back to article Spotify to introduce lossless audio streaming: Better sound or inefficient gimmick?

Spotify will introduce CD-quality lossless audio streaming, in a new service for premium subscribers to be rolled out in selected markets later this year. The streaming biz was early to market with its service, launching in 2008 when music downloads were more of a thing, but is facing competition from Apple Music and Amazon …

  1. Throatwarbler Mangrove Silver badge
    Go

    I think this is more important

    "Talking of dynamic range, the music industry is also notorious for deliberately reducing it in order to increase the perceived loudness of recorded music. The idea is that louder music is more likely to be noticed and therefore streamed or purchased, never mind that it reduces the quality for discerning listeners."

    I can certainly hear the difference in quality between music which was mastered with more dynamic range versus the squashed-together music common today. I imagine that "HD" music combined with an enhanced dynamic range could yield more satisfying audio quality, even if the average user would be hard-pressed to identify the difference. I am happy to see that there are people raising the bar for audio quality--maybe it will drive a revolution in dynamic range as well.

    1. J27

      Re: I think this is more important

      Good luck, if history is anything to go by we'll just end up with super-loud dynamically compressed music so that people perceive it as "high-res".

      1. Gloddata

        Re: I think this is more important

        Bingo. If they can screw up CD audio then they can screw up HD.

        If the master is very good and I am listening at home in a quiet room with my best headphones and best DAC then I can _BARELY_ tell the difference between 320kbs and CD-quality Flac. But it's only in certain tracks in certain areas were it's very quiet, etc. Any sort of air conditioning blowing or background noise going on and it's going to be impossible to tell the difference.

        Lossless audio streaming is a 'why not' scenario. If you are on metered wifi networks then it's dumb, but if you are at home, then why not? It's a nice to have.

        However the HD formats are pure 100% audiophile hucksterism. It's for the same market of people that buy Monster HDMI cables for 50 dollars, lust over nitrogen gas infused copper audio cables, and wear copper wristband bracelets with magnets to help with the RSI.

        Having "high resoluiton" audio is nice for studio work you need to do a lot of manipulation of multiple tracks and do different audio mixes to get the best results. The real purpose to them is to makes a audio engineer's job easier.

        But for a end user format it's a scam.

    2. This post has been deleted by its author

    3. slimshady76

      Re: I think this is more important

      It's not just about the compression. Over its life, any music piece will be submitted to one or more "remasters", which ultimately means it will be loudified and compressed beyond recognition. Back at the start of the 90s my brother gave me Pink Floyd's Dark Side of the Moon on CD for my birthday. It was an original, UK edition. A couple of years after that (and many, many joyful listening sessions) I handed it to a soon-to-be ex girlfriend, and ultimately forgot about it. So one day, after the millennium change, I craved for more DSotM once again, and went berserk trying to find another copy. I could only find a "remastered" version on CD, and with heavy cursing, I bought it. Shortly after I ran into my former girlfriend, and asked her for the CD. She gave it back to me, and I still use those two editions to show my fellow melomaniacs how a "remaster" by some half-assed, anonymous sound engineer could destroy the work of those involved in the creation of a masterpiece. This still holds true when talking about streaming. Original masters are chopped and chewed by who-knows-who and then spat into our ears.

      TL/DR: As the singer from Gogol Bordello put it a few years ago, "We basically have a couple of generations who grew up listening to music on shit speakers".

      1. Anonymous Coward
        Anonymous Coward

        Re: I think this is more important

        Not a music fan at all - think the last CD I bought was David Bowie's Greatest Hits, and that was before he died.... Anyway, daft question alert - why do they remaster stuff in the first place?

        1. Martin an gof Silver badge

          Re: I think this is more important

          Possibly because you can re-issue a song or an album, and fans who bought the original will probably buy the new version whether you have made a total hash of it or not. Possibly because it allows you to re-issue a song in a more "streaming-friendly" or "radio-friendly" format (long songs are cut down while short songs might have sections repeated), which might get you some more sales.

          Often though the motivation is genuine. You have to remember that the "master" is not the original studio recordings; it is the version of the song which is copied to make the record, cassette or CD. If the original recordings were made in an analogue studio and edited on analogue equipment it is highly likely that all subsequent releases have been made from copies of that one final "master tape". This is the edited mix and it is used again and again because it's the definitive version.

          What you hear on your 7" single isn't what was put down on tape by the musicians, it isn't what was heard in the recording studio's control room, it isn't what was heard in the edit suite once they'd run it through whatever post-processing they used and it isn't even what the editor heard through his speakers when he played back the master tape. After that last editing effort, the "master tape" was copied to create a library master and maybe a backup master, and those were copied to make reproduction masters which were sent to the factories which actually produced the vinyl, cassette or CD. In the case of vinyl the reproduction master is used to cut a master stamping disc, which itself will wear slightly each time it is used to stamp a wiggly spiral into a lump of warm plastic.

          Each copy suffers "generational loss" and in some cases there could easily be half a dozen generations or more between the multitrack machine in the studio and the 0.125" of thin rust-coated plastic you slot into your Walkman.

          Since the 1980s the copy masters are probably digital so you don't have to worry about them becoming worn out through over-use, but technology has moved on, and the generational losses that were an inevitable result of analogue processing, mixing and editing are not a problem for digital.

          So in its simplest form, a "remastering" will go back to the original studio recordings and digitise them directly. A new mix can be created from these tapes and it could easily be exactly the same as the original mix, except that there is now no generational loss so the copies will be much "cleaner" and contain more detail. The classic example is whichever Beetles release it was where Lennon can now be heard swearing quietly just after the song finishes, something which was lost in the noise of previous releases.

          On a more subtle level, getting rid of analogue tape noise actually makes the job of a lossy encoder much easier, so even when encoded at a low bit rate you will get a much better result than you would have from encoding the original hissy master.

