hifi

Romancing the Sound: A 24 bit Sound Odyssey

There are two schools of thought when it comes to music. There are those for whom the quality of the sound plays no role in the musical enjoyment. Only the notes matter. If music is played with some feeling and it’s reasonably on key, then the truth of the music is preserved. Then there are those for whom searching for the perfect listening experience is half the enjoyment. They experience the enjoyment of music and the details of its performance – with their whole bodies – feeling the vibrations and “seeing” the music in space ahead of them, savouring every nuance of the instruments and the acoustics of the recording.

For someone like me, who grew up with music in the 1970s when hi-fi technology was becoming a reality, Hi-Fi was as fascinating as smartphones are today. There was genuine excitement over what technology could accomplish. Today, it could all seem moot, given the enormous improvements in sound quality that have occurred over those 50 years. Today, anyone with a good pair of headphones and a smartphone can enjoy better hi-fi sound quality than the best sound systems then. But that doesn’t mean that there isn’t something far better than that default experience, for those predisposed to seek it.

There is some suspicion of hi-fi enthusiasts by the first type of listener—are they more interested in the technology than the music? Sometimes, that can be true. Some listeners buy accessories rather than sound equipment, and the politics of it are quite ridiculous (if your equipment looks too nice, then you are not a serious sound person, just an accessorizer, sigh).

Actually, the general understanding of what hi-fi is about is also patchy. If I had 10000 kroner for every time someone told me that “better loudspeakers” was the way to better sound, I could afford those best speakers in the world. But actually, the loudspeakers are probably the last item on the list of priorities for an excellent hi-fi experience. Hi-Fidelity doesn’t mean “sounds expensive”, it means “accurately reproduced”. Fidelity is faithfulness. The idea of hi-fi is that what you hear at home sounds exactly like what was recorded in the studio, not that you maximize the bass and treble. True hi-fi doesn’t have equalization controls, just a volume knob.

Too hi or not too hi?

In truth, if your go-to genre of music is pop or rock, you can relax. These forms of music are traditionally so processed and so compressed (in order to sound clear on lo-fi radio) that there is nothing a hi-fi system can do to make them sound any different. They are stripped of dynamic range and extreme frequencies, and mixed for the masses. Occasionally digital remasters can add a little integrity to old recordings, but on the whole it’s not much.

The real benefits of hi-fi come to those who enjoy complex music, especially orchestral music, and acoustically recorded instruments and voices, where the soundscape comes from the raw sound rather than the production process. Film music enthusiasts must fall into this category, since many of its classics have been recorded with the best of modern digital equipment, especially since George Lucas went on a mission to improve sound recording after Star Wars.

Linn Akurate DSM

Today, music is delivered to us on a variety of media. There is streaming audio (mp3 and similar varieties of compressed music file). There are CDs. There are also super CDs, though I’ve never seen one, nor have I seen a player that could play one. There is vinyl, with its romantic retro appeal. And finally, there are hi-res digital audio files, which can be downloaded or heard on blu-ray and other high density media.

Does it matter which we choose? Well, yes – but only if your music playing equipment is up to the task. A Linn Akurate DS based hi-fi is the one expensive thing I bought in my life, apart from a house – and it continues to give me pleasure every day.  If you listen to pop and rock, you won’t hear much difference in the amount of detail from these sources, but if you listen to symphonic music, there is a world of difference to be had. Vinyl will add its own `romantic’ pops, scratches, and E.M. noise; whether vinyl contains more information than a digital sampling (as some try to argue) or not depends on the recording, the mastering, the production, and the quality of your player. What we can say is that vinyl can’t compare to the best digital sample files for fidelity. You might like the sound better, but it can’t compete objectively with digital.

Film orchestration

Film music enthusiasts may well find that the sound they hear at the cinema or on blu-ray is  better than the CD they buy afterwards. That’s because the sound recording is in 24 bit digital, whereas CDs are only 16 bit digital, with lower sampling rates.

