Recommended reading: https://en.wikipedia.org/wiki/Loudness_war
edit: compression as in dynamic range compression, not data compression like mp3 in audio
Compression is the best tool we have for accurately reproducing the musicality and emotion of a musical performance. Without compression, most recordings would be unlistenable.
Don't confuse the foolishness of the loudness wars for "compression is bad". That's like saying the internet is bad because there's porn on it.
Compression is a style. There is far more to musicality and emotion than compression. The problem compression solves is that the environments where industrialized cultures now listen in are not dedicated listening areas, but alternately loud and quiet places, so compression makes all parts of the music almost equally loud so there are no drop outs where the quieter parts would be. There is no need to compress music in headphones, for example, to the extent that it is currently compressed.
I find compression and other techniques such as removing vocal breath sounds, makes most recordings unlistenable. They don't sound like humans anymore, but synthetic puppets animated by humans with conflicting values. Take the Foo Fighters, for example. They're popular, sure, but all of their songs sounds like one continuous din. Between the compression induced by the guitar distortion settings and the compression added to the recording, then the compression added by the radio station, it just sounds like a waterfall with a few bandpass filters changing between the verse and chorus.
Also their vocals have no dynamics. When he yells loud, the vocals don't get louder but the timbre changes. That changes it from cathartic to strained. The dynamics have all been flattened.
Why do you think the indie rock movement and bands and styles with wide dynamic range like the Pixies, Nirvana and dubstep got so popular? They eschewed the trend of hardline compression with alternating loud and quiet parts. They match the rhythm of human thought and motion which has fast and slow, detail and empty parts.
> That's like saying the internet is bad because there's porn on it.
Yes but on the internet you can go where there is no porn. Where can you find music with no compression?
That is your artistic choice, as it should be.
"The MP3 only has 5 percent of the data present in the original recording. … The convenience of the digital age has forced people to choose between quality and convenience, but they shouldn’t have to make that choice." -- Neil Young. [0]
Not every artist wants this to happen. They have no choice and listeners get a fraction of the sound recorded. This was not the case with vinyl.
@mborch, the exact compression method is of less importance than recognising that for all the compression being discussed, is a retrograde step from vinyl. Why?
[0] http://allthingsd.com/20120131/neil-young-and-the-sound-of-m...
The source song, PSY's Gagnam Style, is the epitome of modern pop. I got a 3/10 on the listening test on a decent pair of Sennheiser headphones in a quiet room.
Some people are commenting that modern pop pairs well with 16-bit because of the heavy-handed mastering techniques and that older music thrives under 24-bits. Well, Audio Check offers the same 16 vs 8 test, using a Neil Young track from 1989... I couldn't fool myself into hearing any differences between the source WAVs at all and didn't even attempt to score the 10 soundbites.
Yes, it has much more dynamic range, but it sounds wrong. Compression basically emulates what our ears naturally do when hearing very loud material, so compression gives one the feeling that the music is LOUD.
Yes, dynamic range compression is currently used/abused extensively in pop music productions, but if mixes weren't compressed they would have a much wider range for the sounds to play around in.
Not only is Justin Bieber's My World 2.0 louder than
Metallica's The Black Album, it's louder than The Sex
Pistols' Never Mind The Bollocks.
But yeah, it's obvious when I have my car stereo nearly on max to listen to classical or jazz, and then if I turn the radio and get a pop music station my ears are about to explode.
The good thing about compression is that it allows you to save your hearing quite a bit. Some music have dramatic parts that get super loud, which can have awesome emotional response; but it does take a toll on your hearing, unless you are in a very silent environment or have fantastic headphone insulation (I have none) -- so compression actually allows me to hear everything the music has to offer. I also use compression tool on my soundcard to play most games, specially FPSes that have incredibly loud bangs and yet you need to hear footsteps and quiet environmental noises -- with a compressor that's possible without blowing up your ears.