          Of course there are problems. From an artistic point of view the original studio engineer may not be around to re-create his mix, or the artist may wish to have more of an input and change the mix, or the record company may want to apply their new whizz-bang toy multiprocessor to the thing to give it more "punch" or "presence" or "body". These sorts of things are a problem mainly for fans.

          Sometimes it is difficult to find a working machine to play a particular tape. This is less an issue for audio formats than video, but it can still happen. Sometimes the studio machine was slightly mis-aligned or poorly set-up, and the tape will only play back correctly on that one specific machine, or another one modified to have the same flaws.

          A more difficult issue is that the studio tapes may have deteriorated through old-age, poor storage or if they are are of a certain vintage, from poor materials used in manufacture.

          I think it was Ampex tape (but quite happy to be corrected if not) from the early 1970s that suffers very badly from poor glue binding the rust onto the plastic tape. Often when the studio tapes are brought out of storage they are found to be glued together in one lump, or when they are unwound the rust falls off the tape. Incredibly, many of these tapes can be rescued by "baking" them at a low temperature for several hours, but then they are played once only, digitised in as high a resolution as possible and then retired.

          Lastly, and this is possibly slightly controversial, those digitised studio tapes can be digitally "cleaned" before remastering. If the tapes are poor quality this may be imperative, but for good quality originals some people would miss some of the analogue "foibles".

          So there are lots of reasons for "remastering" and not all of them are bad reasons, in fact many remasters are done with the best of intentions and in some cases a fan with an older copy of a song may not "like" the new version, because it doesn't sound like the copy they have been listening to for thirty years, even though the remastered version may actually be much closer to the sound the artist, producer and engineer intended at the time the song was recorded.

          M.

          1. Anonymous Coward
            Anonymous Coward

            Re: I think this is more important

            Wow! Learnt something new!

            Thank you Martin for a detailed answer to my question!

    4. Korev Silver badge
      Mushroom

      Re: I think this is more important

      This is one of the worst examples out there.

      Linking to the above video in this thread is probably an act of war... -->

      1. ICL1900-G3

        Re: I think this is more important

        It will never replace music.

      2. slimshady76
        Mushroom

        Re: I think this is more important

        You want the nuclear option eh? I see your Imagine Dragons and I raise you a Despacito:

        https://youtu.be/kJQP7kiw5Fk

        1. Korev Silver badge
          Mushroom

          Re: I think this is more important

          My God!

    5. jason 7

      Re: I think this is more important

      The 'Loudness War' was mostly the reason why I stopped buying CDs soon after the turn of the century. I kept locked into releases recorded or issued on CD before 2000.

      I also blame the easy availability of PC based audio recording for the drop in all round quality of pop & rock. Before, you had to go into a studio and hire a producer/engineer and that presented a fabulous barrier to entry/filter to those with more ambition than talent.

  2. Anonymous Coward
    Anonymous Coward

    AB Testing

    TLDR; Too old, can't tell the difference, saved loads of cash to spend on other things.

    About five years ago I thought about upgrading my music collection to hi-res audio so did a bit of research including AB testing. I couldn't tell the difference between a hi-res and a 'standard' audio source. This was with a hi-res audio player and Sennheiser studio monitor head phones etc. Saved a fortune by sticking with MP3s. Tried again recently with Amazon HD and normal Amazon Music. Again no difference. It didn't lead to better quality or more satisfying listening. Even ripping my own CDs to FLAC does not really show any difference. I admit that I am in my fifties, and attend far too many rock concerts. I am not an audiophile and can't tell the difference between speaker cable at 100 GBP per metre or 5 GBP per metre which may be why I can't tell the difference between hi-res and normal quality.

    1. sreynolds

      Re: AB Testing

      Philistine. You know all those valves on certain high end amps are there not for their quadratic gain not for any other acoustic properties. It's all about snobbery. How dare you criticize these high end brands.

      What make me laugh too is these 200 squid + microphones that don't use anything better than a 2 buck MEMS device to do the ADC.

    2. Anonymous Coward
      Anonymous Coward

      Re: can't tell the difference between speaker cable at 100 GBP per metre or 5 GBP per metre

      Then your ears are fucked.

      Changing nasty cheap speaker cable to fat, 100 core copper cable was the only time I've ever heard a clear, noticable difference when upgrading my hifi equipment.

      Don't listen to this AC, bad quality speaker cables can cripple an otherwise good quality audio setup!

      (Everything else they say I agree with!)

      1. desht

        Re: can't tell the difference between speaker cable at 100 GBP per metre or 5 GBP per metre

        Sure, cable quality makes a difference. But absolutely nobody needs to spend £100 per metre of cable - you can buy perfectly decent cables for a fraction of that. Anyone who thinks £100/m cables is a good idea probably has shares in Monster.

        1. Anonymous Coward
          Anonymous Coward

          Re: nobody needs to spend £100 per metre of cable

          I said using 100 core copper cables beat nasty cheap ones. I didn't pay a quid per core!

          I agree noone needs to spend £100 a metre, but you should get good speaker cabling.

          1. Return To Sender

            Re: nobody needs to spend £100 per metre of cable

            13A UK mains flex (not solid core) does the trick. Saw it used (carefully hidden from the eyes of the great unwashed) at an audio event for some pretty serious (read expensive) kit. Had a bit of a chuckle with the guys behind the stand about it, their professional take was that cable thickness/no. of cores was the important bit unless you run unfeasibly long lengths...

            1. sreynolds

              Re: nobody needs to spend £100 per metre of cable

              The only problem with the house wiring stuff is that some people might think that you are playing mains on the speakers.

              I don't know why but someone put two phase wire in a stove, even though one phase is fairly standard. The second active was left floating at the switchboard but wired on the stove (what, leave a wire hanging?) When the crappy stove broke down under warranty the techie was just so scared he wanted to run away.

