop back, it's a hook into the windows drivers where you can capture the digital data from any audio source and write it to disk. Try it yourself with Audacity, play a ripped cd and capture it with Audacity and its loopback function, subtract from original and the result if aligned correctly is zero (32 bit float rounding errors not withstanding). In my case it's output to an external receiver using hdmi, and the receiver doesn't have an adc. But that's irrelevant since the signal path between spotify and Audacity doesn't touch the receiver.
As for no 'significant', that depends where you place the threshold.
I can pick out this particular song 10/10 with blind abx test. With random songs 7/10 since I don't have an audiophile grade setup and some songs compress particularly well without much loss. Just 4 diy 3 way speakers and a cheap denon receiver (x1500).
@genericusername5909: Spotify is the best for portability because with environmental sounds you can’t hear the difference anyway. For literally every other context they are the worst and I wish my tech was compatible enough to leave
@gehanameresekere1345: I find Spotify better than mp3.
@AudioGuyBrian: Buffering in the middle of a song is nice too. Streaming sucks. Give me a cassette I recorded on my Nakamichi or an LP any day.
@hifirulezzz: 100% true, I frequently mention dropouts with change a bit in a song! And both Apple Music and Amazon Music are doing that on Hi-Res Music! I thought that this was Digital Rights protection - i.e. song encoded to the player. But your point is definitely worth taking into account.
@hifirulezzz replies to @hifirulezzz: Yes, and this is on the 1Gbit internet channel btw.
@Waffledogchat: In my opinion, anything after the master tape is impure. Whether you like cd lp cassette or Spotify is OK as long as you like it. I had to turn away from true hifi as I kept attempting to pursue perfection, when in reality it doesn't exist. It spoilt my listening of music and I became hifi bonkers. I don't mean to sound dismissive, but please folks, just enjoy the music, but of course, set the Spotify to high bit rate if you can afford it. Good video as always. Mark.
@nickwallette6201: As someone who does understand how the Internet works, I can't make heads or tails of what that dramatization is trying to convey. It all just looks like the stupid computer drama trope: "Let's pretend developers spend their time huddled around a frantic whiteboard session until someone has a stroke of genius, takes an exhaustively long time scribbling out a nonsense analogy that their peers wouldn't need to understand the concept, until another developer catches on and finishes their sentence, and then everyone erupts in raucous applause."
What "red light" problem are they talking about? There's definitely a delay between a request from a client and a response from a server, but the solution isn't to "run the red light" -- whatever that's supposed to mean in this context.
The ONLY thing I can think of is that it might make sense to transfer the first second or two of the song in a reduced bitrate to help ensure timely and complete delivery before the buffer fills up enough to smooth over any gaps. I would also think UDP would help here, since you can fudge a lost packet or two, and there's no benefit in receiving a packet of audio data after that region of the song has already been "played" with whatever data existed on the client at the time. Other things might help, like interleaving chunks of audio data, so a lost packet isn't a lost portion of time, but just lost detail in a slice that already had other bits of data delivered previously, etc... I'm sure it's hell on transients, but it's not like those exist in modern music anyway. ;-D
Final gripe: I really hate when TV and movies try to illustrate technical concepts like this. It's "exciting" ... I guess ... but if you're going to make up some utter BS, at least make up exciting BS that has something to do with the actual problems they solved. The normies won't understand it either way, but at least it can be understood, and when it is, the viewer would appreciate its attention to detail. This nonsense just sounds like the service crammed packets into the Internet tubes harder, and bowled over some Hotmail packets in the process.
@igorbeuk4068: I have to say this because i worked in Collab with a telecommunications engineer who works on a 5G network, so I know that the Internet is not capable of wide traffic distribution and 5G would be a solution, but Huawei is banned and they are already working hard on the new to a system that is even faster than 5G, but so that the internet is unison and not a hybrid Repeater on every other building and the rays are bad for health... 5G in Europe and America is not even close to 5G, but as announced and unfortunately it is not possible to create an adequate infrastructure, lossless is a solution for many regions because people should not live only in cities.
@igorbeuk4068: So i have to sucribe to another streaming service to watch it 😂 but if EP 6 is focus i already understand because # 6 is number is anouncing some trouble
@Harjawaldar: I've tested Tidal by using my Denon receiver (with Heos integrated) to stream a song from tidal, and then listen to the same song with a CD transport (NAD, optical cable) - same DAC. Unfortunately for my wallet, this has only increased my affinity for CDs. In most cases they sound a lot better!
@rezsbc: I'm not sure how much I'd read into that clip. I'd be surprised they aren't just using TCP + buffering. Sure they'll lower the bitrate if the data-rate is so low it's that or suffer a stop in music. But the TCP + buffering thing was long developed (likes of youtube) by when Spotify launched. In terms of MP3 / lossy it's mostly fine. I happen to opt for lossless Deezer myself, but either way I think mostly you "get what they send" and it's decent quality. Audio is quite low bitrate in the grander scheme of things, it's not been a technical challenge to stream since the demise of dial-up and 2G.
And streaming services? In general they are great. You can find most everything, make your own playlists, all in good quality. Much easier than organising a large music collection on CDs or Vinyl and trying to remember where everything is / goes, and vastly better than trying to do so through file system folders or iTunes library style software. Streaming is where it's at tbh. Although I still love a 12" single :P
@melaniezette886: Frankly most people don't care at all about audio quality
@melaniezette886 replies to @melaniezette886: How many people stop doing anything while listening?
@melaniezette886 replies to @melaniezette886: Lots of tests have been done, hq compressed audio can't be distinguished except by very trained professionals. And even in this case you no longer listen to the content. Though I listen to flac on good old hi-fi components system for decades from LPS to files...
@anotherdamn6c: The buttons. So many buttons. And then I have to look closely and wiggle the mouse, choose the path, type, whatever. Make sure all the subscriptions and permissions are positive. By the time I get it all straight the cat has taken my chair and the wife has found the 'honeydews'. Even if I've streamilined the process it's still so damned complex that 10% of my listening brain is under an umbrella of uncertainty. A few clicks or hiss from the TT is a small price to pay for the distracting mental overhead. I LIKE separating the computer and the Internet from my music. It's why I learned to play music myself and why I close my eyes to listen.
Thanks, DM for the thinky fodder. Subscribbled.
@AudioMasterclass replies to @anotherdamn6c: I agree about the computer. I work at my computer all day and the last thing I want is for it to meddle with my relaxation. DM
@lukastemberger: This is so wrong. Compressed music is not a "tenth of the original song" just because the file size is one tenth. At worst you lose a few percent of the actual track.
Try compressing a wave file into 320kbps AAC. Then convert that AAC back to wave and compare the two waveforms in software. They'll be over 99% exactly the same and if you listen to them side by side there's no chance you'll hear a difference.
