Audio compression

This is clearly written by someone who did just enough research to evoke buzzwords haphazardly. "Headroom" and "dynamic range" are bandied about in the same marketing-speak employed 30 years ago to sell the CD as "perfect sound". "Nyquist" is cited as the bedrock of audio science, though it's long been discredited in theory and practice. Saying that higher sample rates have "failed to catch on" - no one can arbitrarily impose an entirely new standard - is like suggesting that more cars would have left-hand drive if only the public wanted it, ignoring that manufacturers refuse to provide it.
The author momentarily gets into the groove explaining the sacrifices inherent in mp3 compression, but soon veers off again when comparing CD quality to DVD. The reason that Neil Young embraced Blu-Ray is because it sounds more like the original recording than CDs do. Blu-ray by definition has "lossless codecs" since it is uncompressed, but so do CDs. The reason Blu-ray sounds better is higher sample rate (96 K vs CD's 44.1) and bandwidth (24 bit vs 16 bit.) Incidentally, this is why people believe that vinyl can sound better than CDs: vinyl preserves more of the original information. While it's true that with mp3s, as previously with cassettes, consumers were willing to sacrifice some quality for convenience and price, it didn't hurt that hip-hop listening habits consisted of turning down the treble and cranking up the bass, rendering fidelity largely irrelevant.
It's an economic reality that CDs revitalized the record industry because Sony/Philips and the labels were able to persuade record buyers to not only buy new music on CD but also to buy their old libraries again. Musicians certainly have no economic reason to complain about new technology; some chafe because their music sounds less good than their original recordings.
The author's hilarious conclusion that people simply prefer that to which they are accustomed regardless of fidelity is especially fatuous considering the pretense that this is a scholarly work. Imagine how devastated Victrola owners must have been when forced to abandon their tinny horns for full-range speakers: "In reality, they had simply become so attuned to the clicks and crackles, as well as the limited dynamic range, of the older format that the familiarity made them feel comfortable."
Thank you for suffering my rant, and kindly excuse me while I retire to my comfy old straw mat on the floor - just can't get used to them newfangled bedsprings...

Apologies if anybody else has brought this up.

Beyond the more technical inaccuracies and improbabilities raised in this article, the likelihood that the youth of the future will be listening to music on Blu-Ray or, indeed, any other kind of disc is essentially nil - and the idea is rather quaint.

Codec and algorithm aside - and much as I miss the illicit hometaping and cd burning of my past - physical media won't be making a comeback.

Personally, I *tried* to be an audiophile but couldn't cope with the nonsense - and am now very happy with Spotify.

Muffled BASE?
Something "fishy" about that!!
It is BASS for a low note.
Bass trombone.. stringed bass... bass clarinet... bass singer
ecc.

What a load of hogwash. Compression, low bit-rates and MP3 files have killed the music industry. Music is not as interesting to listen to anymore because of these practices. MP3 is to great audio as McDonalds is to fine restaurants. It is remarkable that the photographic industry has pushed the concept of ever-higher resolution while the poor music industry is left with crap.
Thanks Mr Jobs and Apple Inc.

I've done some controlled listening tests on a small sample of high school and college students to see if they prefer CD-quality versus MP3 (128 mbps), and accurate versus inaccurate loudspeakers. The students generally preferred the more accurate CD-quality formats, and the more accurate loudspeakers. None of them preferred MP3 over CD. While this study is still ongoing the early results refute Professor Berger's finding,s which were informal and never published.
Kids prefer accurate sound reproduction when given the opportunity to directly A/B good sound versus bad sound. Unfortunately, with the demise of the boutique stereo store and the growth of the electronic sales on the internet, the opportunity to audition and compare audio gear,etc is diminishing. To me this suggests an opportunity for educating the public about good sound, and the music/audio industries need to come up with a marketing strategy to sell kids a higher quality listening experience. They can hear just fine but they need to be shown the path to achieving a better quality listening experience.
References
See 1) http://seanolive.blogspot.com/2010/06/some-new-evidence-that-generation-...
2) http://www.aes.org/e-lib/browse.cfm?elib=16108).
Cheers
Sean Olive
Director Acoustic Research
Harman International

Surely you know "RIAA equalization". Low frequency sounds generally become too fat in grooves of vinyl records, easily disturbing the adjoining grooves. There therefore a beautiful technique is used: sounds in lower frequency region are artificially supressed when they are recorded, and artificially amplified (equalized) when regenerated, through the pre-amplifier, both normally according to RIAA specification.
Yes, there's also a kind of compression/decompression process in good old days vinyl records.

Interesting article, however as Gordon L points out, there is a renewed interest in vinyl recordings that the article does not address.