Music is compressed in order to squeeze it onto a CD. The sampling frequency is usually what people focus on. 44.1 kHz is standard for CD, but pop and rock music is often compressed far below that level. A well mastered classical recording will more than saturate the amount of detail that can fit onto a CD, so there will be loss – not in the frequency range really, but in the dynamic range. So-called high resolution (hi-res) music is 24 bits sampling, with the depth to capture the full range of an orchestra. The difference is astonishing. Suddenly, details that were muddled on a CD can be preserved with a totally new clarity.  If you are a fan of vinyl, you will not be able to approach the dynamic range or the frequency range of hi res digital on any level—though no one can deny the tactile appeal of vinyl, and the ingenuity of the technology. Music may be compressed even further by mp3, but that may not matter too much for pop and rock, which are its main uses.

Star Wars, Episode IV: A New Hope (1977)

If you want to test some of these assertions, you can follow the evolution of sound recording in the familiar Star Wars music. A New Hope is one of the most re-recorded orchestral soundtracks, and its many versions (both in different masterings and releases of the same recording, and in rerecordings) span the time from before digital recording began. The original recording from 1977 was harsh and brittle sounding, with a lot of microphone noise and more than a few bum notes from the orchestra. After that, each movie had notable improvements in the sound recording quality.

A lot of music from the 1970s had poorly tuned instruments, sharp or flat. Part of the reason was that the equipment was not as sensitive to these subtleties then, so they were less obvious– but they sound painful on modern equipment.

The modern recordings are mastered digitally, and the old recordings have also been resampled and cleaned to reveal all their inadequacies perfectly. Comparing, say, the music from The Last Jedi (in 128kHz/24 bits) with the recording of A New Hope will leave you laughing at the difference.

I used to find the entire Star Wars catalogue hard to listen to, because the music is quite dense and brassy (two qualities that suffer most from compression). The new 192/24 recordings had me listening with a whole new set of ears, and without feeling the effects of “brass fatigue”. The choir in «Duel of the Fates», which sounded flat and two dimensional on CD, was suddenly three dimensional and cathedral-like. For the first time, I listened long enough to hear all the amusing hidden references in Revenge of the Sith, where John Williams seems to quote passages from other movies (James Horner’s Aliens, Jerry Goldsmith’s «V’GER» theme, and even a passage from Jurassic Park when the dinosaurs run wild).  You can also compare all the old and new CD recordings and compilations of Star Wars, recorded in better studios over the years too. John Williams has at least three separate recordings of the Star Wars suite, with 10 years between each, charting an interesting course through the annals of sound quality.

Where to get it

The problem is that very little music is released in higher than CD quality – mostly we are cutting back to even higher compression rates – and very few listeners have equipment to play anything of higher quality anyway.  HDtracks.com and a few other companies release some music in hi-res. Your smart phone might even be able to play some of it, but alas movie music is not their priority. Rock and pop (which have the least to gain) dominate their catalogues.

Big budget movie soundtracks do tend to get released in hi resolution. So the Star Wars catalogue (the most popular of all time) is now complete, and in maximum quality, with amazing results. Jurassic Park’s 20 year anniversary was released too. Personally, on the John Williams front, I am hoping for the Harry Potter movies to get the same treatment. Beyond these predictable cases, the choices seem quite random.

The Red Sparrow (2018)

James Newton Howard’s unusually beautiful score from The Red Sparrow is a great addition that makes full use of the the high resolution medium. The Hobbit movies (Howard Shore) are nicely recorded, as is the haunting music from The Revenant (Ryuichi Sakomoto, Alva Noto, Bryce Dessner). Then a few Hans Zimmer scores are available: The Dark Knight, Interstellar, for example. Alan Silvestri has a few titles: Back to the Future, The Avengers and Ready Player One, Night at the Museum. And, for Star Wars or How to Train Your Dragon fans, John Powell’s score for Solo: A Star Wars Story is also available in hi-res (better without the movie) – with a bizarrely similar theme to Silvestri’s Ready Player One. Bear McCreary’s God of War soundtrack (for the computer game) makes a stunning recording of beautiful music (a strong recommendation for music aficionados), but alas his superb Battlestar Galactica music is not available (the taiko drums alone would definitely benefit from a 24 bit mastering).

