Edoardo
500+ Head-Fier
- Joined
- Feb 10, 2009
- Posts
- 856
- Likes
- 23
(EDIT) DISCLAIMER: this is just a listening guide, not a programming, nor an engineering nor a harmony lesson. I tried to keep the "scientific" part as brief as possible, mainly avoiding it, with the purpose of explaining everything in very poor words, just to make the reasons of some hopefully useful tips to be easily understandable to everyone.
I've always been kind of surprised by the number of people claiming that there is no audible difference between playing Mp3 files or CDs.
Many of them stated that you would have to use the "lastest" Mp3 algorithms or the like.
So I thought... Maybe it's just me, my Mp3 copies sound crappy because I can't find a software with the latest algorithms...
Well, yesterday I attended a test that encouraged me to write - finally - this.
I'd like to clear out why some people can say that there is no remarkable difference between compressed and uncompressed music, why they are also right, and also why HD music is still such a niche.
Let's start.
1. Where and how compression takes place
It's all about loudness, or else, about compressing the size of the file by compressing the dynamic range of the audio tracks.
Any algorithm "selects" the highest-volumed signals, and cuts off the lower ones.
So you have saved space on your hard disk, but you have a copy that contains only the volume peaks of the original, and that is dynamically much flatter than the original one.
2. What's the dynamic range
In poor words, we define dynamic range as the difference between the lowest-volumed and the highest-volumed audible signals recorded on a track.
Compressed music (and here I could mean "compressed" in both senses) has a dynamic range of less than 5 dBs, against the tens of dBs of dynamics possible to record and playback.
3. Why this?
Money. That's half a joke.
Most people play music while commuting, or on their cars, which are noisy environments.
Also, music is played as background in bars, pubs, clubs, supermarkets and the like.
So... Nobody wants to stay at the volume knob all the time, or to risk a crash when the chorus or the main theme come.
These are the same reasons behind the so-called loudness war (a.k.a. loudness race).
For those who haven't heard about it yet, we call "loudness war" the phenomenon for which the major labels have been producing recordings and masters with increasingly lower dynamic range, for those same reasons.
But... These are also the reasons behind the success of the compressed files.
(Of course, one more reason was the 10-year-ago internet bandwith. Try to up and download a 600MB CD at 56k. Lighter music files were needed for online music stores to succeed. But this does not concern this thread, so let's just focus on the dynamic range thing.)
4.0 Why is having a high dynamic range so important?
Usually, when we think about the benefits of a high dynamic range, we think mostly about the recording of factors other than the mere instruments' sound, such as the reproduction of the soundstage or of environmental elements.
Of course, you cannot record nor reproduce wall reflections nor a fly buzz and something else playing out loud at the same time, without a dynamic range of the needed width.
But it's not just about that. I'm going to explain why having a pretty wide dynamic range is essential to the riproduction of the single musical instruments and to the listening of a composition as it was intended by its author.
4.1 Recording and playback'ing instruments
Every note produced by a musical instrument is composed of various, and theoretically infinite, harmonics.
Again in very poor words, the harmonics are the elements in which a sound can be divided, and of which a sound is composed.
When we play an "A" - which has a fundamental frequency of 440Hz - on a pitched musical instrument, we do not produce just a 440Hz sound wave, but a more or less complex sound composed also by a series of sound waves at 880 Hz, 1320Hz, 1760Hz, etc.
The distributions, decreases and volumes of the harmonics in playing a note, give every instrument or voice their peculiar timbres, or colors.
Let me borrow a pic from the internet, showing the frequency range of the fundamentals and of the harmonics which some musical some musical instruments are able to produce.
Don't be misleaded by the way the graph is drawn down here, though.
Any musical instrument produces harmonics at any height. What's colored lighter here are just those frequencies that the instruments do not reach as a fundamental frequency, but only through their harmonics.
Of course, harmonics play at increasingly lower volumes than their fundamentals. But yet, they are there.
And it's them to give each instrument its peculiar sound, or timbre, or color if you prefer.
So... In conclusion, a certain dynamic range is required to playback any musical instrument decently, since a pretty wide dynamic range is required to get not only environmental sounds, but also the very instruments' sound!
4.2 Delivering the musical message
In commercial music, usually, even if you have more than one instrument playing at the same time, they all happen to be mixed at the same (relatively high) volume, so the problem within the fidelity of the recording falls in the previous analysis.
But once upon a time, before music identified itself in the discographic industry and before loudness became a must, musicians also exploited volume difference between the instruments, or the playing volume of the whole band or orchestra, as a living part of the composition itself, as instrument, key or tonality choices are.
Of course, this exploitation is far more evident in orchestral music than in any other one.
Now, I really don't wanna beat dead horses like "which genre is the better", so please just take this statement as neutrally as I hope you will.
But that's a fact. The instruments' playing volume is clearly indicated by the good old great composers on their music sheets.
Now it would be half a commercial suicide for any rock band to do that. Instruments just play as loud as the others or don't play at all.
Musicians have their hands tied on that part.
Therefore, I'll conclude this part stating that a wide dynamic range is required not only to listen to some great compositions as they were meant to, by their own authors (Else, a fundamental part of the very composition is missing), but also to let the musician using one more tool to craft its work.
5. Where and what to look for
Given this introduction, I wouldn't be surprised if I didn't find any difference between an Mp3 compression of some commercial artist, poorly and/or loudly recorded.
If the dynamics achievable by the CD standard are not exploited at all at the source, its Mp3 copy becomes not a lossy, but just a space-saving one.
So it's easier to spot the differences, if any, with unplugged/acoustic instruments, and within bands and orchestras.
Of course, well-recorded and without "loud" compression.
So... Testing anything, actually, it's just about listening to what's happening at the lower volumes.
What's there? Instruments playing in the background, their harmonics, environment reflections, the sound of those strings picked one or two seconds before that keep vibrating...
Mind that higher frequencies give our brains the direction from which the sound is coming, and therefore, the smaller or bigger richness of the harmonics can be "translated", by our brains or put into words as worse or better soundstage also.
But... Before testing, and before making any judgement on the quality of the various formats, mind the quality of the recording and the one of the instruments recorded there.
Spotting difference between various formats may be impossible within a dynamically poor track.
Hope I could help, feedbacks are welcome.
I've always been kind of surprised by the number of people claiming that there is no audible difference between playing Mp3 files or CDs.
Many of them stated that you would have to use the "lastest" Mp3 algorithms or the like.
So I thought... Maybe it's just me, my Mp3 copies sound crappy because I can't find a software with the latest algorithms...
Well, yesterday I attended a test that encouraged me to write - finally - this.
I'd like to clear out why some people can say that there is no remarkable difference between compressed and uncompressed music, why they are also right, and also why HD music is still such a niche.
Let's start.
1. Where and how compression takes place
It's all about loudness, or else, about compressing the size of the file by compressing the dynamic range of the audio tracks.
Any algorithm "selects" the highest-volumed signals, and cuts off the lower ones.
So you have saved space on your hard disk, but you have a copy that contains only the volume peaks of the original, and that is dynamically much flatter than the original one.
2. What's the dynamic range
In poor words, we define dynamic range as the difference between the lowest-volumed and the highest-volumed audible signals recorded on a track.
Compressed music (and here I could mean "compressed" in both senses) has a dynamic range of less than 5 dBs, against the tens of dBs of dynamics possible to record and playback.
3. Why this?
Money. That's half a joke.
Most people play music while commuting, or on their cars, which are noisy environments.
Also, music is played as background in bars, pubs, clubs, supermarkets and the like.
So... Nobody wants to stay at the volume knob all the time, or to risk a crash when the chorus or the main theme come.
These are the same reasons behind the so-called loudness war (a.k.a. loudness race).
For those who haven't heard about it yet, we call "loudness war" the phenomenon for which the major labels have been producing recordings and masters with increasingly lower dynamic range, for those same reasons.
But... These are also the reasons behind the success of the compressed files.
(Of course, one more reason was the 10-year-ago internet bandwith. Try to up and download a 600MB CD at 56k. Lighter music files were needed for online music stores to succeed. But this does not concern this thread, so let's just focus on the dynamic range thing.)
4.0 Why is having a high dynamic range so important?
Usually, when we think about the benefits of a high dynamic range, we think mostly about the recording of factors other than the mere instruments' sound, such as the reproduction of the soundstage or of environmental elements.
Of course, you cannot record nor reproduce wall reflections nor a fly buzz and something else playing out loud at the same time, without a dynamic range of the needed width.
But it's not just about that. I'm going to explain why having a pretty wide dynamic range is essential to the riproduction of the single musical instruments and to the listening of a composition as it was intended by its author.
4.1 Recording and playback'ing instruments
Every note produced by a musical instrument is composed of various, and theoretically infinite, harmonics.
Again in very poor words, the harmonics are the elements in which a sound can be divided, and of which a sound is composed.
When we play an "A" - which has a fundamental frequency of 440Hz - on a pitched musical instrument, we do not produce just a 440Hz sound wave, but a more or less complex sound composed also by a series of sound waves at 880 Hz, 1320Hz, 1760Hz, etc.
The distributions, decreases and volumes of the harmonics in playing a note, give every instrument or voice their peculiar timbres, or colors.
Let me borrow a pic from the internet, showing the frequency range of the fundamentals and of the harmonics which some musical some musical instruments are able to produce.
Don't be misleaded by the way the graph is drawn down here, though.
Any musical instrument produces harmonics at any height. What's colored lighter here are just those frequencies that the instruments do not reach as a fundamental frequency, but only through their harmonics.
Of course, harmonics play at increasingly lower volumes than their fundamentals. But yet, they are there.
And it's them to give each instrument its peculiar sound, or timbre, or color if you prefer.
So... In conclusion, a certain dynamic range is required to playback any musical instrument decently, since a pretty wide dynamic range is required to get not only environmental sounds, but also the very instruments' sound!
4.2 Delivering the musical message
In commercial music, usually, even if you have more than one instrument playing at the same time, they all happen to be mixed at the same (relatively high) volume, so the problem within the fidelity of the recording falls in the previous analysis.
But once upon a time, before music identified itself in the discographic industry and before loudness became a must, musicians also exploited volume difference between the instruments, or the playing volume of the whole band or orchestra, as a living part of the composition itself, as instrument, key or tonality choices are.
Of course, this exploitation is far more evident in orchestral music than in any other one.
Now, I really don't wanna beat dead horses like "which genre is the better", so please just take this statement as neutrally as I hope you will.
But that's a fact. The instruments' playing volume is clearly indicated by the good old great composers on their music sheets.
Now it would be half a commercial suicide for any rock band to do that. Instruments just play as loud as the others or don't play at all.
Musicians have their hands tied on that part.
Therefore, I'll conclude this part stating that a wide dynamic range is required not only to listen to some great compositions as they were meant to, by their own authors (Else, a fundamental part of the very composition is missing), but also to let the musician using one more tool to craft its work.
5. Where and what to look for
Given this introduction, I wouldn't be surprised if I didn't find any difference between an Mp3 compression of some commercial artist, poorly and/or loudly recorded.
If the dynamics achievable by the CD standard are not exploited at all at the source, its Mp3 copy becomes not a lossy, but just a space-saving one.
So it's easier to spot the differences, if any, with unplugged/acoustic instruments, and within bands and orchestras.
Of course, well-recorded and without "loud" compression.
So... Testing anything, actually, it's just about listening to what's happening at the lower volumes.
What's there? Instruments playing in the background, their harmonics, environment reflections, the sound of those strings picked one or two seconds before that keep vibrating...
Mind that higher frequencies give our brains the direction from which the sound is coming, and therefore, the smaller or bigger richness of the harmonics can be "translated", by our brains or put into words as worse or better soundstage also.
But... Before testing, and before making any judgement on the quality of the various formats, mind the quality of the recording and the one of the instruments recorded there.
Spotting difference between various formats may be impossible within a dynamically poor track.
Hope I could help, feedbacks are welcome.