@dancemusicorganisation: Spotify royalties starting with .000 😂, I laugh, but it aint no joke to us artists, who can't afford the electricity it takes to power our devices in order to make the music in the first place, due to the minuscule amount received from these platforms!
@AudioMasterclass replies to @dancemusicorganisation: My royalties commonly start .000 - Probably Sir Paul's or Sir Elton's start in a much grander way. DM
@dancemusicorganisation replies to @dancemusicorganisation: @@AudioMasterclass Yeah, I get it, I see the same stats from my distro. Though, it's not as lucrative even for the successful artists. DEFECTED RECORDS attended BMC 2019, and gave a talk about the current situation of streaming, and they claimed that one of their tracks received 19M plays on Spotify and they received £26K , it is only the fact they still run events and are heavily into the Ibiza scene and associated clubs, are they able to break even, not make a profit, break even! Can you imagine the royalties from selling 19M records in the 70's / 80's / 90's and even 00's!
@JohnFraserFindlay: I thought I saw Dr. Strangelove hovering in the background of that clip……..
@TheExtremenarcissist: Many people using streaming services are now using Bluetooth headphones to listen to the streamed audio. Is there further loss of information when the streamed audio is sent to the headphones over Bluetooth?
@AudioMasterclass replies to @TheExtremenarcissist: Bluetooth audio has come through some evolutions since its inception. It's still lossy compression though. I'd like to be able to test it but the complication is that I'd need a headphone that has both wired and Bluetooth. They do exist so it might be possible for a future video. One interesting thing I heard is that if you use Apple's AirPods, and possibly other Bluetooth earbuds too, your device transmits both channels to one AirPod which then transmits the opposite channel to the other AirPod *through your head*. DM
@TheExtremenarcissist replies to @TheExtremenarcissist: @@AudioMasterclass thanks for the reply. Just discovered your channel; great info.
@dfpolitowski2: As long as its cd quality I'm more then happy. When CD came out music quality for me when off the charts, I could hear words, and high playing instruments I didn't know were in that song before. .
@Misfitnz8: I hate streaming , owning the physical music is the only way to go
@SrEstebanG: This is why I have abandoned spotify and searching for other streaming services like tidal. Also not good but far better quality than spotify.
May you explain why samsung does not implement last aptx loosless codec on its devices?
@jeffty60: As a lifelong musician and recordist of a 'certain age' I still prefer my library of SACD, Blu-ray and DVD-A discs along with CD's for serious listening, something I make time for daily. That being said I recently started subscribing to Apple Music+ mostly because of the better quality of their lossless streams plus I find some of the Spacial Audio mixes to be a fun enhancement, although not always. Their vast library is also a big plus, it's rare when I cannot find a vintage album when I want to. I enjoyed watching "The Playlist", I thought it was likely pretty close to the reality of launching Spotify. Thanks for your insightful videos, I really enjoy your wry sense of humor...
@mattmarriott5506: I'm not sure your comment about compressed music sometimes being only 10% of the original is quite right. My experience of early mp3 was that the raw wav would be in the region of 50mb and the mp3 would be in the region 5mb, a 90% reduction in size, yes but once decompressed, not 10% of the original signal. That wouldn't be compression, just removal is information. I suspect if you do carry out an experiment on modern streaming services you're going to get the "boring" answer, they're to all intents and purposes as good as the original. I'll be interested though.
@AudioMasterclass replies to @mattmarriott5506: Thank you for your comment. I'd have to say that if 90% of the data has gone missing, then only 10% of the music is left. But as humans we can't hear most of the 90% that is lost. We're getting into philosophical territory here but I'd probably have to conclude that MP3 works amazingly well considering what it does. DM
@nickwallette6201 replies to @mattmarriott5506: @@AudioMasterclass I have to agree with Matt, here. If I tell you "1 + 1 = [ redacted ]", are you now missing 33% of the information? Technically, yes, but it is fully recoverable -- so in essence, you still retain 100% of the data meant to be conveyed.
This is more of an analogy for RAID than perceptual encoding, but it illustrates the concept well enough, I think. Storing the actual value of every sample is a rather inefficient means of representing audio data. It's convenient -- because seeking to a given sample is deterministic, and editing one sample has absolutely zero impact on the ones around it.
Other forms of storage may reduce the bit-depth of particular samples, with the assumption that you don't need 16-bits to convey the delta between any two adjacent samples, unless you're encoding square waves at the sampling rate. The data size is smaller, but the result can be perfect. Stuff like MP3 and AAC shift the data into the frequency domain, so we're not talking about samples, but sinusoids at various frequencies and amplitudes. This will always be lossy, as the translation isn't flawless, but with high enough detail, it can be essentially transparent.
Given realistic bitrates (which sacrifice perfection for size), once converted back to PCM (which it has to be for playback), the difference is not really 90%. It was just transported more efficiently. It was a temporary reduction in data that resulted in a permanent, but much less substantial, reduction in actual audio fidelity.
It can be difficult to measure the actual loss, since there is some disruption of phase and so on, but I've done tests inverting a decoded copy and summing it with the original, and IIRC, a 192kbps MP3 encoding in quality mode with a version of LAME that I used some 15 years ago yielded a difference signal of around -60dBFS on average. Hardly anything to get excited about.
@xaverlustig3581 replies to @mattmarriott5506: @@AudioMasterclass Looking at the raw data is misleading. A useful analysis is a comparison of the waveforms original vs compressed. There are some videos around where someone did that, computing the difference waveform. It's not zero obviously, but some noise that pulses with the music.
@EddyTeetree: We pay for Spotify. I do use it for background music but mainly its a great way to check out new music before buying the physical recording which in our case is vinyl. we dont own a streamer or even a cd player although we would love a B&O with those sexy sliding glass doors!
@EricB256: It's interesting to know that streaming services automatically switch the user to a lower bandwidth version of the content when the bandwidth of the internet connection goes down. Since I have more than 90% of my music on physical media, I have never thought about that before. But it is totally understandable why streaming doesn't work any other way. It's kind of like when your LTE phone automatically switches to a GSM frequency when you're driving in between cells and the cell in the location you are entering doesn't have any more open slots of LTE available: your connection is still going but it's way different than before.
For me, streaming is merely the current era's "pre-listening session before buying" which in the heyday of the CD era, people used to have at players inside of stures. Other people use it as personalized radio. Will those even notice when a song gets deleted (for whatever reason) from the catalogue of the streaming site and doesn't come on anymore? They probably won't if they only use the random button / AI playlisting tool, but rather only will if they remember that the song used to be in a playlist they created by themselves. In the late 00s, I created a handful of playlists for other people to enjoy. Suffice to say they didn't last the proverbial lifetime of listening enjoyment.
