This episode brings us a target-rich article written by Daniel Kumin for January 2011 issue of Sound and Vision Magazine titled “How to Rip Your Music (Part 1)”.
The article starts fine, talking about the need for metadata (album art, etc.) in addition to the music data itself. But unfortunately, the quality of the writing and information goes down from there on.
Starting with nitpicking, a codec is short for compressor/decompressor or coder/decoder. I have not heard the encoding end being called “code.”
Codecs do not convert music data from one dialect to the other either. They compress the data into a shorter form. The decoder does the reverse and restores the music into a playable form.
Given the above, we don’t talk about codecs when the file is saved in its uncompressed format. No codec is used at all in that instance.
All in all, this is the worst introduction I have seen to audio compression.
As if to prove my third point above, he goes on to say:
So he concedes no codec is in use when ripping the original files as is. But then goes off saying a bunch of obscure stuff that makes little sense. If this is an intro to ripping music, then we are always talking about 16-bit PCM data from CD. The word “not necessarily exclusively” does not apply. If he is referring to ripping DVD-A and SACDs as being the exception, then he should have said so rather than making the user think his CDs may not actually be 16-bit PCM data.
Now I am on a war path with the guy for not mentioning the built-in lossless format in Windows which my group developed, namely, WMA Lossless
. It is true that FLAC has more universal playback ability in the wild but WMA Lossless has a strong following due to it being the only lossless option in Windows Media Player and holds its own as far as popularity with Apple’s.
Continuing on, after showing off that he knows what MP3 stands for, he goes to give us this jewel:
Where to even begin…. AAC is an international standard and it stands for Advanced Audio Coding. It has never had anything to do with Apple. Indeed, neither iTunes, nor iPod existed when AAC was defined. Apple simply licensed it for use in its products.
On the other hand, when MPEG went out looking for a container for hold MPEG-4 audio/video content, Apple bid its QuickTime file format which (in rigged process) got adopted as part of the standard. So if he wants to call anything “Apple,” he could have associated that with MP4, not AAC.
As to MP3 and AAC being the only two that matter, it can be argued that WMA, which my group also developed, and was in existence years before Apple came up with iTunes, has arguably even more content ripped into it than AAC. Either way, it would have merited a mention at least and not make it look like the only thing the author cares about is the Apple world. I told you I was on a war path with him with his anti-Microsoft bias .
And what to make of the comment that MP3 and AAC are interchangeable? Interchangeable in what way? That they are compression formats you could choose? That doesn’t make them interchangeable but rather, the choices you as a user have, depending on the media player/ripping software you use.
Ok, this is getting tiring. Why does he keep adding “MP4” to the front of AAC? Yes, that can be the container for AAC but who cares? You don’t see people referring to WMA as “ASF/WMA” because ASF is the format of the container for WMA audio format. If this is an article for beginners, let’s not feed them terminology which is uncommon.
Second and the real issue, he can’t possibly understand the difference between these two codecs to have made the statement he did. The first letter in AAC stands for “Advanced” for a reason: it is a substantially improved codec compared to MP3. It doesn’t just sound better on “certain kinds of sounds.” That’s like saying if you drive a Porsche a certain way, it handles better than a Ford!
AAC has much improved fidelity especially at 128 Kbps and lower. At higher bit rates, the difference between all codecs shrinks although to a trained listener, that difference is audible all the way to the maximum rates of these codecs. While AAC/WMA/WMA Lossless can approach lossless status at their highest data rates, MP3 never quite gets there.
OK, help me out. When did the standard for writing go so far down in a commercial magazine? “Big-time?” And why does he write this as if it is surprising that a lossy codec would take up less space than a lossless one? There would be a hundred ways to say lossy codecs take up less space than the confusing sentence he wrote.
Huh indeed! If we take the 1.5 megabytes and convert it to bits we get 12 megabits (1.5 * 8). We then divide this by the 64 Kbits/sec of the MP3 file and we get a song length of 187 seconds.
Now let’s do the math for the CD source. We have two channels of audio at 44,100 samples/second with each sample taking two bytes (16 bits each). So for 187 seconds we have 2*2*44100*187 = 32 megabytes. In other words the compression ratio is 32/1.5 = 21:1, not 10:1. And his file size for the CD track is wrong.
There is of course a simpler way to get to above. Uncompressed CD has a rate of 1.4 mbit/sec. So if we divided 1400/64 which is also equal to 21:1 compression.
He goes to this outlandish claim:
Now, really? Let’s start with simple stuff. MP3 encoders at 64 Kbps severely chop off the high frequencies before they even start to trim the samples perceptually. Here is an example:
It is kind of hard to see the scale but the above graph is the original and the bottom the MP3 at 64kbps. Everything past 8 KHz is filtered off. Sampling rate is likely to be 22 KHz ( so bandwidth of 11 KHz) and some extra filtering thrown in there for good measure. Even at 128 Kbps, standard MP3 encoders roll off some of the high frequencies below that of CD. Net, net, even if the user is untrained, they can easily tell that the MP3 at 64 Kbps is muffled due to lack of high frequencies. Pile on compression artifacts on every transient and the statement he is making is quite laughable.
Besides, when did MP3 encoding at 64 Kbps become typical? Even 10 years back the standard was 128 Kbps. To wit, the MP3 encoder in Windows Media Player does not even go below 128 kbps. I don’t know that I have ever run into anyone ripping music at 64 kbps using MP3.
I can’t wait to read the second installment of his article. Let’s hope he reads the primer I wrote on such matters before inking that one: Digital Audio/Video and Communication Rates and Lossless Audio Compression.