              Probably two core multistrant automotive cable might work well too.

            2. genghis_uk

              Re: nobody needs to spend £100 per metre of cable

              I seem to remember that What HiFi have used 13A mains lead in a speaker cable test - late 90's or thereabouts. It came out much higher than you would expect too. Something like 6th out of 10 cables meaning 4 were worse than mains lead!

              A quick google does not show it and I can't be bothered to do a long search. I will say that I was a cable sceptic - it's LF, why would cable make a major difference? But I replaced the rubbish wire on an old hifi with something half decent and was gobsmacked at the difference... I still can't explain it scientifically but my ears could definitely tell

              1. Anonymous Coward
                Anonymous Coward

                Re: nobody needs to spend £100 per metre of cable

                Back (way back) in my university days we had a subject on this.

                There was a lab where we compared every kind of speaker cable available.

                Using real spectrum analysers, high quality signal generators, correctly calibrated environments, etc, etc, we had to *measure* the distortion, attenuation, frequency response, etc, etc of each setup.

                (The purpose of the lab was to train us in how to use all of the equipment, which was to be used in later labs.)

                The best cable by far was a really fat figure eight power cord with plenty of strands in the cable.

                It was provided by one of our students who worked at the ABC (Australian BBC-a-like) TV studios. It was what he used at his work. Cable thickness (and multi strand count) was the best indicator of quality. Single thread, thin, speaker wire 'ribbon' was the worst, and was included in the box with most speakers.

                Monster cable was good, not the best, but good. It was also the most expensive by far. The power cable was the cheapest by far.

                Of course, the result was that we all chipped in and bought a roll of several hundred meters of the power cable, and took our share home. Decades later, my (new) speakers are still wired up with that old cable.

              2. Anonymous Coward
                Anonymous Coward

                Re: nobody needs to spend £100 per metre of cable

                I seem to remember that What HiFi have used 13A mains lead in a speaker cable test

                Relying on the results of tests in What HiFi is unwise.

                But I replaced the rubbish wire on an old hifi with something half decent and was gobsmacked at the difference... I still can't explain it scientifically but my ears could definitely tell

                Ever heard of the placebo effect? Your brain is telling your ears the new cables are better because it knows the cables were replaced and you spent money on the new ones. Try doing a proper double-blind test and measuring the results. Which is something that I doubt What HiFi has ever done in its history.

      2. Anonymous Coward
        Anonymous Coward

        Re: can't tell the difference between speaker cable at 100 GBP per metre or 5 GBP per metre

        Original A/C here who mentioned cable price.

        You might be right about my ears but I was really making the comment that there are diminishing returns on HI-Fi costs. A 5 GBP per metre budget should get you a decent 100 core cable. I was talking about the difference between good cable and over priced cable. So I am in total agreement that you need good cables, but cables become good at a very low price. Have an up vote.

      3. IGotOut Silver badge

        Re: can't tell the difference between speaker cable at 100 GBP per metre or 5 GBP per metre

        Search for Audiophile coat hanger test.

      4. jason 7

        Re: can't tell the difference between speaker cable at 100 GBP per metre or 5 GBP per metre

        The only thing that really affects speaker cable is the basic physics of 'shorter and thicker' IMO. That covers 99% of any real changes.

        I've often said that that £100 a meter cable was probablty originally specced to be used in aircraft or a CAT scanner and cost 0.05p a meter. The Hi-Fi companies do not have their own undustrial smelting works to make cable. It's massive industrial scale work that only is viable in runs of many kilometers a go. That £1000 a meter cable Russ wants to sell you, did NOT cost 99% of that to make. They just buy in a huge consignment from China or Eastern Europe where they still make the stuff and relabel it. The Mk2 version is just the old stuff running out and having to order in something else.

        People still want to believe Hi-Fi companies have Cable factories and in-house metallurgists...

        Just get some VanDamme or similar for £3-£4 a meter and be done with it.

    3. Zippy´s Sausage Factory
      Devil

      Re: AB Testing

      The only thing it'll really do is to clog the Internet with streams that really could be very much more efficiently done as 256K MP3s. As if Netflix, Disney- and all the other streaming services weren't enough to choke the net anyway.

      Oh and of course mobile phone companies will love it. Spotify now uses 4 times as much data to stream the same song? Jack up the mobile data prices, Ernie, we're gonna be (even more) rich...

    4. Steve Graham

      Re: AB Testing

      I'm not aware of any convincing research that shows that people can tell the difference between well-encoded, high-bitrate MP3 and CDs.

      However, I've been in the studio with good producers and engineers and they definitely can hear stuff that I can't. Actually, given my age and the abuse my ears have suffered, it's surprising that they work at all.

      Up until recently, I bought physical CDs, but my last album purchases have been MP3 downloads, bought directly from the artistes or labels. And I'm happy.

    5. katrinab Silver badge
      Flame

      Re: AB Testing

      "I am not an audiophile and can't tell the difference between speaker cable at 100 GBP per metre or 5 GBP per metre"

      No, the reason you can't tell the difference between a £100/m cable and a £5/m cable is because there isn't any difference.

      1. Anonymous Coward
        Anonymous Coward

        Re: AB Testing

        "No, the reason you can't tell the difference between a £100/m cable and a £5/m cable is because there isn't any difference."

        There might not be any difference in audio quality but there is a difference: £95/m.

  3. David 45

    Compressed to within an inch of its life

    Little point in offering high bit-rate audio on the internet as, these days, there is an incredible amount of compression, processing and butchering applied at source. Can't recreate what's been taken away. Most CD's and radio (both FM and DAB) sound absolutely foul, as there is very little in the way of proper dynamics left after being pumped through such monstrosities such as "Optimod" and the like. It's only too obvious when fed into an audio editor like Audacity. It almost flatlines! Not a scrap of life left in the music, any semblance of a decent dynamic range having been completely wrung out to dry. Fit only for background music! Every production is now trying to sound louder than anybody else's, with the result that EVERYTHING now sounds awful. Decent sound quality now seems to be a thing of the past.