I can't remember exactly, but year-on-year sales in the UK are up by a substantial percentage, albeit from a low base.

"A future generation—trained to hear a recording’s subtleties burned by a lossless codec onto an audio Blu-ray Disc—will be puzzled by their parents’ preoccupation with sizzling songs rather than an authentic replica of the music the performer actually created."

Got news for you bub.

My 17 year old is maddly keen to buy vinyl records.

One part of the problem is a lot of us spent a lot of money on CDs, which we were fortunately able to turn into MP3s once that market took off. Now I own a lot of music in that format. I suspect it would cost me a lot of time, if not money, to update my collection to FLAC. And then, of course, will future players even support the format?

The problem described here has been going on since AM car radios became popular and studio engineers did their final mix while listening to the pop song through a single standard 6x9" dashboard speaker.

There's an inherent conflict between what an audiophile wants to hear when sitting down to do nothing but listen to the music, and what the much more typical listener wants while driving a car or working in a factory or at a construction site. When you have a lot of ambient noise, a concert-hall dynamic range is not just useless, but it's annoying. If you make it loud enough to hear the quiet parts, the loud parts are ear-splitting, and if you turn it down so the loud parts sound good, the quiet parts are inaudible. One solution that has been available for a long time, but is rarely offered, is an analog dynamic range compressor. Part of why it hasn't been offered is probably that customers don't know they'd like to have it. The other part is that it can only react after the fact, which means it causes an annoying "pumping" or "breathing" effect.

Digital audio codecs finally provide the ideal way to satisfy both kinds of listeners with a single digital audio file. The codec itself could include dynamic range compression selectable by the listener at playback time. Since the codec, unlike an analog compressor, "knows" what is coming in the music as well as what has already happened, it could apply any desired level of compression to the whole song before it's played back, with the result being perfect compression to any desired extent, and no distortion. The cost would be merely software, which is essentially free today.

As an aside, the time may soon come when storage is so cheap that there is no need for lossy compression. In that sense, we'd be migrating backward from .mp3 files to .wav files. There would then be no need for a "codec" in its current sense of the term, but it could be replaced by a playback processor that applies equalization and amplitude compression functions to an unencoded data stream.

As storage and bandwidth become cheaper yet, artists could choose to give the customers all the tracks they laid down in the studio in raw form, and let the customer play "recording engineer" by mixing them however they wanted. Maybe this would be too much of a gimmick for most artists and listeners, but I could at least see a return to giving the listener more than two tracks. 8-track or quadraphonic, anyone?

Ultimately, the best way to utilize digital music will be to store as much information as possible (more samples/second, more bits of resolution, more channels of input) in the music file, and let the playback device optimize it for a given listening environment and listener preference. This need not overwhelm the non-technical user with unwanted choices. It would be fairly trivial, for example, for a car stereo to adjust its amplitude compression depending on ambient noise and adjust its equalization based on both the ambient noise spectrum and the source material power spectrum.

"Rock-and-roll, as usual, is leading the way. Bands such as Pearl Jam and Metallica have used FLAC to sell recordings of their concerts online."
This is not true. They're simply adopting and commercializing a long-existing free approach (see, for example, http://en.wikipedia.org/wiki/Etree) created by fans of the Grateful Dead. There have been audiophiles trading lossless recordings of live shows for decades (think analog torrents), and they're the ones who have been "leading the way" in experimenting with and driving demand for new formats.
The hippies led the way with a free-for-all model, and the rock-and-rollers simply commercialized it (at least this time the rock-and-rollers didn't try and strangle it, cough napstercough).

'Their preference is similar to the way audiophiles from a previous generation swore that vinyl LPs produced a warmer, richer sound than CDs. To their ears, they did.
In reality, they had simply become so attuned to the clicks and crackles, as well as the limited dynamic range, of the older format that the familiarity made them feel comfortable.'-Nonsense.

High Res digital will probably never approach the sound quality of vinyl. Think about it. For digital to be reproduced it needs to pass thru thousands of transistors in the DAC. I have a $38,000 CD/SACD player and though it sounds much better than lesser players it's no match for vinyl. SACD is only slightly better than CD. Vinyl builds a more solid and stable soundstage that puts you back in the recording venue. Even as the music fades away at the end of songs this solidity and stability is uncanny. Would it still sound that good if my phono stage was made of thousands of transistors?

'The compression-decompression algorithm (“codec”) used relies on psychoacoustical tricks to remove less audible parts of the signal—like a quieter sound masked by a louder one occurring at the same time, or notes near the limit of human hearing.