The article starts fine, talking about the need for metadata (album art, etc.) in addition to the music data itself. But unfortunately, the quality of the writing and information goes down from there on.
Starting with nitpicking, a codec is short for compressor/decompressor or coder/decoder. I have not heard the encoding end being called “code.”
Codecs do not convert music data from one dialect to the other either. They compress the data into a shorter form. The decoder does the reverse and restores the music into a playable form.
Given the above, we don’t talk about codecs when the file is saved in its uncompressed format. No codec is used at all in that instance.
All in all, this is the worst introduction I have seen to audio compression.
As if to prove my third point above, he goes on to say:
So he concedes no codec is in use when ripping the original files as is. But then goes off saying a bunch of obscure stuff that makes little sense. If this is an intro to ripping music, then we are always talking about 16-bit PCM data from CD. The word “not necessarily exclusively” does not apply. If he is referring to ripping DVD-A and SACDs as being the exception, then he should have said so rather than making the user think his CDs may not actually be 16-bit PCM data.
Now I am on a war path with the guy for not mentioning the built-in lossless format in Windows which my group developed, namely, WMA Lossless
. It is true that FLAC has more universal playback ability in the wild but WMA Lossless has a strong following due to it being the only lossless option in Windows Media Player and holds its own as far as popularity with Apple’s.Continuing on, after showing off that he knows what MP3 stands for, he goes to give us this jewel:
Where to even begin…. AAC is an international standard and it stands for Advanced Audio Coding. It has never had anything to do with Apple. Indeed, neither iTunes, nor iPod existed when AAC was defined. Apple simply licensed it for use in its products.
On the other hand, when MPEG went out looking for a container for hold MPEG-4 audio/video content, Apple bid its QuickTime file format which (in rigged process) got adopted as part of the standard. So if he wants to call anything “Apple,” he could have associated that with MP4, not AAC.
As to MP3 and AAC being the only two that matter, it can be argued that WMA, which my group also developed
, and was in existence years before Apple came up with iTunes, has arguably even more content ripped into it than AAC. Either way, it would have merited a mention at least and not make it look like the only thing the author cares about is the Apple world. I told you I was on a war path with him with his anti-Microsoft bias .And what to make of the comment that MP3 and AAC are interchangeable? Interchangeable in what way? That they are compression formats you could choose? That doesn’t make them interchangeable but rather, the choices you as a user have, depending on the media player/ripping software you use.
Ok, this is getting tiring. Why does he keep adding “MP4” to the front of AAC? Yes, that can be the container for AAC but who cares? You don’t see people referring to WMA as “ASF/WMA” because ASF is the format of the container for WMA audio format. If this is an article for beginners, let’s not feed them terminology which is uncommon.
Second and the real issue, he can’t possibly understand the difference between these two codecs to have made the statement he did. The first letter in AAC stands for “Advanced” for a reason: it is a substantially improved codec compared to MP3. It doesn’t just sound better on “certain kinds of sounds.” That’s like saying if you drive a Porsche a certain way, it handles better than a Ford!
AAC has much improved fidelity especially at 128 Kbps and lower. At higher bit rates, the difference between all codecs shrinks although to a trained listener, that difference is audible all the way to the maximum rates of these codecs. While AAC/WMA/WMA Lossless can approach lossless status at their highest data rates, MP3 never quite gets there.
OK, help me out. When did the standard for writing go so far down in a commercial magazine? “Big-time?” And why does he write this as if it is surprising that a lossy codec would take up less space than a lossless one? There would be a hundred ways to say lossy codecs take up less space than the confusing sentence he wrote.
Huh indeed! If we take the 1.5 megabytes and convert it to bits we get 12 megabits (1.5 * 8). We then divide this by the 64 Kbits/sec of the MP3 file and we get a song length of 187 seconds.
Now let’s do the math for the CD source. We have two channels of audio at 44,100 samples/second with each sample taking two bytes (16 bits each). So for 187 seconds we have 2*2*44100*187 = 32 megabytes. In other words the compression ratio is 32/1.5 = 21:1, not 10:1. And his file size for the CD track is wrong.
There is of course a simpler way to get to above. Uncompressed CD has a rate of 1.4 mbit/sec. So if we divided 1400/64 which is also equal to 21:1 compression.
He goes to this outlandish claim:
Now, really? Let’s start with simple stuff. MP3 encoders at 64 Kbps severely chop off the high frequencies before they even start to trim the samples perceptually. Here is an example:
It is kind of hard to see the scale but the above graph is the original and the bottom the MP3 at 64kbps. Everything past 8 KHz is filtered off. Sampling rate is likely to be 22 KHz ( so bandwidth of 11 KHz) and some extra filtering thrown in there for good measure. Even at 128 Kbps, standard MP3 encoders roll off some of the high frequencies below that of CD. Net, net, even if the user is untrained, they can easily tell that the MP3 at 64 Kbps is muffled due to lack of high frequencies. Pile on compression artifacts on every transient and the statement he is making is quite laughable.
Besides, when did MP3 encoding at 64 Kbps become typical? Even 10 years back the standard was 128 Kbps. To wit, the MP3 encoder in Windows Media Player does not even go below 128 kbps. I don’t know that I have ever run into anyone ripping music at 64 kbps using MP3.
I can’t wait to read the second installment of his article. Let’s hope he reads the primer I wrote on such matters before inking that one: Digital Audio/Video and Communication Rates and Lossless Audio Compression
.