This does degrade the quality, sometimes heavily. Nevertheless it is done by mastering engineers (they rarely enjoy it) as well as by radio and tv stations extensively because of the psycho-acoustic fact that a songs appears to be better if it is played louder. This gives them an advantage over the competition: On average, people searching for a radio station are more likely to listen to your radio station if it is louder than the competition.
The main issue lies in the fact that the current peak measurement of audio signals does only marginally correlate with the perceived loudness and heavy compression is used to trick this system. The broadcasting industry is aware of this. An open and quite effective loudness measurement algorithm [0] has been introduced a few years ago and it gets slowly adapted all over the world by new broadcasting laws: AGCOM 219/09/CSP (Italy), ARIB TR-B32 (Japan), ATSC A/85 PRSS CALM Act (US), EBU R128 (Europe) and OP-59 (Australia). iTunes Soundcheck is also based on [0] and since this year Youtube applies this to newly uploaded videos as well [1]. Even games use [0] to keep their audio at a consistent loudness.
So slowly, the over-usage of compression does not give music producers and broadcasters any advantage anymore and beautiful dynamic music will be competitive again.
I have collected some links [2] about this topic. Because of the lack of any affordable implementation at the time I created one myself [3] with some additional notes [4].
[0] ITU-R BS.1770, http://www.itu.int/dms_pubrec/itu-r/rec/bs/R-REC-BS.1770-4-2... [1] http://productionadvice.co.uk/youtube-loudness/ [2] https://www.klangfreund.com/lufsmeter/manual/#about_loudness [3] https://github.com/klangfreund/LUFSMeter [4] https://github.com/klangfreund/LUFSMeter/tree/master/docs/de...
Edit: Some useful educational material to read before moderating: https://en.wikipedia.org/wiki/Loudness_war .
Maybe if you really mangled your audio by encoding at extremely low bit rates.
But in general, no.
I have a 96khz/24bit interface that I use and ATH-M30X headphones, and I can tell a difference between at least some 24bit FLAC files and 16bit highest-quality-possible MP3s. I was mixing my own music and the difference was quite obvious to me. The notable thing was that drum cymbals seemed to have a bit less sizzle and such.
Now that being said, if I hadn't heard the song a million times in it's lossless form from trying to mix it, I probably wouldn't have noticed, and even then it didn't actually affect my "experience".
I'm one of those guys that downloads vinyl rips as well, but I do that mostly just to experience the alternative mastering, not that I think it's higher quality or anything. (though I have heard a terrible loudness-war CD master that sounded great on vinyl with a different master)
They're pointless for playback.
That is really the central issue. It's much like imaging since the time of Ansel Adams: the sensor can capture more dynamic range than the human eye can experience. The producer may have use for that range when editing, but the audience will never know what was -- may have been -- missed. And we're not talking about limits of reproduction. We're talking about the human sensors both instantaneous and absolute upper and lower bounds.
If I airplay a song from my iPhone and have the volume at 50% set in software, then a few extra bits can help. Not sure if it makes a noticable difference, but it's a digital mixing scenario occurring at playback. If you play at extremely low volume it should be noticable.
The basic problem: the quieter a sound or detail gets, the fewer bits of resolution are used to represent it.
In 16-bit recording, there simply aren't enough bits to represent very low level details without distorting them with a subtle but audible crunchy digital halo of quantisation noise.
In a 24-bit recording, there are.
Talking about dynamic range completely misses the point. It's the not the absolute difference between the loudest and quietest sounds that matters - it's the accuracy with which the quieter sounds are reproduced.
This is because in a studio, 0dB full-scale meter redline is calibrated to a standard voltage reference, and both consumer and professional audio has equivalent standard levels for the loudest level possible.
These levels don't change for different bit depths, and they're used on both analog and digital equipment. (In fact they've been standard for decades now.)
This is why using more bits does not mean you can "reproduce music with a bigger dynamic range" - not without turning the volume up, anyway.