    1. Sgt_Oddball

      Re: Compressed to within an inch of its life

      There are still some out there that really value it but usually you're talking small record labels who also spend time making vinyl (Alien Transistor for example).

      Album on flac and record played through the same sound card and save for the nuances from my current cartridge you'd be hard pressed to tell between them. Spotify stream though was enough to make not bother and wait for the album to physically be shipped (brexit be damned).

      Also helps that I recently got some studio monitor headphones, my previous set trended to mask issues with recordings by being less analytical and muddying the sound more.

    2. sabroni Silver badge

      Re: EVERYTHING now sounds awful.

      Yeah, that's right grandad. You go and have a lie down and I'll ask the kids to be quiet.

      1. genghis_uk
        Joke

        Re: EVERYTHING now sounds awful.

        The problem is that kids don't know when they are wrong...

        Too much ego and too little experience - put the phone down and learn something!

        <wasn't like this in my day.. mumble, mumble, pass the tea dear>

  4. Pascal Monett Silver badge
    Flame

    "it reduces the quality for discerning listeners"

    The music industry doesn't give a damn about discerning listeners. The music industry is all about whatever the young people care about, because they are the ones glued to their phones and most susceptible to spend money on streaming services.

    The discerning listener can go buy vinyl.

    1. AW-S

      Re: "it reduces the quality for discerning listeners"

      "The discerning listener can go buy vinyl"

      Absolutely agree - and on the day my limited edition copy of Mike Lindup's Changes album arrived - made of top quality orange vinyl.

      1. sreynolds

        Re: "it reduces the quality for discerning listeners"

        I quite like vinyl because you past music playing has left an imprint on the needle thereby slightly affecting all subsequent music that is played thereon in a subtle and almost indubitable way, but one that only the owner of the device can truly appreciate as he bores you to near death telling you about it.

        1. Anonymous Coward
          Anonymous Coward

          Re: "it reduces the quality for discerning listeners"

          Never mind vinyl, I remember many many years ago going on a school trip to a collection of pianolas nad other mechanical instruments ( I was astonished a few years ago while visiting Kew to find there was a new "Museum of Mechanical Music" which appears to be based on the same collection) where the man who ran it went on at length that people spending £100s of amps, turntables, speakkers etc (this was pre-cd!) were wasting their money as the only true way of experiencing musical recordings was on something like a pianola where the exact timing and pressure from the fingers of a famous pianist had been recorded and were then replayed on real piano to "repeat a performance in a way these electrical boxes can never do"

          1. Anonymous Coward
            Anonymous Coward

            Re: "it reduces the quality for discerning listeners"

            Stravinsky, famously, produced a version of his Rite of Spring that he cut directly onto pianola roll to deliberately eliminate any effects introduced by the pianist.

          2. Someone Else Silver badge
            Happy

            Re: "it reduces the quality for discerning listeners"

            Can't wait for the bass guitar pianola!

        2. David Nash Silver badge

          Re: "it reduces the quality for discerning listeners"

          Forgot the joke icon there...I wonder if the downvotes are from people thinking "he's talking BS" or from people who think it's true and object to your last point!

    2. hoola Silver badge

      Re: "it reduces the quality for discerning listeners"

      And most people now listen on their phones, either with Bluetooth connected headphone (earbuds or whatever) or a Bluetooth connected speaker.

      All are total garbage due to bandwidth. Even if you have something that supports LDAC on Bluetooth it barely makes any difference.

      But one cannot get away from all the shite processing that goes on before something is streamed or broadcast. DAB is a total waste of time due to the bit rates.

      I know my ears are buggered but the sound from something like Classic FM compared to the same CD or even an MP3 off the same CD is night and day. It has been completely mangled to make up for the fact that most people listen on headphones/earbuds that cut everything off below about 1KHz.

    3. Someone Else Silver badge

      Re: "it reduces the quality for discerning listeners"

      The discerning listener can go buy vinyl.

      Have you not been paying attention? The best vinyl can put out something like 63 dB dynamic range. Even the worst CD can put out 90dB dynamic range. (Whether they are allowed to is another story; cf. discussions on uber-compression on this forum). So with vinyl, you start out with a high level of effective compression, and go downhill from there.

      And I guess "discerning listeners" like pops, crackles, and wow?

      1. jason 7

        Re: "it reduces the quality for discerning listeners"

        I abandoned it as soon as I felt CD players reached a decent level of maturity and never looked back.

        Whether its a Tandy or a fully setup Thorens, all I hear by the end of track two is massive distortion. I just cannot unhear it with vinyl. Like nails down a blackboard for me.

  5. rnturn

    FLAC

    I ripped all my CDs to FLAC and can hear the difference between those that I originally ripped to 320Kbps MP3s. It's not dramatic but content that has a lot of high-frequencies -- cymbals, plucked strings, some electronic music, etc. -- benefits. It surely sounds better than most streaming content that I've heard. I can't imagine Spotify streaming FLAC-encoded music, though.

    1. rcxb Silver badge

      Re: FLAC

      MP3 is not the best lossy audio format available, for many reasons I won't get into here, and will never sound perfect.

      Try your test again with Musepack at standard (~160kbps) or MP2 (Musicam) at 192kbps and you won't have the distortion of cymbals and the like. Both are temporal domain codecs. MP2 has a bad reputation from DAB, but that's only because broadcasters insist on using it at terribly low bitrates.

  6. Version 1.0 Silver badge
    Boffin

    Lossless is only one third of the issue

    I've been working with audio all my life and my experience has been that while lossless is sold as important, that attitude ignores the conversion of pure, accurate electrical signals into sound pressure. And even if you get both of those perfect, you still have to consider the environment that you are in when the perfect sound is playing.