@mcrunk1977: Im a bit different I guess! - I listen to Tidal Hi Fi through an early Naim 1 box streamer with a CD player in it. Tidal beats the CD. I ripped all my CDs to 360 MP3 or flac but only ever listen to the hard drive recordings if they are not on Tidal. Every now and then I think it’s so good I can get rid of my vinyl until I put a record on and realise how good it could sound.
@bocaratonopera409: I love my Qobuz subscription! I have a trained ear and I cannot say that it sounds any worse than a CD. I would even say that some recordings seem to sound even better! But of course, I can't scientifically say this is true. I can assure you all, though, that soundwise, I miss nothing free from a CD or even an SACD, DVD-Audio, Blu-ray Audio, etc. listening to Qobuz (except for the multichannel content, which streaming does not offer). Of course, when there's a bad recording, the streaming isn't the problem. But comparing superior quality recordings, I'm very happy with Qobuz.
Cheers!
@6teeth318-w5k: Back in the old days, you had shops who generated revenue. people who had work, streets that thrived, meeting other people. Today, Spotify gets all the money and pay very very little if any tax in many countries. They are keeches.
@davidaston5773: I find that Spotify and YouTube are the ultimate jukebox. ENDLESS music and non music audio.
The problem you lose the feel of a Vinyl record (album or single), cassata or even CD gives you a connection to the object. Plus, online music doesn't have the ability to become a piece of art.
But, the good thing, it will create new fans of a wide variety of artists who will buy their music on vinyl, cassette and CD.
We are now living in an age where we can listen to music on ANYTHING. And anywhere.
Music isn't dead. Just everywhere.
@cjay2 replies to @davidaston5773: 'new fans who will buy their music on vinyl cassette and CD'? Seems to me, they'll just stream it like you do.
@nickwallette6201 replies to @davidaston5773: @@cjay2 This is the problem, to me. It's so "easy" that it really does make any physical distribution completely pointless. And then, if everything is only available by subscription to a service that has to have the license to keep that music in their library, there's absolutely nothing to stop large catalogs of music from just vanishing at the stroke of a lawyer's pen.
So, no, I'm not on board with streaming. Physical for life. The cost is too great, in the long run.
@SimonBlandford: I use Amazon Unlimited (lossless) and listen to non-mainstream music which on the whole sounds fantastic. The problems arise when I want to listen to something like Pink Floyd and can only find super-duper remasters which usually mutilate the sound in an attempt to be heard to have done something. The best versions of the classics I heard were late 80s such as the Led Zeplin remasters done by Jimmy Page, the CD with the crop circle on it. That was before digital denoising came along and ruined everything.
@AudioMasterclass replies to @SimonBlandford: In my opinion we've had enough re-mastering. What we need now is DE-mastered releases. DM
@moritz_schoenermann: As someone who was lured to realease songs on streaming, my first issue with it is how bad they pay, specially spotify. Could mixing your mastered song with a phase inverted version of the streamed audio of it be enough to point out the diferences?
@AudioMasterclass replies to @moritz_schoenermann: Thank you for your comment. Inversion can be an interesting test but it doesn't work if there are timing differences, which there may be due to temporal masking. The results could be misleading. DM
@moritz_schoenermann replies to @moritz_schoenermann: @@AudioMasterclass Does temporal masking also occurs with downloaded audio? :)
@AudioMasterclass replies to @moritz_schoenermann: @@moritz_schoenermann Temporal masking is a feature of lossy compression. I don't know whether there's a codec that doesn't use it so my answer, subject to any possible further insight from someone with higher knowledge, is yes. DM
@moritz_schoenermann replies to @moritz_schoenermann: @@AudioMasterclass When I read about conversion and all things involving streaming, temporal masking was not described, for sure they would not include something that could show loss of quality.... :/ I know for a fact that my original files and the ones converted for streaming differ a bit on BPM. The converted ones are a bit faster. thanks ofr the knowledge. :)
@goat-of-neptune: what do you think of dolby atmos, specially on the Jazz rekeases from blue note and some ckassical recordings? is there a solution yet for the differences in loudness, when comparing them to the stereo versions? A 5.4.1 plaback is a lot softer than the binaural rendering, and even the renderers from apple and Dolby differ. I would love to hear your comments on that.
@editingsecrets: It comes down to the purpose of the design.
McDonald's was designed to answer the question, "How can we use unskilled teens to provide a consistent hamburger at the lowest price and make a profit?" It was never intended to answer the question , "What is the best of nutrition for each individual's best health ever?"
The Internet's TCP/IP standards suite was designed to answer the question, "Given trusted people inside trusted institutions connected by trusted networks, how can we share computer resources even if connecting nodes are randomly up and down?" It does that brilliantly. It was never designed to answer the question, "How do we protect network resources when bad people can access any of them remotely?" and we are paying that price today.
Spotify was designed to answer the question, "How do we maximize minimally adequate, lowest cost per user audio delivery to listeners who prize fast response and convenience and have minimal interest in highest quality sound, and we do so at a profit for us?"
Neal Young tried to answer the question, "How do we design digital audio for highest sound quality even if that's more expensive and somewhat less convenient?" and not enough people cared about that answer. Spotify also wasn't designed to answer the question, "How can we maximize the opportunity of musicians to make a living?"
For many people, "I can tell it's got a good beat and I can dance to it" matters more than anything about perfect sound quality.
@markvanosdol8120: CDs are my choice, but Spotify works for party nights where tending to the music interrupts the games .
@editingsecrets replies to @markvanosdol8120: I think you're Spotify's perfect customer!
@peters7949: Apple Music (on Apple TV & Mac) often cuts off the start of a track, due to the fact it takes about 1s for the audio path to un-mute when one presses play. Not always an issue when an album’s subsequent tracks play.
This affects both streamed audio and audio stored locally.
@cjay2 replies to @peters7949: Because they don't care and you pay them for this.
@drbassface: I noticed uploading a HD Master of one of my originals sounds better on SoundCloud than on YouTube… both streaming but YouTube must cut down the quality.
@AudioMasterclass replies to @drbassface: This is something I might look into in a future video. DM
@drbassface replies to @drbassface: @@AudioMasterclass Cool. Here is my Remix on YouTube. The SoundCloud link for the same song is in the video description….if you want to compare… https://youtu.be/Fti8jGYxOL8
@teekay_1: So here's the deal. Spotify didn't invent this, the internet supports streaming using a protocol call UDP which doesn't care if packets get lost. Every bit of music and video online follows the principle, and it's baked into the basic design. Assuming this is sinister (or even novel) just displays a lack of knowledge about how i.e. the internet works.
Even when using "lossless" music services, they use the same exact design as spotify. So it's never literally lossless, unless you're streaming ripped CD's from your server in your home. When these services say they're "lossless", it means they're streaming 16-44 audio (CD quality), not that they don't lose data. Some lossless services offer up to 24-192 audio, which is theoretically better than a CD., but even those are subject to the rules of traffic on the internet.