What actually happens is that the maximum possible volume of a playback system stays the same, but quieter sounds are reproduced with more or less accuracy.
In a 16-bit recording quiet sounds below around 50Db have 1-8 bits of effective resolution, which is nowhere near enough for truly accurate reproduction. (Try listening to an 8-bit recording to hear what this means.)
You might think it doesn't matter because they're quiet. Not so. 50dB is a long way from being inaudible, ears can be incredibly good at spectral estimation, and your brain parses spectral content and volume as separate things.
There's a wide range between "loud enough to hear" and "too loud" and 24-bit covers that whole range accurately. 16-bit is fine for louder sounds, but the quieter details just above "loud enough to get hear" get audibly bit-crushed.
The effect isn't glaringly disturbing, and adding dither helps make it even less obvious. But it's still there.
24-bit doesn't need tricks like dither - because it does the job properly in the first place.
Now - whether or not commercial recordings have enough musical detail to take full advantage of 24-bits is a different question. For various reasons - compression, mastering, cheapness - many don't.
But if you have any kind of aural sensitivity, you really should be able to A/B the difference between a 24-bit uncompressed orchestral recording and a 16-bit recording using an otherwise identical studio-grade mixer/mike/recorder/speaker system without too much difficulty.
Besides, mp3 [audio] compressions have difficulty in handling specific samples, or type of samples (eg. sharp attacks), and they may manifest artifacts independently of the bitrate; MP3, AFAIK, also has a ceiling of 320 kbps within the standard specification, which certainly doesn't help.
Secondly, I'm not sure if you process further the MP3s (when you refer to mixing), but if you do, you're definitely going to make noticeable, artifacts which weren't so in the unprocessed MP3 form.
It's possible you are just hearing the difference between codecs. You'd have a fairer comparison with 24-bit vs 16-bit FLAC.
Even 128Kbps MP3s render cymbals better.
Yes, non-linear effects can be sample rate sensitive. However-- this really means that their internal model is aliasing and not faithfully simulating an infinite sample rate system.
In an ideal world, effect that needed more sample rate would internally upsample/downsample (or be constructed in a way that they didn't need to). Then they would behave consistently across rates; though doing this would waste cpu cycles.
In any case, the article is all about distribution. Having excess rate in mastering is cheap and harmless, and-- because of these reasons, can be practically pretty useful.
The difference you hear is the difference between flac's lossless format and mp3's lossy format it has nothing to do with 16 bit versus 24 bit.
I was listening to Marvin Gaye on my friends system and I could hear that there were several different backing singers all moving and at different distances from the microphone.
Are there any double blind trials anywhere of Vinyl/CD/24-192khz with super high end hifi systems? Mostly I see people suggesting that these tests are performed from the phono output of a mac with a pair of average ear buds...
You were listening to £20,000 worth of amps and speakers, and you were most likely in an acoustically treated room.
Also, novelty is almost always euphonic when it isn't overtly bad. This fact is often neglected. You hear something you didn't hear before and your brain immediately tells you that it sounds better, even if it doesn't actually represent higher fidelity. Actually making an objective judgement requires a careersworth of experience, or a test lab and the skills to use it.
For example: you were listening to vinyl, which is covered in delicious noise and warm harmonic distortion, and is mastered differently. Highly euphonic, very novel if youve only ever heard the CD version before, but definitely not higher fidelity.
BTW higher end DACs do sound better, but the rest of your signal chain needs to be really good for you to notice it. It's often to do with better phase accuracy between the left and right channels, which affects the soundstage, or stereo image. If your speakers/amp have loose timing however, you'll never be able to tell.
This hasn't passed the blind tests either. A good, 100 dollar dac (a schiit or an odac) will sound just as good as a 1000 dollar dac.
This fact alone should cause you to question your subjective experience. You have no idea what part of that system was contributing to what you found pleasant. Someone who knew what they were doing could probably build a $2000 system that would blow you away just the same.