For those cases where a CD is the only alternative, modern recordings and remasterings are always superior to old CD pressings. An excellent example was the iconic soundtrack to Alien by Jerry Goldsmith. The original mastering was clipped and distorted. It took 20 years to get a remastering that was both complete and non-distorted. Today, everything is digital from end to end, and its almost impossible to get a bad quality mastering, but that quality is mostly lost.

Sounds of the Unexpected

The effects of sound quality are often not what you expect. We are so used to hearing music compressed for distribution that it can be unexpected when it sounds like the real thing! The sound picture is not simplified for us to make it easier to cut through background noise. You might have to listen more carefully, but you will hear hidden depths.

In truth, there are many things you can do to improve a listening experience without paying more for a new stereo. You will always hear more from a proper sound system with speakers than from headphones, though headphones might emphasize certain details. Sound quality starts with the source, and the weakest link will be your bottleneck. Often, arranging for proper acoustics in your listening room may actually be the most noticeable improvement. New clean cables can also make an unexpected difference.  If you listen to jazz orchestras, there are more options to buy quality. Joni Mitchell in front of an orchestra on her album Both Sides Now is available in 192/24, for instance. Less compression means you hear the whole orchestra, not just her voice artificially made louder on the CD. It pays proper respect to the musicians behind her.

Film music is written to add emotional resonance to the telling of a story. It conjures pictures and feelings, as few other kinds of music is designed to do.  If the story is blurred and compressed, something is lost. Anyone who appreciates the beauty of the orchestra, even when mixed with electronica, can rediscover musical performance in hi res recordings. I’m certainly hooked.

Visits: 74

  1. Sigbjørn Vindenes Egge sier:

    Interesting article, Mark. It really shows you’re a passionate listener for sound quality.

    I noticed you write about compression without really defining it. I take it you mean dynamic range compression?

    1. Mark Burgess sier:

      Thanks Sigbjørn. On the compression issue—compression is quite a complex topic, involving both frequency filtering and dynamic range compression (loudness). These days, digital compressions (like lossy mp3) may reduce both frequency and dynamic range in order to minimize file sizes. In pop and rock, it’s common to do this already at the mixing desk to avoid changes in the experience during mastering. Then there is FLAC (lossless compression) which is a purely file size compression, which does not affect frequencies or dynamic range. The true benefit of 24 bit lies in dynamic range though—no matter how much hi-fi enthusiasts focus on frequency, most of us are completely deaf to frequencies about 18kHz. The loss of dynamic range tends to emphasize certain voices over others, meaning you hear a simpler picture (great for radio and noisy backgrounds playing in the car, etc). To hear a full orchestra with the scratching of the bows and the breathing into the brass, its the dynamic range that counts.

  2. Sigbjørn Vindenes Egge sier:

    I agree there’s no great benefit of increasing the sampling frequency beyond that of the CD. When it comes to the dynamic range, I’d argue that the CD’s range of 96dB is quite sufficient – that it’s not always fully utilized is a different matter. After all, whisper is on the level of 20-30dB, while 110dB is the average human pain threshold. Therefore, I’m having difficulties seeing why a greater available dynamic range would do the music any good – the record producer won’t be able to use the whole range anyway.

    1. ^^^^ Agree totally! The CD’s 96db is more than sufficient. The reason some “high rez” music sounds better is because of the remastering process; it has nothing to do with the resolution itself.