@AudioMasterclass replies to @teekay_1: This isn't my understanding of the clip, which I don't think explained the issue fully anyway. It isn't about lossless versus lossy, it's about how quickly the music starts to play. In the clip, it's said that the standard internet protocol is what's causing the delay, so they find a way to work around that, causing some of the audio, or perhaps some of the resolution of the audio, to be lost, presumably at the start before there's time to buffer. But that isn't the real issue here, it's the fact that - according to the clip - no-one will notice and that doesn't matter. I think it does matter and as technology improves I expect to hear the whole of the music, exactly as the mastering engineer heard it when it left their studio. DM
@Richard-bq3ni replies to @teekay_1: Actually Spotify uses TCP not UDP. So packet loss is not much of an issue.
@Kwippy: You and I care about the fidelity of the music but for the vast majority of people out there, it only has to be good enough. What should matter more to every one, is what music you choose to listen to and how accessible it is.
But now that music is inexpensive to listen to and so widely available, the value has declined for most people.
@editingsecrets replies to @Kwippy: Billiionaires of media access hardware and online services have successfully and ruthlessly commoditized their complements. Complement is whatever you need to add to something to make your experience completely satisyfing. Commoditize is to make it interchangeable and only sellable at the lowest possible price, regardless of how many providers it puts out of business.
Car makers haven't been able to make gasoline free or only a penny a gallon, even though that would make cars more appealing. Washing machine makers haven't been able to get soap makers to accept only getting a penny a packet. But companies that provide our equipment and services for music have succeeded at making music essentially worthless, boosting the value of our gadgets and subscriptions.
Meanwhile Spotify's billionaire owner haughtily lectures musicians that they just need to do better if they want a living from their cute little irrelevant creativity which boosts his service's value proposition.
Henry Ford eventually realized he needed to pay the people who built his cars enough to buy the cars they built. There's no Henry Ford for the music industry yet.
@MostlyBuicks: I use Spotify to determine if there is any music I would like to buy on CD. But now most new music isn't even released on CD. I have 5000 CDs, but now I MIGHT buy 3 or 4 CDs a year now. And I NEVER buy dowloads.
@CraigPMiller: I only stream when fidelity isn’t the issue. 🧐🤨👍
@focuspixvideoaudioservices68: its practically unimportant...having said that Soundcloud messes with my low end! I'm guessing data transfer will become cheaper ..
@Seiskid: CD players do the same thing. If they mis read a section - and the error correction can't rebuild it - guess what - they mute. The theory is provided the mute is less than a certain fraction of a second, you won't even notice it. You'd notice the corrupted garbage. But not the silent section. Especially if you are not expecting it to be missing.
@medonk12rs: After 10 years with Spotify I tried Qobuz and Tidal last year. Strangely enough, Qobuz sounds better - even when arguably the same album / album version / master is played. Maybe has to do with normalization and/or watermarking.
Anyway, happy Qobuz user now. 🙂
@daveuk1324 replies to @medonk12rs: Agree! Me too.
@thepuma2012 replies to @medonk12rs: I listen via Qobuz too, I think they sound better because they use lossless FLAC, tidal s mqa seems to be not lossless. Only comment on qobuz I have there is difference in how loud tracks are played.
@medonk12rs replies to @medonk12rs: @@thepuma2012 ... in my comparisons between Tidal (FLAC / non-MQA) and Qobuz (FLAC) Qobuz always won.
Since I don't have MQA capable equipment, I cannot comment on TIDAL+MQA.
@thepuma2012 replies to @medonk12rs: @@medonk12rs ok thanks
@teekay_1 replies to @medonk12rs: @@medonk12rs My DAC supports MQA, but I haven't tried it since it costs significantly more than flac 24-192, and I don't see how it could be better.
Plus, didn't MQA just declare bankruptcy?
@DWyn-xq4yf: Besides listening to or watching music videos on Youtube, I do not consume music through streaming services. You lose quality for convenience. I check stuff out on Youtube to get an idea about what I am investing my money into when buying music. I want quality.
@carlosquijano2827: Frankly I don’t listen any difference between my CDs and the Spotify Premium version…
@DAMN11KIDS: What about downloading lossless? :o
@editingsecrets replies to @DAMN11KIDS: When it eventually arrives, you have your own lossless copy on your own device. What you don't have is instant playback of the other billion songs in the streaming service's library, which is the point of the video.
@alvaromedinagarcia: I don't care if some of the music is lost, as long as it's not too perceptible. I use Spotify (and Youtube, BTW) for convenience. I want the real deal I pay the download or CD. But perhaps that's just me, who was used to "guess" the music over the hiss in 10th-generation cassettes.
About the tests: I uploaded myself the tracks of my band's albums to CDBaby/Spotify, tracks that were ripped directly from the original CDs, burned from the master in a professional facility. When the tracks showed in Spotify I did an A/B test and didnt' find differences. Of course it can be some link in the chain that degraded the master so it didn't matter, or just I have bad ears.
@CORVUSMAXYMUS: We are free to choose.We live in a free world.If we want quantity over quality YES choose streaming ( which is open to all kind of errors( data loss, internet speed, etc ) if you want want CD quality dont be penny wise and buy CD .
@ALOISC1 replies to @CORVUSMAXYMUS: "It's a free market, Manny. Not a free world." - 2 Guns (2013)
@rogermuggleton8127: Practically all the music i hear is classical music, and my reference is the live concert. You might think this ensures pure undamaged music but of course it does not. The listening experience will vary depending on where I sit (or stand), the people around me, other extraneous noises in the concert hall, and where I am looking. If I move my head the sound balance changes. If I stand at the front then I am usually nearer to the leader (violin) than most of the other orchestral players, each of which hears a different balance. If I stand in the middle of the hall, or sit at the back, my experience will be different. If I sit on the floor, the sound will change, as it will if I rest my head on the back of my seat. And being long past retirement age, my ears distort very loud music, and filter out the higher frequencies.
This summer I will travel hundreds of miles to listen to two or three dozen classical concerts at the Proms in London. I will stand at the front, among friends who I've known for years and who all love the music. The experience of this greatly surpasses any musical or audio imperfections, and even the ache in my legs and the various smells emanating from others there!
Listening to recordings is very different, regardless of the recorded quality. But you can still be transported to another place, a place that only those who allow themselves to be moved there by the music will know.
@frogandspanner replies to @rogermuggleton8127: I find that live music has distortions: the outer ear distorts, and at Symphony Hall, on the Ledge, positioning of my head relative to the seat was important. I use the past tense because I have hardly been to a concert there since Rattle left, and his band was made to play easy-listening classical music.
The greatest distortion was at the climax of a piece that had four percussion sections, and the percussionists and brass donned ear-defenders before the four-gong conclusion ©. If anybody has any suggestions as to what that piece might be I'd like to hear them - and hearing was more than I could do after those gongs.