And if you were playing vinyl, there wasn't even a DAC present in the signal chain :)
Vinyl mastering is sometimes better than CD mastering though, due to the loudness war.
I would love to sell my turntable and vinyl collection and rely purely on digital formats. Takes up less space, technically superior format, etc.
But one thing keeps me buying vinyl:
AWFUL mastering on CDs. A significant portion of LPs are released with more normal mastering on the vinyl, while the CD will be brickwalled all to hell.
I listen to metal, and rock as a broader genre is particularly bad about it. One of my favorite albums of last year, Fallujah's The Flesh Prevails, had a dynamic range of 2 to 3 on almost every track on the CD. The vinyl master? 9 to 10. Still not great, but leaps and bounds better. The CD actually clips if you convert the songs into MP3.
Until they go back to not murdering CD mastering, I'll continue buying vinyl :(
(I know your comment isn't directly about vinyl being bad or anything - I just have a compulsion to bitch about the loudness war any chance I can)
Not on a laser stylus turntable.
I don't know about double blind trials but people do tests on their own. It's further complicated though because the hardware you use could be optimized for certain types of music, e.g. have a read through http://arstechnica.com/gadgets/2014/07/some-of-the-worlds-mo...
The article mentions just such a study performed with high end equipment.
That said, there's one thing the article does not address and that is "beating", or really inter-modulation distortion from instrumental overtones.
Instruments are not limited to 20-20kHz. They can have overtones well above this range. Additionally, note that short pulse-width signals, i.e. transients, like drum strikes, especially involving wooden percussion, can have infinite bandwidth. (Not really infinite, but pulse-width is inversely proportional to bandwidth.
In a real listening environment (i.e. live performance) these overtones have a chance to interact with one another in the air. It is possible that these overtones may beat with one another and cause inter-modulation products in the audible range. For an example of this, play a 1000 Hz tone through your left speaker, and a 1001 Hz through your right speaker. You will hear a distinct 1 Hz "beat". The audibility of these are largely dependent on listening position and amplitude, but it is possible to occur with instruments. Since most recordings are done using a "close mic" technique (placing the microphone very close to the source) the interactions such as this are never recorded.
However, if full bandwidth of the producing instruments is preserved, these interactions of the overtones can be reproduced in a playback environment given equipment having a wide enough bandwidth and degree of quality.
The comparison of a 1hz beat to a 1hz sound should be absurd on its face: you need about 20-30hz to become audible, and it's a low rumble more felt than heard. Very low frequencies sound absolutely nothing like intermodulation beats.
Second, as far as I know our hearing is composed of linear excitation elements (they have a definite bandwidth), and this is confirmed pretty well by experiments with human hearing -- you can see the threshold of our hearing at about 20kHz and that we experience tones of different frequencies fairly independently. Those assumptions imply that two tones, one at e.g. 50kHz and another at 50.001kHz are inaudible, end of story.
You can actually do this experiment yourself if you have a signal generator that can do 1Hz amplitude modulation and drive a transducer with a non-negligible sensitivity in that range.
[1] AFAIK most music is not recorded like that, instruments are recorded separately and then overlaid; but then adding realistic-sounding "beats" based on whatever positioning the sound engineer envisions should be possible in software?
Beating and intermodulation distortion are entirely different things. They look similar on an oscilloscope, but they're not and they don't sound the same.
>Instruments are not limited to 20-20kHz. They can have overtones well above this range.
Correct. You can't hear the overtones beyond the upper portion of the hearing range (many people believe you can).
>In a real listening environment (i.e. live performance) these overtones have a chance to interact with one another in the air.
In reality they do not unless you're driving the air so hard the trough rarification is approaching hard vacuum. (That's not actually impossible. It's how ultrasonic audio 'beaming' devices work). Some performances are powerful enough to get close, eg, if you're sitting six feet from the pipe organ.