    Remember the old days of TV when everyone was concerned about "ghost" images? Well that exists in sound too and affects the quality - a perfect sound applied to a room with a hard ceiling or a mirror means that it is not the original sound. Yes, it might have been lossless but that mirror or ceiling has changed it - and this is assuming that you have a perfect set of speakers ... LOL, most people are listing on earbuds or their phone.

    1. Anonymous Coward
      Anonymous Coward

      Re: Lossless is only one third of the issue

      Well, the lossless audio is probably a modern track which has been mixed and balanced on the basis that everyone will be listening on earbuds (probably via a heavily lossy format) and have been engineered to work around the limitations of that so getting a "perfect reproduction" of this may not actually be desirable in any case!

  7. Androgynous Cow Herd

    Analog kid in a digital world...

    Is anything digital truly lossless or merely hi-resolution?

    For some things, you can’t beat a proper analog signal chain.

    Just hard to send it over the internets...

    1. phogan99

      Re: Analog kid in a digital world...

      Nyquist Shannon Sampling theorem, for a band limited signal (which audio is), a sample rate of twice the highest frequency to be sample is enough to reproduce the signal with perfect fidelity. The limiting factor is the equipment it's played on.

      1. Kobus Botes
        Boffin

        Re: Analog kid in a digital world...

        @phogan99

        "The limiting factor is the equipment it's played on."

        In my second year of electronic engineering studies our electronics professor had much to say about that. At the time (late seventies) wow and flutter (or lack thereof) was all the rage and manufacturers of tape recorders/players and turntables usually very prominently quoted how low its products' w&f were. I cannot remember exact figures, but I seem to remember premium products would be in the region of 1-2 %.

        The prof's rant was about how we wasted money doing that, as the best and most expensive speakers of the time (Bang & Olufsen comes to mind) could not better 5%. And if I am not mistaken, the human ear cannot do much better in any case (any sound engineers who can weigh in on that?).

        1. Neil Barnes Silver badge

          Re: Analog kid in a digital world...

          Er, loudspeakers are not susceptible to wow and flutter (other than that they will show you you have it); perhaps you have misremembered your professor?

          I would argue with him that money spent on improving turntables was wasted (if it were something about which you cared) - the whole turntable thing is such a fragile reproduction system that any improvement over the basic 'goes round at vaguely the right speed' is probably worthwhile - though inventing the CD was probably a good thing.

          Wow is low frequency changes in turntable speed; the best way to remove it is perhaps lots of mass in the turntable. Flutter is high frequency changes; that can be handled by good bearings, belt drives to isolate the motor, and/or good speed feedback control to the motor (but servos are hard).

          When you think about it, reproducing sound by dragging a needle across a grooved scratched in plastic doesn't seem the most elegant way of doing it...

          1. R Soul Silver badge

            Re: Analog kid in a digital world...

            "perhaps you have misremembered your professor?"

            Most definitely if he remembers B&O kit was the best.

          2. Someone Else Silver badge

            Re: Analog kid in a digital world...

            Wow is low frequency changes in turntable speed; the best way to remove it is perhaps lots of mass in the turntable.

            Correct, but wow is also caused by off-center cut spindle holes, and by warped vinyl...both of which have infected my vinyl collection in the late 70s and early eighties.

        2. R Soul Silver badge

          Re: Analog kid in a digital world...

          Bang and Olufsen kit was never the best. It was usually the most expensive though.

          Their $$$ CD players had exactly the same electronics and firmware as Philips players that cost a tenth of the price or less. Apart from the price, the only difference between them was the prettiness of the box that housed said electonics. Oh and you couldn't buy Bang and Olufsen kit in Dixons.

      2. cornetman Silver badge

        Re: Analog kid in a digital world...

        > ...is enough to reproduce the signal with perfect fidelity.

        Is that really true? I always understood that twice the frequency was the *minimum* required to reproduce a signal. I don't know how double frequency sampling would adequately capture the differentiation between a sine wave, a square wave or a sawtooth.

        1. Neil Barnes Silver badge

          Re: Analog kid in a digital world...

          Your ears can hear a sine wave at 'up to' 20kHz or so, if you're young enough and haven't spent too much time at music gigs or driving heavy machinery.

          Which means that from 10kHz, a sine wave is all you can hear since the harmonics of any other wave form are outside your hearing range. In all probability you can't tell the difference between a sine, square, or sawtooth at anything greater than 5kHz; you simply can't hear enough harmonics.

          Allowing for practical filters, a signal sampled at 2.2 times its highest component can be adequately reproduced. Square, sawtooth, and triangle signals have significant energy in higher harmonics - and since those frequencies fall outside the range which can be reconstructed, it is important not to allow them to be sampled at all. You filter the input to ensure that any frequency over 1/2.2 of the sample frequency is not passed through, as it would simply alias to a lower frequency on reconstruction and would be impossible to remove.

          So, for high frequency signals you can't tell the difference between different waveforms because the harmonics aren't there, but you couldn't hear the difference anyway.

          1. Someone Else Silver badge

            Re: Analog kid in a digital world...

            You know, this discussion of Nyquist frequencies is fascinating and enlightening. but it seems to be ignoring one point: The beat frequencies between inaudible harmonics is (or can be) audible. The beat frequency between 30kHz and 31kHz is 1kHz, and that is audible even by a long-past-his-prime rocker like me. Yes, it is true that the amplitude of that beat frequency might be very low; perhaps below the amplitude necessary to hear it in and of itself. But that beat frequency will subtly affect the signal of other frequencies around 1kHz, and all those interactions is what makes symphonic music (and even electronic music) have all that "nuance", "life" and "warmth" that the audiophiles rave on about.