@fredjohnson9856: No self repespecting audiophile listens to Spotify. And you're right, if ppl understood how the internet works, specifically TCP/IP, they wouldn't worry about losing packets. VINYL FOREVER!!! and some streaming is good too...
@teekay_1 replies to @fredjohnson9856: "No self repespecting audiophile listens to Spotify."
Even in their cars? And they likely use UDP not TCP. Probably most services do that. Look up the difference between UDP and TCP.
@fredjohnson9856 replies to @fredjohnson9856: @@teekay_1 that was all tongue in cheek - glad to see somebody knows networking
@teekay_1 replies to @fredjohnson9856: @@fredjohnson9856 When it comes to audiophiles, you can never be sure.... ;)
@Richard-bq3ni replies to @fredjohnson9856: @@teekay_1 Spotify used TCP not UDP
@zibikonti8901: I have listen to streaming service for last 15 years. First Spotify until ~2014 but I found the quality of the music, well spotty. Migrated to Tidal Master but confused with MQA controversy I started to compare tracs with Qobuz. And I found Qobuz the best in terms the quality of music. HiRes is great but honestly I am not sure that I can tell the difference between well recorded lossless CD standard and HiRes on Qobuz. I hope and believe that I get real lossless CD or better honest stream of all true bits delivered to my PC on time. However, with the upgrade of my gear to HiFiMan HPhones and RME ADI-2 Dac I have started to question the quality of the recordings. I think that it is fascinating subject to cover. We all are focused on the quality of our gear (streaming/transport, DACs, Headphones amps, power amps & speakers) but what about the quality of the master recording? Proper mixing? I dare to say distortion on the master ? Why most recordings are so excellent (examples: Sting, Sade, most of Miles Davis, Pink Floyd) and other sound flat, compressed (emp: U2) or even poor and distorted (old Simon & Garfunkel recording). I am not complaining about musicians who use distortion as a way of creative expression like John Fruciante who distorts his guitars and / or vocals quite frequently. I hope on purpose :) Interested in your comments.
@6643bear: Hi , another interesting and informative video, it transpires that Spotify on the premium sub is ACC 256kbits/sec, how ever they we’re supposed to bring out Spotify platinum or Spotify hifi back in 2021 but still waiting. Also talking about in charging $19.99 for this service .regards mark
@AudioMasterclass replies to @6643bear: Lossless on Spotify has been a long time in coming, and may still be a long time yet. We live in hope. DM
@Anybloke: Interesting. Yesterday, I listened to One World by John Martyn via Spotify and simultaneously recorded it using Audacity. I bought the LP back in the 80s btw. I burned a CD from it and listened to it on my hifi expecting it to sound better than the record. Needless to say, it didn't (apart from the lack of clicks and pops etc). If anyone's interested, the Spotify app for Brave browser means you get it sans ads without a subscription.
@rabit818: I have a Bluetooth speaker in the bathroom and listen to Spotify when I shower. It’s not for critical listening.
@2134yanto: I sold/donated/gave away my system and over 1500 items of physical media 18 months ago just before we emigrated.
For me now, streaming is my only option. I use a pair of Kef LS50W2 with 2GB Fibre internet. Obviously I am unable to compare cd/lp and stream but suffice to say that I’m extremely satisfied with the experience I get from this method of listening.
I completely understand those who will only ever want physical product - been there, done it, had the t-shirt. But sometimes priorities change as do necessities. As long as I can listen, I’m happy.
Also I don’t consider myself an audiophile but a musophile
@QueenFondue: I think there's something to be said for just how EASY it is to find music using streaming. It's all free, or if you want no ads for a low subscription. It's all easy to access. A WORLD of discovery is at your fingertips, and there's so many countless playlists for different genres and artists and vibes that you just couldn't realistically achieve on a physical medium. Streaming is an absolutely fantastic way to find new music the exact way you want to.
When I've listened to an album enough on streaming, that's when I start to think about graduating to a physical copy. I think using streaming to find new music and CDs or other physical media to keep the music you like is honestly kind of the perfect system, because you're not locked into buying an album just to listen to it only to realize you don't like it very much.
@AudioMasterclass replies to @QueenFondue: I can agree with this. If I see a film I like on Netflix, Prime etc, then if I really like it I'll buy the blu-ray. DM
@maidsandmuses replies to @QueenFondue: I totally agree with the try-before-you-buy policy. But I only do that using "free" ( = ad-supported) services, such as this platform. Then if I really like it, I buy the CD. (DVDs/ BluRays I rarely buy, I prefer music to films)
The music streaming subscription cost may seem low like you suggest, but with so many "services" now moving to a subscription model, it all adds up at the end of the month. Here in the UK it is all too easy to start spending a thousand Pounds per year if not more on subscription services if you are not careful!
@AudioMasterclass replies to @QueenFondue: @@maidsandmuses This is what I said in the 1980s, that there would come a point where people would pay £1000/year for their home entertainment. The cost of music streaming is low, but the providers surely must want to segment the market so there's a rock channel, pop channel etc, all individually subscribed. And then $10/month for Taylor Swift, $10/month for Ed Sheeran. £1000/year might be cheap compared to what's coming. DM
@maidsandmuses replies to @QueenFondue: @@AudioMasterclass Indeed. Add cloud services for your music "downloads" and your family & holiday photos and it is easy to predict where this is all heading.
Call me old-fashioned but I'm still buying CDs and storing my own data on my own storage media; that way they won't have me over a barrel (quite so easily). I don't need instant access to all my digital data anywhere in the world (when did that become a necessity anyway?) 😉
@cjay2 replies to @QueenFondue: There's a reason it's all 'free' and easy to access. Figure it out.
@robertwestinghouse4098: Cannot see the value in streaming service. Call me old.. I like to have the music on my PC/CD and play it when and how often I like
@vincesnetterton5868: most people nowadays want wireless, and bluetooth, and mp3, not hifi.
@worthingtonmodelrailway8628: I have with the advent of lossless availability been streaming more than I used to! However it has quickly become obvious that the streaming files are not the same as the CD versions. Many stream files are marked as “remaster” and have basically been f’d with by the streaming service. I enjoy the simplicity of streaming but sometimes am disappointed with what I’m hearing. I’ve started using streaming to preview what I’m going to order on CD knowing I’ll probably enjoy the CD more!
@cjay2 replies to @worthingtonmodelrailway8628: Unless they f'd with the CD as well. Pretty likely these days.
@sportscarnut: The test idea is great. Just do a bit difference.
@tbrown6559: Oddly, I often find Apple Music to sound the best of the three I use, Spotify and Amazon music being the other. Go figure.
@qtduck7082: "When we're listening to streaming music ... we're listening to about a tenth of the music."