Once you're driving air so hard it becomes nonlinear, thus introducing intermodulation distortion in the air, that distortion produces actual audible-range distortion products. And because the distortion you're hearing is in the audible range, a recording will sample and reproduce it accurately.
You're hearing the audible _result_ of IMD, you're not somehow listening to the distortion curve itself.
> It is possible that these overtones may beat with one another
You're continuing to confuse beats and IMD, but here you're talking about beat frequencies, so Yes. But beat frequencies are a sort of auditory illusion. If one of the frequencies that would produce a beat is inaudible--- there's no beat. Easy to test, go try it.
> and cause inter-modulation products in the audible range.
IMD is not a beat. Inaudible ultrasonics will produce audible artifacts when the underlying reproduction system is nonlinear (another way of saying 'there's intermodulation distortion'). However, that's a playback artifact. If the IMD products were audible in the original signal, audible range sampling would reproduce them.
If it wasn't audible in the original performance, it should not be part of the recording, and it should not be part of the playback.
24bit also means we don't have to record at 0dBFS, which saves a lot of time.
If you have the time, watch the two videos that xiph.org did a few years ago[0]. There's a great in-depth explanation, as well as a hands on demonstration to demonstrate this reality.
Those who don't have a oscilloscope can see the picture here: http://i.imgur.com/wY0wzcW.png
See the digital media primer 2 for more information on that: https://wiki.xiph.org/Videos/Digital_Show_and_Tell
If humans were able to hear audio above 22kHz (or what not) in any meaningful way, we'd expect to be be able to demonstrate that effect in carefully controlled studied and then that lack of low-passing may matter; but that isn't what the best evidence so far shows.
The amount of misinformation / junk-science in the audio world is preposterous. There's a religious-cult of an industry that feeds off the ignorance and placebos of its participants. I have many friends who swear by their What.cd 24/192 FLAC vinyl rips and spend hundreds of dollars on audiophile AC wall outlets. Not to say that there are no differences in high-end audio equipment, but so much of what's "good" is subjective.
Unfortunately, there is no point to distributing music in 24-bit/192kHz format. Its playback fidelity is slightly inferior to 16/44.1 or 16/48, and it takes up 6 times the space.
This has all been known to anyone with actual signal processing and/or audio engineering knowledge for a long time now. As in, common knowledge to the kinds of folks attending the AES conference at least back to ~2001 or so. The high sample rate/bit depth stuff is useful for production process, but irrelevant for final distribution.
Edit: christ, I mixed up bitrates (e.g. 192kbps) with sampling frequency (e.g. 192kHz) again. I was referring to 64kbps streams.
Edit: apparently my memory is worse than I thought.
In order to decimate a signal to 44.1 or 48khz, and preserve high-frequency content, high frequencies need to be phase-shifted.
This phase-shift is similar to how lossy codecs work.
For what it's worth: I'm a big fan of music in surround, and most of it comes in high sampling rates. When I investigated ripping my DVD-As and Blurays, I found that they never have music over 20khz. It's all filtered out. However, downsampling to 44.1 or 48khz isn't "lossless" because of the phase shift needed due to the Nyquist-Shannon theory.
I still rip my DVD-As at 48khz, though. There isn't a good lossless codec that can preserve phase at high frequencies, yet approach the bitrate of 12/48 flac.
Your understanding of sampling theorem is incorrect. Sampling alone (not quantization, of course) is completely lossless under the critical frequency.
We demonstrated this in a very clear way near the end, at about 21 minutes in, on the primer two video: http://www.xiph.org/video/vid2.shtml where we show a square wave being phase shifted tiny fractions of the intersample length.
> In order to decimate a signal to 44.1 or 48khz, and preserve high-frequency content, high frequencies need to be phase-shifted.
What do you mean by high frequency? If you mean frequencies below but near the Nyquist frequency then no, there is no phase shift. If you mean at or above...
I'm struggling to avoid a blatant appeal to authority here, but your position is that the author of the Ogg Vorbis coded doesn't understand digital sampling, which seems challenging to believe.
> So the math is ideal, but what of real world complications? The most notorious is the band-limiting requirement. Signals with content over the Nyquist frequency must be lowpassed before sampling to avoid aliasing distortion; this analog lowpass is the infamous antialiasing filter. Antialiasing can't be ideal in practice, but modern techniques bring it very close. ...and with that we come to oversampling.
if you accept that the limit of hearing is around 20 kHz, then you must also accept that frequencies above that can freely be removed without loss of fidelity to the human ear.
the article notes that higher frequencies can be heard, but only in the form of ultrasonic intermodulation distortion. (i.e. not in fact the higher frequencies at all)
Because digital filters have few of the practical
limitations of an analog filter, we can complete the
anti-aliasing process with greater efficiency and
precision digitally. The very high rate raw digital
signal passes through a digital anti-aliasing filter,
which has no trouble fitting a transition band into a
tight space.
My understanding: If you have a an analog filter of a given steepness the only way to further reduce aliasing effects digitally is oversampling. Or less steep (cheaper) analog filter plus oversampling is the same as steeper (more) expensive analog filter. People tend to say digital anti-aliasing filters when they really mean oversampling.
"24/192 music downloads make no sense" seems to be a thoroughly researched and carefully written article. It explains oversampling very well, possible confusion with digital filtering (anti-aliasing or not) is out of question. But then it goes on to talk about digital anti-aliasing filters, which makes me afraid I could be wrong.
Do digital anti-aliasing filters exist?
> My understanding: If you have a an analog filter of a given steepness the only way to further reduce aliasing effects digitally is oversampling. Or less steep (cheaper) analog filter plus oversampling is the same as steeper (more) expensive analog filter. People tend to say digital anti-aliasing filters when they really mean oversampling
You're right, and it's actually both. The ADC can run at a much higher sample rate with a cheaper analog filter, and then that digital signal is again passed through a digital filter and downsampled.
If more people prefer the sound at the higher bitrate and sampling rate, then that's the better format, even if there's no technical reason why that format is superior.
Much like how some people prefer the "warm" sound of tube amps, even if that means more distortion.
>Empirical evidence from listening tests backs up the assertion that 44.1kHz/16 bit provides highest-possible fidelity playback.
You can read the article if you want to find the actual references. No one is arguing that higher rates/bits produces any sort of distortion that anyone would prefer.
The difference from my perspective is that an amp is a tool for sound production while a digital music format is a tool for sound reproduction. When producing sound, choosing more distortion over less distortion is a valid choice. When reproducing sound, the goal should be accurate reproduction of the original.
I've never seen the hype from artists about 24/192 as being about better listening experience. It's about handing their consumers a better master so as to encourage and enable more of them to be remixers.
Or generate a tone sweep in audacity. Generate->chirp http://www.audacityteam.org/
You loose the ability to hear high frequency sounds as you age.
Personally I can hear up to about 14kHZ
That's a pretty cool feature for Ozone 7, for sure! I'm still using Ozone 5 and don't feel a need to upgrade, but that might make it...
Previous discussion https://news.ycombinator.com/item?id=3668310
People today are often amazed when they listen to CD or turntable content through 70's era crossover speakers. Back in the 70's you'd have a stereo with 2 "speakers" that each had 3 subspeakers for a total of six speakers. The fad today is to have 5.1 sound with a single driver in each satellite, also a total of six speakers. The spatial resolution increase is good for movies, games and TV but surround sound in music is marginal. An amazing number of old "classic rock" recordings were done in quad and anything by Donald Fagan will sound pretty good w/ Dolby Pro Logic, there are some more recent Bjork recordings, but almost everything is mixed for stereo and what you loose in frequency response is not compensated by anything, except perhaps the ability to produce more volume with more speakers.
It made sense to me, and I love how the speakers sound. Understanding is not inserting distortion makes even more sense.