            1. Neil Barnes Silver badge

              Re: Analog kid in a digital world...

              Those beat frequencies are real. They're there on the input signal. If they're below the input sampling filter frequency cut off, they get recorded. Nothing mystical about them.

            2. This post has been deleted by its author

    2. Neil Barnes Silver badge

      Re: Analog kid in a digital world...

      In a previous life - thirty-odd years as a broadcast engineer - I demonstrated that a room full of audiophiles couldn't tell the difference between their fancy cables and a bit of mains cable with a one ohm resistor in parallel with an 1N4001 diode in one leg...

      All audio reproduction is a question of 'which distortion does sir prefer the sound of?' My broadcasting colleagues used to despair that after all our efforts to get the best possible quality at the bottom of the transmitter mast, people would be listening to it on three-transistor AM radios with a 3kHz bandwidth and a two-inch loudspeaker... nonetheless, the rule is always that once you take something out of a signal, you can't put it back.

      Lossy compression systems work on the well-tested theory that there are bits of the signal which *most* people can't hear if other bits are going on, either louder or close in frequency. These masked signals can be ignored, or be sent with fewer bits (i.e. at lower resolution). It's not surprising that some people can tell the difference and equally unsurprising that many people can't (given enough bits). Given that most people aren't likely to be listening in studio conditions, if they're happy with what they hear, fine.

      If they're not happy, that's when they will want to spend some money and look at alternative approaches. The job of the audiophool industry is to keep them unhappy...

      1. Paul Kinsler

        Re: a room full of audiophiles couldn't tell the difference

        My impression is that a lot of the "audiophile" angle is more about the ceremony of listening, rather than actual the sounds hitting their ears. The idea that everything is set up just-so with high quality components adds to their enjoyment.

        For example, I have some very nice drinks glasses. Although I have no expectation that the beverage within actually tastes any different; sitting down and relaxing with a drink from a very nice glass is just generally a more experience pleasant than slugging something down from a generic cylinder whilst in the middle of some household chore. It's not just about the beverage, it's the whole experience package.

        Mind you, audiophile pricing is just crazy. Even my nice drinks glasses were less than £10 each. I do find it hard to believe the "ceremony" bonus in listening can really be worth that much. Still, it's their choice.

        1. Neil Barnes Silver badge

          Re: a room full of audiophiles couldn't tell the difference

          Well yes. But I'd rather have thirty thousand quid's worth of music and a carefully considered three hundred quid pair of speakers than thirty thousand quid's worth of hifi and one CD to listen to...

          Each to their own. The world would be very boring if we were all rational in the same way.

        2. Annihilator

          Re: a room full of audiophiles couldn't tell the difference

          To be fair, based on an Infinite Monkey Cage episode I heard recently, the taste of something is definitely affected by the receptacle it's served on/in.

          https://www.thenakedscientists.com/articles/questions/does-metallic-cutlery-affect-perception-food-taste

          I didn't hear the episode very well though, because it wasn't a lossless encode and my speakers are wired up with coathangers.

    3. Anonymous Coward
      Anonymous Coward

      Re: Analog kid in a digital world...

      Be interested where you're getting a "proper analog signal chain". Vinyl a) is still a facsimile of the original, not a faithful recreation thanks to many many limiting factors in the surface of the vinyl - part of the skill of mastering is working around/with these limitations and b) there will almost certainly be a DAC somewhere before the vinyl lathe.

      But yeah, as someone has pointed out, Nyquist-Shannon sampling.

      1. Gloddata

        Re: Analog kid in a digital world...

        I always have found the obsession with vinyl formats hilarious. Not just because they are inferior to digital recordings in a technical sense, but because people somehow believe that they are more 'faithful' to the original when vinyl records use analog form of compression in order to get their long play times!

        It's called 'RIAA Equalization Curve' and it is done to allow more audio to fit on a vinyl record. Which is fairly lossy, btw.

        The problem is that low-frequency sounds have to move the needle a significantly further distance then high frequency sounds. If you go and look at a large subwoofer it needs to move much further and displace much more air to reproduce it's sounds then a little tweeter for the same volume. The same issue occurs on the surface of the record.

        So the solution was to use a high-pass filter that effectively mutes the low-frequency sounds, which is then the audio format recorded to the surface of the vinyl record. This allows the lines on a record to be much closer together, prevents needles from being damaged or thrown from the surface of the record, helps to reduce the effect of small bits of dust on the surface, and allows the record to last longer.

        Then when you play it back the low-frequency sounds need to be amplified (equalized) much more then the high frequency sounds to try to undo the high-pass filter.

        With digital formats this isn't a issue. The only thing that is used for digital formats to work is to have a low-pass filter that filters out frequencies that no human can hear and no speaker can reproduce. Which allows much smaller files and 100% accurate frequency reproduction using Nyquist-Shannon mathematics.

        Sure there is a sort of 'bounce back' effect if you try to reproduce sound that has a hard cut off at a high frequency. Which is why CD audio format uses 44.1Hz sampling rate, which is more then twice the frequency people can hear.

        1. Martin an gof Silver badge

          Re: Analog kid in a digital world...

          It's called 'RIAA Equalization Curve' and it is done to allow more audio to fit on a vinyl record. Which is fairly lossy, btw.

          Not sure what you mean by "fairly lossy", but the main benefit of the RIAA curve is a reduction in reproduced noise. Yes, the reduction in amplitude of low frequencies will allow closer grooves and indeed reduce the likelihood of the needle jumping, but even "quiet" vinyl has quite a lot of random noise in the groove, mostly at high frequencies. By boosting the level of higher frequencies in the signal which is recorded, and reducing them on playback, the background noise from the groove itself is also reduced. Similar systems have been applied to most analogue formats, from cassette tape to optical film soundtracks to FM broadcasts.

          M.