I guess it's in reference to the size of the data after compression, which is typically 1/10 of the 'original' 16-Bit/44.1KHz data size. Actually these days studio masters are more like 24-bit/96K to 32-bit/192K. So using this size logic, when we listen to a CD, we're already only listening to about 30.6% (16/24 * 44.1/96) to 11.5% (16/32 * 44.1/192) of the original. But wait, there's more. Sound waves, before digitization, are continuous. A time interval, no matter how small, contains infinitely many samples. So even if you sample it at a million Hz per second, you're still getting exactly 0% of the music. Oh bummer. (Actually I lied. We know from quantum mechanics that time has a minimum interval called the Planck interval. So sound wave is not exactly continuous. But I'll leave that calculation to the concerned reader. Hint: it'll still be very close to zero.)
Here's a little Zen puzzle. If you record a minute of total silence (say, in an acoustic chamber), and compress it, will you be listening to only 1/10 of the silence?
At this point, my head hurts so much I have no more bandwidth to analyze the Spotify issue.
@AudioMasterclass replies to @qtduck7082: I skimmed over that but the reality is that a compressed file can consume a tenth or less data than the original WAV and still sound passably good, to the point that few people would notice the difference. Regarding silence - As far as I understand it, John Cage's (who I have worked for) 4′33″ is meant to focus the listener's attention on the sounds that are normally considered to be silence. Noise doesn't work well with data compression, so the artefacts might be interesting. DM
@editingsecrets replies to @qtduck7082: This is why I only listen to minimalism on streaming services. If a measure of a Philip Glass piece is missing, you can just rerun the previous measure. This is why minimalism, Philip Glass measure. Is why mimimalism, streaming Philip Glass rerun.
@AudioMasterclass replies to @qtduck7082: @@editingsecrets Philip Glass through lossy compression - even more minimal. DM
@hellomeatrobots: While we're talking about imperfectly delivered music at the beginning of songs, how about pre-echo on LPs? I'll wager that's far more audible than whatever tricks Spotify plays to get minimal response time from their play button. I mean, if I can hear it, than it's gotta drive golden eared audiophiles nuts.
@editingsecrets replies to @hellomeatrobots: You mean you're not using the platinum core, dodo feather wrapped $3,000 cables that optimize pre-echo warmth??
@stephenrubke8960: For me, there is an important distinction between critical listening and casual listening. 95% (or more) of my music consumption is casual. And while I love sitting down in my living room and playing albums on my hifi, most of my life I’m out and about using my phone and Bluetooth headphones.
The thing is, all of the loss that occurs between the multiple compressions before the music gets to my ears and brain seems pretty negligible to me. I’ve done the lossy v lossless blind test and my 42 year old ears cannot tell the difference reliably.
Having grown up on cassettes and then CDs through decidedly lofi equipment, I’m genuinely impressed by what we can get through these lossy formats and relatively affordable headphones and speakers. I contend that this is a golden age for music lovers of all ages. Quality is so important, but I think anyone who believes Spotify’s highest quality setting sounds bad is just being an audiophile gatekeeper. Whatever degradation is there, it’s so subtle and negligible that it doesn’t detract from the enjoyment of the music. And the numbers of users/subscribers bears out that fact.
@cjay2 replies to @stephenrubke8960: So the majority of users opinions suddenly count and are correct? Interesting how your ears don't hear the 'subtle and negligible' degradations. Go. Enjoy your music.
@Gjoa1906 replies to @stephenrubke8960: @@cjay2 sarcastic dope..
@Citizen-1a: My streaming service of choice is Apple Music, and my preferences are set to play lossless when possible. When I drive to work, I go through a mobile data deadzone for about 20 minutes. Apple Music very seldom stops playback, even when I'm playing a play list of stuff I've never heard before. This tells me that the player is streaming far ahead of playback, and is able to rely on its buffer for quite a long time. Of course, prior to that deadzone I'm getting 5G data a rates good enough to stream 4K video with a multi-channel audio. We have very fast mobile data now, and audio data rates, even 24 bit lossless, are not much of a challenge for most of us. (And yes, as a consumer, I do like music streaming services, even if at the same time I recognise artists aren't getting much of a fair deal).
@SteveWille: I’d love to see Spotify make the leap to lossless as they have been promising for several years. As I have recently become to understand, the delay is not technical; Spotify employees have been beta-testing this successfully for some time now. The issue is financial; Spotify negotiated licenses from record labels for lossless/HD streaming several years ago when they thought they would be able to provide it at a financial premium to customer. In the meantime, both Apple and Amazon have introduced it at no-charge. Spotify can’t compete, providing it for free, without losing money on the licenses in place. Along with the entirely passible sound of Spotify’s highest quality setting when played through a competent digital stack, I fear we may never see Spotify “HiFi”.
@teashea1: Amazon Prime Music has the best - lossless streaming available. Quite excellent.
@G8YTZ: Interesting video. Here is my take: My sources are Spotify, CD, FM radio and Vinyl. My equipment is a Quad 66 CD/34/FM4/405 with Rogers LS7 Loudspeakers, Garrard 401/SME 3009 v2 and Linn Akurate stream receiver. I use Spotify at home and in my car and also have an extensive FLAC/24 bit music collection on a Synology RAID media server.
What I have learnt is there really is no discernible difference between the CD player and Spotify on the Linn (@320kb/s) playing the same album in an A/B comparison. This was auditioned by myself (an ex-BBC engineer) and a friend who is a current BBC sound supervisor and musician. He has a very similar system to me, but with the Rogers Studio 1 loudspeakers. For vinyl my 401/SME/in an SMD Acoustics granite plinth with A&R P77 that's been re-tipped by Expert Stylus Co makes old vinyl sound absolutely fantastic. I have some very high quality Archiv Productions classical pressings from the 1960s & 70s as well as some very good modern pressings. Vinyl is great if you want to hear old recordings in their original form in the way they were originally intended to be heard. Listing to old music on digital media is a bit like looking through modern double glazed windows installed in a Georgian house, yes the image might be better, but the imperfections in the old glass ad the glazing bars is part of the charm and the experience.
My conclusion is that I've no idea what kind of magic fairy dest Linn sprinkle on their stream receivers (I guess it is down to a very stable clock, great codec, jitter management, PSU etc.) but whatever they do but they seem to achieve the impossible by making a service like Spotify sound as good as an old, but well designed, CD player that uses the Philips chip set. At the time this chip set was always considered to be the best and was also had first over sampled DAC (4x). Older CD players, like the Quad, also support equalisation switching used on many analogue mastered CDs. Modern players seem to have forgotten about the two equalisation standards on a CD.
As you say, other so called "lossless" services (lossless means in my book not loss free, just loses less than other compression systems) seem to be more based on marketing nonsense than actually better sound quality. The best theoretical sound quality is available with streaming radio stations such as the German station Mother Earth Radio (96kHz/24bit FLAC, 4.6Mb/s) but the Linn Radio which is 24bit/320kb/s sounds just as good. I trialed both Apple Music and Tidal, neither impressed me, plus the only two mainstream streaming services that support continuation albums that I've found are Spotify and Deezer, yet the Columbia Long-Playing record supported this feature from day one of its launch in 1948 and tape supported this even earlier! At the tine of writing this is March 2023, will they ever catch up?