      Good article on this here:

      https://www.whatsbestforum.com/threads/is-the-dynamic-range-of-cd-sufficient.13461/

      Quoted:

      “…others have argued that the 16 bit dynamic range and high frequency issues, quite apart from practical filtering problems, indeed pose theoretical problems. I had contended that the dynamic range is more than sufficient, since you can’t get below something like 25 dB background noise (very quiet living room, or basically empty concert hall except musicians) and an orchestra only on very rare occasions reaches 110 dB (which is screamingly loud). This would make for a maximal dynamic range for orchestral music of about 85 dB (at 110 dB – 25 dB). A similar dynamic range for orchestra (around 70-80 dB) has been cited on several websites, including on the Wikipedia page on dynamic range with a book citation on this. For recording that dynamic range, the dynamic range of 96 dB of 16/44 CD is more than sufficient.

      Yet even so, trying orchestra at maximum live levels at home in everyday listening situations is not recommended, since 110 dB is close to ear-damaging levels even for just short amounts of time (I never go beyond 95-97 dB for short orchestral climaxes at home, which is really really loud, and usually I keep max levels somewhat below that — after 2 minutes of the final brass chorale of Bruckner’s Fifth Symphony at 95-97 dB, reproduced without appreciable distortion, I already feel pressure in my ears). So I would say that makes even an 85 dB dynamic range unnecessary in home listening situations. Your noise floor in your listening room is a minimum of 25 dB — at least it is around 25 dB in mine, in a very quiet neighborhood, with sound-suppressing windows and no electrical gear except the stereo system, thus under really optimal circumstances. So for practical listening situations that would make for around a 95 dB – 25 dB = 70 dB effectively needed dynamic range.”

    2. Mark Burgess sier:

      Actually it’s a bit more complicated than that. Forgive me for putting on my physicist hat for a moment. Using frequency range and decibel dynamic range can be misleading. First, there is the issue of how many bits per second can be processed by the hardware. Digital formats (e.g. mp3) can choose this effective average rate by changing the encoding, regardless of whether the baseline is 16 or 24 bit. A good 16 bit sampling is better than a bad 24 bit sampling. That’s why digital master CDs give a big improvement, while making a 24 bit version of the same remaster will not necessarily improve anything. Unless your hardware is capable of responding to the encoding faithfully, you won’t get much benefit either. Dynamic range is a complex beast. The decibel scale is a power ratio. For compressed music, this is simply a matter of a pretty constant average signal to noise. But for something like an orchestra there are huge variations, say from PP violins to whacking a bass drum. A bass drum uses a huge amount of energy (power) to move air. You will lose a lot of information on transient signals like drums and plucks unless you keep a lot of high frequencies. The combination of deep and fast uses the most energy. High frequency implies fast changes (Fourier analysis). So you can try to keep the dynamic transients at the level of sampling using 128/24. But then, if the electronics that read this can’t deliver the power — this is why a good amplifier is the next line of defense in sound experience. Linn refused to make a CD player because their research showed that you can’t reliably sample 16 bits faster enough (error free) to render music correctly. So they jumped straight to digital from hard disk.

      I think the true differences between 16 and 24 bits are subtle and hard to reason about technically. You can only listen and hear the difference (or not). For me, I keep multiple versions to test the difference. It’s really on well-recorded “bare” orchestral music that the differences are most noticeable. I’d be happy to demo the difference one day. It still warms me inside every time I listen to those best recordings.

      1. I guess my only real comment here, analysis aside, is that the proof would be in comparing 16 bit and 24 bit recordings level matched and double blind. All the tests I have either read about or taken part in have found little to no audible difference. And the tests that did reveal an audible difference were only when barely audible content was increased in volume to such a level that when a loud transient hit you would either go deaf or the speakers would literally get blown out. In other words, it involved turning barely audible content up to totally unrealistic volume levels – to the point where HVAC noise in the recording studio was the loudest and most prominent part of the recording. And even there, the only difference was that this low level material was rendered with a bit less noise in 24 bit than in 16 bit. But I think it’s important to keep this all in a real world perspective – no one who values their hearing would ever listen at these kinds of levels.

        So for me, at this point these are just claims that have not been validated in any controlled scientific manner. Until such time someone can demonstrate audible differences in controlled listening tests, I remain skeptical. Please know that no disrespect is intended, I’ve just spent so much time involved with blinded and non-blinded testing, that I tend to think most reported audible differences that can’t be replicated under controlled conditions can most realistically chalked up to expectation and confirmation bias. We are all susceptible, me included, which is why I am such an advocate of blinding any and all listening comparisons when possible.

        Much of the double blind research into audible differences between sampling rates and bit depth is summed up in Chris Montgomery’s famous article about high res music – if you are not familiar, it is available here:

        https://people.xiph.org/~xiphmont/demo/neil-young.html

        What he says lines up almost exactly with what the acoustics researchers at Harman and the Canadian NRC have found in their testing (usually the work of Floyd Toole and Sean Olive). A couple of brief quotes from the above (though I recommend reading the article itself since he links to relevant studies, demonstration videos and sound clips):

        “It is also worth mentioning that increasing the bit depth of the audio representation from 16 to 24 bits does not increase the perceptible resolution or ‘fineness’ of the audio. It only increases the dynamic range, the range between the softest possible and the loudest possible sound, by lowering the noise floor. However, a 16-bit noise floor is already below what we can hear….

        Listening tests:

        Understanding is where theory and reality meet. A matter is settled only when the two agree.

        Empirical evidence from listening tests backs up the assertion that 44.1kHz/16 bit provides highest-possible fidelity playback. There are numerous controlled tests confirming this, but I’ll plug a recent paper, Audibility of a CD-Standard A/D/A Loop Inserted into High-Resolution Audio Playback, done by local folks here at the Boston Audio Society.

        Unfortunately, downloading the full paper requires an AES membership. However it’s been discussed widely in articles and on forums, with the authors joining in. Here’s a few links:

        The Emperor’s New Sampling Rate

        Hydrogen Audio forum discussion thread

        Supplemental information page at the Boston Audio Society, including the equipment and sample lists

        This paper presented listeners with a choice between high-rate DVD-A/SACD content, chosen by high-definition audio advocates to show off high-def’s superiority, and that same content resampled on the spot down to 16-bit / 44.1kHz Compact Disc rate. The listeners were challenged to identify any difference whatsoever between the two using an ABX methodology. BAS conducted the test using high-end professional equipment in noise-isolated studio listening environments with both amateur and trained professional listeners.

        In 554 trials, listeners chose correctly 49.8% of the time. In other words, they were guessing. Not one listener throughout the entire test was able to identify which was 16/44.1 and which was high rate [15], and the 16-bit signal wasn’t even dithered!… It’s important not to cherry-pick individual papers or ‘expert commentary’ out of context or from self-interested sources. Not all papers agree completely with these results (and a few disagree in large part), so it’s easy to find minority opinions that appear to vindicate every imaginable conclusion. Regardless, the papers and links above are representative of the vast weight and breadth of the experimental record. No peer-reviewed paper that has stood the test of time disagrees substantially with these results. Controversy exists only within the consumer and enthusiast audiophile communities.”

        If I don’t stop I’ll end up quoting the whole thing, lol. Again, I recommend reading the entire article since it gets into far greater depth.

      2. Sigbjørn Vindenes Egge sier:

        You’re most welcome putting on your physicist hat – that’s my kind of hat as well. 🙂 It would indeed be interesting to experience a proper comparison between CD quality and high res. In the end, what matters is indeed the audible differences, and that can only be properly compared by deriving the two qualities from the same master.

  3. Mark Burgess sier:

    I stand by my remarks–.including when you will / will not be able to hear a difference. I suspect that there can be no convincing scientific account of this (I am a physics PhD, so I don’t say this lightly), because there are just too many variables. But as someone who had been involved in music and its production all my life I am confident that I am not fooling myself, which is why I keep multiple versions of music to remind myself of the benefits of 24 bit masterings for those who have appropriate equipment and listening conditions.

    The goal is to be happy with one’s own experience. so my article is addressed to those who may be interested in what can be achieved with some dedication (and plenty of money).

    Enjoy your music!

Legg igjen en kommentar

Din e-postadresse vil ikke bli publisert. Obligatoriske felt er merket med *