  8. Kevin McMurtrie Silver badge
    Thumb Up

    I might try it

    I quit buying AAC and MP3 albums years ago. For me, an MP3 album has many annoying defects from bandwidth deficits while an AAC album has around 0 to 3. I don't like those odds when lossless purchases aren't significantly more expensive. Even if it's background music while I'm working, those squeals or fuzz-outs are distracting. I could see myself occasionally paying extra for lossless streaming too, but not always.

    The problem for Spotify is going to be convincing phone makers to restore headphone jacks or convincing the Bluetooth SIG to get serious about audio quality and compatibility.

  9. DudleyDuoFlush

    Tidal Hifi vs Current Spotify

    I tried a direct comparison between Tidal Hifi and Spotify using Sony Bluetooth headphones connected using LDAC. I played the same tracks and switched back and forth. Tidal Hifi certainly sounded a bit better with more space between instruments, more balanced, more detail etc but the difference wasn't big enough for me to move away from Spotify. I'll certainly give Spotfiy Hifi a go but don't expect a massive improvement.

  10. fpx
    Boffin

    CD vs MP3 Blind Test

    Twenty years ago, German c't Magazine arranged for a blind test of CD vs MP3, using high quality equipment in a high quality sound room. Some test persons were chosen for their professional background in music, some were chosen at random. Music was played, and they had to choose whether it was from MP3 or CD.

    The bottom line is that the test persons could distinguish 128 kBit/s MP3 versus CDs better than average, though not reliably. For 256 kBit MP3 versus CD, it was a coin toss, i.e., the test persons chose correctly as much as incorrectly.

    Original article in German: https://www.heise.de/ct/artikel/Kreuzverhoertest-287592.html

    1. Mike 137 Silver badge

      Re: CD vs MP3 Blind Test

      A while back I gave a lecture on digital audio, during which I played a number of replicates of the same track recorded at different bit rates and sample depths. I played them in order of increasing bit depth/sample rate asking the audience if they could distinguish between them without prompting about their parameters.

      The recordings were made at bit depths from eight to 24 and at rates from 44.1 to 96 kb/s. From the slowest/shallowest right to the fastest/deepest members of the audience detected gradual improvement. However they were all resampled to 44.1, 16 bit before the talk, so beyond that point they were actually all the same. Psychologically, the detectable improvement among the first few tracks conditioned an expectation of continued improvement that didn't actually exist beyond that point.

      Sensory perception is extremely subjective.

  11. Anonymous Coward
    Anonymous Coward

    All recorded music is shit

    Can’t remember who it was, but years ago What-Fi or some such magazine had an article about musicians hi-fi setups. One guitarist said expensive hi-if was a complete waste of money. He reasoned that once you’ve stood in front of a valve amp and a 4 x 12 cab with a Strat plugged in, everything else sounds shit in comparison.

    1. Androgynous Cow Herd

      Re: All recorded music is shit

      I depending on how the valve amp was set up....everything else sounds like shit afterwards due to tinnitus.

    2. Someone Else Silver badge
      Happy

      Re: All recorded music is shit

      There was a time in my misspent yout' where I would play rock records through my Fender Bassman amplifier (input wye'd between the two channels) connected to a '69 vintage Bassman cab (the bigger ones, 2 x 12") and a Fender Showman cab (1 x 15"). No, it was not stereo. But put out on the balcony of my dorm at 7:30a on a Saturday morning, playing Grand Funk Railroad's Live Album full-out, after a hard night of partying...it was Reverie for the Gods!

      I would concur with said rock musician! (At least, back then I would...)

    3. R Soul Silver badge

      Re: All recorded music is shit

      WTF?

      How could anyone possibly hear anything after standing in front of a valve amp and a 4 x 12 cab?

      1. genghis_uk

        Re: All recorded music is shit

        And yet, in a band situation the bloody guitarists still complain that they can't hear themselves and turn up!!

        As a bass player I can only hear the screaming Marshall stack on one side and the white noise of the crash cymbal on the other (on a good day with bit of snare thrown in). In many gigs I have worked on the air from my cab hitting the back of my legs so I know something is happening...

        Live music may sound great to an audience if well mixed but it rarely sounds good to the people playing it. Even with great monitoring each player has their own mix so they are not hearing the front of house version.

        1. Someone Else Silver badge

          @ genghis_uk -- Re: All recorded music is shit

          Hey...What kind of cab (and amp) are you using? As a bassist, one of my goals is to find a rig that, when I play my low E string, it knocks my feet out from underneath me. Sounds (pun intended) that you may have found that rig.

          All you have to do, it seems, is turn it up a bit...

          1. Anonymous Coward
            Anonymous Coward

            Re: @ genghis_uk -- All recorded music is shit

            One that has a volume control that goes to 11?

          2. genghis_uk

            Re: @ genghis_uk -- All recorded music is shit

            The rig is a Genz-Benz Shuttlemax 9.2 into a Barefaced cab... Believe me it is plenty loud enough when I can use it properly.

            The problem is usually the venue - boominess that makes the whole lot sound out of tune or more usually, tight for space so I stand next to my cab. As Bass takes a big of distance to resolve, standing less than 4ft in front means I can't actually hear what I am playing despite the volume. Now fixed with in-ear monitors but that has its own complications.

        2. Martin an gof Silver badge

          Re: All recorded music is shit

          Speaking as someone who occasionally "does" PA for small scale gigs, and has to do FoH, a monitor mix and a recording mix at the same time, single-handed, give me an acoustic band any time :-) The less monitoring I have to set up - because the musicians are used to listening to each other - and the less amplification the instruments need, the better.

          The FoH mix starts from a position of "I can hear it perfectly well without any PA" and is set up so that voices can be heard over the louder instruments and quieter instruments aren't lost, while the recording mix has to include everything.