My conclusion is that with streaming services it's largely all down to the the quality of the stream receiver and what it does with the received stream.
@EgoShredder: Not sure I could make it through six episodes of cringe acting and dialogue like that. 🤣 I've never used any subscription service where I do not own the content, but everything is about speed and convenience these days with few other if any considerations.
@fotobrell9593: An interesting test would be to take the lost parts of the music and try if you can recognize the original song
@DanteLikesRock replies to @fotobrell9593: you more than likely would only be able to hear white noise and maybe an overly lo-fi drum beat.
@Animal_lives_matter: I don't understand what the Swedish clip is talking about - packet loss should trigger retransmission so no data should actually be lost at the user's end. If data was lost, the audio quality would be degraded very severely and very noticeably, to a degree that vast majority of users would not find acceptable. Akin to packet loss during a video conference call, where the video freezes momentarily and audio cuts out intermittently.
@AudioMasterclass replies to @Animal_lives_matter: The drama is clearly simplified and we don't know what was lost. If I had to guess, it would be that the stream started at a very low data rate then quickly switched to high, 'high' in relative terms of course. That's just a guess and we'd need a time machine, or inside information, to really find out. DM
@Citizen-1a replies to @Animal_lives_matter: @@AudioMasterclass I think Spotify (and many other streamers) used a transmission mode called UDP, that doesn't not ensure that every packet is delivered. Now, they use TCP instead, and that's detects packet loss and re-transmits at the transport level. As they are using TCP, we can be sure that packets are not lost.
@chrisbartram3034 replies to @Animal_lives_matter: i also have no idea what they were trying to get across. Are they saying that they lower the bitrate at the start of the track to fill a buffer more quickly, so the track starts playing quicker?
@pixadordelterrat2725 replies to @Animal_lives_matter: @@chrisbartram3034 This makes sense to me. Spotify uses (or at least used) the Vorbis audio codec which was not designed to deal with packet loss (as opposed to Opus) so I highly doubt they would use UDP. What Vorbis developers did have in mind was bitrate peeling so it should in theory be able to handle changes in bitrate within the same stream. Starting wth a lower bitrate would need less buffeing to start playing ASAP before the full data gets to the user.
@tylero8595: I have an excellent Gold Note streaming dac. I rarely listen to it. I much prefer vinyl. Its an experience listening to vinyl.
@paulhenner8914: Thanks for confirming what my ears have been telling me !
@calummcgregor3662: Have collected music for past 6 decades ... LPs?... fine in theory but I hated pops, cracks, scratches, warps, crappy pressings and the whole hassle factor ..... loved the gate folds and liner notes though ... collected 1,200 LPs over the years .... cassettes? .. not for me ... CDs in theory should "work" but some of the digitization has not been stellar, had several bronzed CDs and endless rounds of remastering to buy over and over... though, for convenience, and progressively better mastering etc CDs are still worth buying (I have over 3,000) ... I also rip my CDs to easily access my tunes (150,000 plus) ... having discovered the wonders of using an external DAC on my PC to listen to my music I now subscribe to Qobuz for listening / trying out music (Spotify quality was not good) ... and I really, genuinely. do enjoy streaming HiRes from Qobuz to DAC to my ears ... I am really appreciating the oppottunity / pleasure to listen, in great audio quality, to a lot of music each month for a subscription price equivalen to the cost of a single CD ... for me HiRes streaming absolutely "works" ... and I like Qobuz ... (tried Tidal but was not convinced by MQA) ... I still buy plenty of CDs ... but these are more informed choices ... often of artists or genres "new" to me .... purchases based on listening first ... Qobuz does need a better search engine though, but no doubt that will come in due course. I have also been buying a few albums in HiRes on BluRay because I enjoy the listening experience (stereo, not 5.1 or Atmos) ... p.s. I don't have remotely expensive kit to get a great sound ... and not a stylus in sight in my house anymore .. and, apart from a few signed copies, the LPs are gone .... CMcG, Aberdeen, Scotland ...p.p.s. enjoyed the video
@davidfromamerica1871: People don’t know how streaming works. Repeat:
People don’t know how streaming works. Repeat, people don’t know how streaming works.
Repeat, you don’t know how streaming works. If you did you would not be posting this video in the first place.
@garrick262: I’ve had subscriptions to Tidal, Spotify, Prime music and Deezer over the past few years. MQA on Tidal got on my nerves as it was always much loader than other tracks. Loader doesn’t equal better. Spotify has an enormous library but many are very poor quality tracks. Prime has a limited library but the quality was good. Currently using Deezer and it’s a mid point compromise, good library and quality of sound. I have a small record collection (about 160) and a small CD collection (about 100) but rarely play them due to ease of streaming music. I have done some direct comparisons with my CD to the streaming version and they are different but it’s had to pin down which is better. A bit like the analogue verses digital thing, they are different but not sure which I would consider to be better. I would argue streaming is the new ‘radio’ and I/we have to accept we will never actually own our playlists. BTW enjoying your musings 😊
@Konspiration100 replies to @garrick262: Fully agree on that with a hint to try Qobuz maybe for free 3 month long-. Oitried even more and found Qobuz to have the best sound quality (library of HiRes recording in Jazz and Classic at least very huge) and it does not really stand behind when it comes to the overall music library compared to the biggest streaming services.
@Chunksville replies to @garrick262: MQA would have been a good solution 10 years ago but with the now available band width its not needed plus its awash with marketing voodoo. It's always difficult to compare streaming with physical discs as there is no guarantee the master source is the same
@teekay_1 replies to @garrick262: Qobuz is pretty good but without the selection of Spotify, but good enough to be acceptable for $11/month.
And I think you're overstating the problems with Spotify. Spotify is perfect for listening on you PC, in your car, and often in your home.
@Cashmore1671: Yesterday, speaking of streaming, I put an Analogue productions 1/2 speed “master” on my Rega P/10, Jethro Tull’s 2nd album Stand Up. It was good but didn’t blow me away, (which for $60.00) it should have! So I started thinking ah-ha why don’t I stream the same album through my HIFI Rose 250 DAC and sync them both track by track through my Prima Luna Evo300 pre and monoblock Evo 300’s, via Qobuz. Groove to groove as it were and switching between aux1 and aux2, turntable and dac respectively, I was quite astounded at the results; the streamed version was right on one to one as far as timing, furthermore the streamed version though brighter was extremely impressive making me question the 60 bucks I put out for the LP. For reasons only emotion can explain I still preferred the LP, mostly for it’s warmth.
@Konspiration100 replies to @Cashmore1671: How I like your comment after all this enthusiasts who feel sorry for me ( I have sometimes the impression) not to own a 5 or 10000 bucks player with your 60 bucks record and wanna tell me how much better this is than ANY digital media. Sorry guys first when it comes to scientifacally compareable numbers digital music IS superior. One might prefer the analogue sound no problem with that and it may have advantages from case to case but its NOT the only way to listen to high quality music. and often not even the best, beside the fact that mixing and mastering is of even greater importance.
@chordsoforion: Interesting. After watching this video, I decided I would do a test such as you suggest of Apple Music. I selected a song from my 2021 Album "Wood" and played it through Apple Music at full CD quality, recording the audio at 16-bit/44.1 using Audio Hijack. I then compared it to my original master that was submitted to Apple (also 16-bit/44.1). I then trimmed both files to the same starting and ending samples using iZotope RX 10. Here's what I found out:
1) All LUFS levels (momentary, short-term and integrated) are the same. LU is the same.
2) True peak levels are different. The original is -1.49dB and the Apple Music version is -1.30dB
3) Obviously, the 2 files are not going to null due to the difference in amplitude. But - there is another difference. Some kind of weird stretching of the samples. Both files are the same exact length mathematically and start on the same sample. In RX, they show the same numbers of samples. And they do almost null at the start, but by the end of the 4 minute song, there is a 22 sample offset.
I am not an engineer, so I can imagine what is going on, but at least recording the output of the Apple Music player on my Mac produces slight differences. Not sure if there is a way to capture the audio stream direct without going through the Audio Hijack - this could definitely account for the differences.
Now, can I hear the difference? Not me, with my old man ears.
@jamess7589: Streaming fills the gap for listening to new artists and albums. You like it, buy the vinyl or CD and keep it forever
@EgoShredder replies to @jamess7589: Finding new music has rarely been an issue for me, as I check out the names of the musicians on any given release and then see what other bands they have been in etc. You can fork off in all directions from just one, and end up with hundreds of albums to check out. Then there are recommendations from friends occasionally.
@sonic2000gr: Hopefully, at the current Internet speeds the lossy algorithm used by Spotify will probably not need to lose data as the transfer speeds are usually more than sufficient. But there are other problems with Spotify: You get a lot of remasters and remixes instead of original versions. Many recordings are really low quality and suffer from loudness war like CDs used to. And tracks/artists keep disappearing from playlists as Spotify loses the rights. In the end of the day Spotify is mostly convenience than anything else. It is hard to imagine not using Spotify but it definitely gets infuriating at times.
@G8YTZ replies to @sonic2000gr: I fully agree with your comments, remasters are what I hate about Spotify, I with you could specify a filter for this, plus as you say tracks keep disappearing from your playlists. I use it in the car and for playlists. I buy the music I really like and stuff it on my Plex server.
@cjay2 replies to @sonic2000gr: "CDs used to" ? Really? The Giles Martin 'remix' of the Beatles Revolver 50th Anniversary set is compressed and louded. 2022. Just last year. And so is the 2020 Abbey Road box set. It's easy for ME to imagine not using spotify.
@igorbeuk4068 replies to @sonic2000gr: I will explain in shortest way. It's not about your internet speed but in Unison solution in which everything equally fast like 5G but it advanced even more. Current Internet is using peers and every device or repetitor is used for sharing speed. Better sound quality is Music with more dynamic and lower levels and it's a fact unfortunately
@maidsandmuses: Streaming? No thanks. Not an issue with sound quality; simply too many modes of potential failure long-term; from streaming service providers dropping some artists at best, to them going bust or a complete breakdown of internet connectivity at worst. There's the environmental footprint as well; the vast majority of songs I will ever listen to in the future I already own on CD or as downloads, so under that pre-condition the environmental footprint of playing the music I already own on CD or DAP is less than streaming it (over and over) from a data centre through my ISP's infrastructure.
@teekay_1 replies to @maidsandmuses: "There's the environmental footprint as well; "
I don't understand this statement. Can you explain it more clearly?
@maidsandmuses replies to @maidsandmuses: @@teekay_1 Sure. It is about comparing the carbon footprint of streamed music vs. CD in general.
The production of CDs involves oil exploration, distillation & refining for the plastic used, building a manufacturing site, energy for the manufacturing process, packaging, transportation, and distribution.
Setting up a streaming service for music involves the building of one or more data centres (implicitly, they may already exist), manufacturing of data servers, the maintenance and upkeep of the communication infrastructure (internet, ISP etc.)
The play-back of a CD involves the power use of a CD player. The playback of streamed music involves the power use of a computer (or streamer), plus the power use of quite a few data servers in the data centre as well as in the internet infrastructure.
The carbon footprint of these two media are therefore not easy to compare, and would have to be carefully calculated on a per-minute/hour of listening time.
There has been research into the carbon footprint of music streaming vs buying a new CDs now. As expected, the more often you intend to play the same music, the more the carbon-footprint balances in favour of owning & playing the CD. But it is not an easy comparison to make as you would need many more CDs to have access to the same range of different songs you have when streaming. Some (therefore debatable) research suggested that when you play an album for more than 5 hours, buying and playing the CD works out more environmentally friendly than streaming the album (the carbon-footprint of the manufacture & upkeep of the data server & internet infrastructure is rather large, and carbon-offsetting is an accounting fudge that mostly shifts responsibility to other, usually poor, countries).
However, my point was that in my case the music I own I bought on CD already, years or sometimes decades ago. The carbon footprint for the production of those is therefore a past event that cannot be changed. It should be pretty clear that therefore, on a per hour listening basis, going forward playing the CDs I have already owned for years is going to have a lower carbon footprint then streaming the same music. If I were to keep my old CDs (which I will 😁 ) yet constantly stream that same music rather than playing my CDs, I would be more than doubling the carbon footprint of my music listening habit.
@DaskaiserreichNet78: 20 years ago we listened to CD quality music on Dedicated stereo systems.
Today most people listen he heavily lossy compressed music of services such as Spotify. What's more is that this music is played on a lofy cellphone speaker.
The music listening experience has gone downhill big time.
@spandel100 replies to @DaskaiserreichNet78: 😫
@editingsecrets replies to @DaskaiserreichNet78: I'm totally flabbergasted at the miserably fuzzy, bareless recognizable sound quality so many people seem to enjoy hearing from their little phone speakers turned way up.
@wyzkun replies to @DaskaiserreichNet78: @@editingsecrets tbf, they also dont have the gear to listen to lossless either. so i dont think they lose much. some times it just because they never a have a good pair of earphone or speaker. there is just no going back after that
@hellomeatrobots: If the lossless music provided by streaming services wasn't, I'm reasonably certain they'd have been called out on it by now. Yes I love Apple Music. Instant access to tens of millions of songs in at least CD quality for the price of a single CD a month? Yes please!