Such “lossy” codecs are widely used where the loss of some portion of the data is acceptable'- is this really true or are the tecks desperately creating analogies for clueless journalists. I have 3TB of digital downloads and to me MP3 at 320 sounds better than FLAC. FLAC's loudests sounds are at the original level but the rest is brought up in level. The effect is that the FLACs sound like a remix.

nt

This article is probably confusing to the layperson. It starts out talking about dynamic range compression (which is often simply called "compression" by audio engineers) and then jumps into the unrelated area of bitrate reduction (confusing, called "compression" by computer scientists). I suspect the average reader would not even notice that two completely different processes are being discussed.

There are also a number of factual errors. For example, this sentences is not at all correct: "Some engineers argue that a CD has so much dynamic range that a portion can easily be sacrificed for compression." It is not the large dynamic range that allows engineers to sacrifice it for compression (this makes no sense, sacrificing dynamic range for less dynamic range??). Moreover it usually not the engineers who are pushing for less dynamic range -- in fact most engineers would prefer more. Usually artists and/or A/R reps push for less dynamic range because they believe it makes their tracks more "competitive". For more on this, Bob Katz, a top mastering engineer, devotes a entire chapter of his book "Mastering Audio" to the "loudness wars". It is also an oft discussed topic on online audio engineering forums (not to mention the water cooler at recording studios).

Later, the article states "Clearly, the iPod generation is becoming attuned to the “sizzle” caused by a muffled base and clipped high notes that MP3’s lossy codec imparts. Their preference is similar to the way audiophiles from a previous generation swore that vinyl LPs produced a warmer, richer sound than CDs. To their ears, they did." Some early research supported this, and it has unfortunately become common wisdom, but there just isn't a lot of data to back this up: http://www.cepro.com/article/harman_debunks_youthful_music_myths/K536

This makes the article's claims such as "All agree, though, that lossy algorithms of the past need to be replaced by lossless ones" highly dubious. Bitrates are certainly going up, and that's great, but are lossless algorithms inevitable? Maybe, maybe not. Maybe sample rates will go up, but we'll continue to download lossy codecs. Not everyone agrees about the future. Most mastering engineers will tell you that the key to higher sound quality is less dynamic range compression, not less bitrate reduction. Unfortunately, they continue to compress the hell out of tracks because that's what the clients are demanding. It's a shame this article doesn't adequately distinguish between dynamic range compression and bitrate reduction.

A good article.

But to be pedantic 'codec' stands for coder-decoder rather than compress-decompress. codecs often compress but many in the network and in your PC merely re-code for other purposes.

It's not the format at ALL, really. It's the Loudness War that makes CD's sound like crap these days -- even more so when copying a brick-walled CD to ANY format. The digital brick-wall compressors, when run hot enough, CREATE those artifacts, making it sound so awful.

I'm no expert, but from what I understand, visual mediums have a universal audio standard, a certain upper limit on the mean volume of the material. Why compress the audio so high when you can only GO so high? So those transients and that sparkle remain as the original artist, producer and engineer intended, which is WHY DVD's and Blue-Ray sound so much better, and even an MP3 copy will sound better too.

Later I tested several wireless Bruetooth products' sound quality.
There is a tendency that the sounds decay at higher (10kHz or more) frequencies, on a FFT analizer, probably due to CODEC, which may explain their net sound quality.
I have not tested MP3 or others yet.

Very glad to find this hot discussion in the Economist.
Once I have tested 16bit and 14bit sounds from the same source with my hand-made recorder(100kHz sampling), and was surprised --- or beaten --- by the fact. They were not discernible to my poor ears.　It was mostly because I tried to use the full range of 14 bit, I hope.
Since then I've strongly convinced that it is not dynamic range that count, but 'available' dynamic range from the noise floor: 16bit with zero noise is preferable to 24bit with 10bit noise floor.

Audiophiles are deaf to the fact that they are not the majority of the audience for music any more. When recordings were reproduced by a delicate mechanical device, the LP turntable, you had to have a quiet, acoustically isolated environment for listening. With digital playback, even via CD's, the listeners have moved first to cars, where only luxury models are super-quiet even now. iPods and other MP3 players have moved the listening experience out into the noisy real world, where compression is needed in order to get the music level up over the ambient background static, whooshing, babbling, rumbling, roaring, and crashing, without increasing the acoustic intensity to levels that damage hearing.

In order to keep dynamic range for mobile listeners, you have to make the rest of the world quiet. Good luck with that!

"The CD lost out because the record companies were slow to embrace the social and technological changes engulfing their business."
Record companies, not CDs, lost out because they existed solely to parasitize music distribution. There is no longer a business model to support their existence. I hear better quality music, hourly, for free, from unknown artists everywhere, than anything music company parasites have ever blasted into my ears.