          Then you crowd 2,000 noisy people into a small, highly reverberant hall which ideally only sits 200, put a band on stage that only arrived five minutes ago and hasn't done a sound check and the temptation is just to crank everything up to the end stops, add a bit of reverb to the singer and rely on the speaker protection to prevent the tweeters burning out when the bloke with the penny whistle gets a bit too enthusiastic...

          :-/

          M.

  12. Mike 137 Silver badge

    Realistic limits

    While it's perfectly valid to do recording and post-processing at high bit depths (to provide headroom and counter cumulative numerical errors) bit rates higher than 96k are not only unnecessary but problematic. 96K allows a less than perfect Nyquist filter (they're all less than perfect as there's no such thing as infinite cut-off rate) to block frequencies above the audio range very effectively. indeed 48k can be enough with a well defined filter.

    When it comes to playback, 24 bit represents a signal range of 16 megabits. As the typical input level to an audio DAC is 1 volt RMS, a single bit is thus equivalent to 62.5 nanovolts. However in the real world, achieving less than a microvolt or so of noise in any analogue circuit is virtually impossible, which means that beyond 20 bit resolution (roughly 1 ppm) the rest of the bits will be noise, albeit effectively inaudible noise. Conversion to 16 bit can actually improve sound quality if appropriate noise shaping is used to push digitisation artefacts into the upper frequencies where they're less detectable.

    So the ideal is 24 bit 48k or 96k for recording and mastering, but anything over "CD quality" is superfluous for playback, particularly as every stage in the analogue audio chain from DAC to speakers will degrade the signal further (particularly the speakers - the worst stage in any audio system by far). This can be demonstrated using objective measurements.

  13. IGotOut Silver badge

    Is this the same Spotify...

    ...that is saying the world as we know will end should they actually pay artists a fair share.

  14. Martin Summers Silver badge

    2FA

    Spotify can't even implement 2FA on their service, despite an outcry from users in the forums for many years. Many people are having their accounts accessed and used without their permission which 2FA would have easily prevented. I grant you that user password hygiene could be better (I had my Spotify account compromised a while ago thanks to the LinkedIn breach). I'm disappointed they're wasting their time with this new feature when they can't even get security basics right.

    I asked El Reg to look into this a while back but alas no article was ever forthcoming.

    1. Anonymous Coward
      Anonymous Coward

      Re: 2FA

      I had my Spotify account compromised last year. Suddenly my phone started telling me I was listenting to Latin rap & salsa music on another device. Very different from my usual 70s - 90s rock stuff. I reset the password and looked for 2FA, I was surprised they don't offer it.

  15. jdiebdhidbsusbvwbsidnsoskebid Silver badge

    192 kHz??? FFS!

    192kHz sample rate? So that'll be a bandwidth of 96kHz then. A quick check of Shure's own claimed bandwidth for their professional musician standard SM58 microphone is only 15kHz. So standard audio CD bandwidth of 22.05kHz is already more than sufficient in that respect.

    Actually, Shure's own spec sheet seems to claim 15kHz based on a 10dB bandwidth rather than the more commonly accepted definition of 6dB. The 6dB bandwidth looks more like about 12kHz but it's hard to read by eye off the log scale.

    1. Martin an gof Silver badge

      Re: 192 kHz??? FFS!

      The SM58 is a vocal microphone. It has been the 'standard' for performance for decades mainly hecause of two 'killer' features, firstly it has a pronounced 'presence boost' which gives male performers in particular more 'body' in their voice if they sing close to the microophone (look at the frequency plot again and note the 'hump' in the lower human voice frequencies, this is a common trait of cardioid pattern dynamic microphones) and secondly - and arguably more importantly - it is built like a tank. It will shrug off all sorts of physical and electrical abuse.

      Oh, and it's cheap-as-chips.

      You would not normally use an SM58 to record instruments, though it's not too bad to be honest, especially at distances where the presence boost isn't a problem.

      But there are quite literally /hundreds/ of microphones available which have a far greater, flatter frequency response than the SM58 and can easily capture 20kHz or more and - of course - electronic instruments can be directly connected, requiring no microphone at all.

      As has been mentioned several times, high sample size (24 bits) is a boon for processing, giving much more headroom, and high sample rates (96kHz or more) mean that down-sampling for playback can be cleaner.

      M.

  16. Natalie Gritpants Jr

    52% preferred the MP3, 30% the lossless audio, and 185 could not tell the difference.

    Tells me that 92% of people can tell the difference between lossy and lossless, and that most of them picked the wrong one because that's what they are exposed to.

  17. Anonymous Coward
    Anonymous Coward

    Simple Observation

    If you listen to compressed music for prolonged periods, you will become accustomed to the idiosyncrasies of the codec employed.

    This might manifest as a preference for the codec's tonal balance, or "punch'. Alternatively, it might be as a nagging frustration that a favourite track section loses its phrasing, or depth.

    Lossless artefacts are definitely there and sometimes are even deliberate - to enhance the listening and disguise the codec's shortcomings in another area.

    How MP3, Ogg and AAC are perceived depends on the individual, the listening environment and the equipment.

    I know my preferences, but they are irrelevant to anyone except me.

    1. Neil Barnes Silver badge

      Re: Simple Observation

      Can you point me to a handy test set that can measure 'punch', 'phrasing', or 'depth'? I've been looking for years...

      The amazing thing about lossy codecs is not that they work so well, but that they work at all. I mean, you take a 24ms chunk of audio and use it to calculate a gain factor. You then slice it into three segments and convert each 8ms chunk into the frequency domain. You throw away some or all of the frequencies which your perceptual model thinks the listener won't notice, and depending how many bits you have to play with, you throw away some which they will, but hopefully not too much.

      Then you stick it back together, converting back to the time domain, and hope it bears some resemblance to the input signal... generally it doesn't, but the ear is fooled.

      Amazing.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon