| |
 |
CD Recording FAQ
7. AudioCD
7.4. Writing MP3 files on CD
The most obvious option for mp3 audio files (or audio files in other formats) is to store them as data files. In this way, you can store many hours of music on a single cd.
Also, many modern cd players can play back data cds with mp3 files, so this option is certainly appealing if you own such a device.
If you wish to write an audio cd from mp3 files, you must first convert them into wav files. This can be done with many programs, but also many cd mastering applications allow you to import mp3 files and the software will convert them into proper wav files on the fly during the recording process. If you are not interested in accuracy issues, you can stop reading here.
Keep in mind that there is no substitute for the original mp3 files. To understand this, let's take a look at the way audio compression works:
Audio is written on a cd in uncompressed format, which means that the density of samples is homogenous. These samples often include audio information whose abscence can be forgiven by human perception, especially if cheap speakers are used.
This simple asumption, of throwing away something that is not considered really needed, forms the basis of audio compression, in various formats, including mp3, vqf, aac and real audio.
Of course, the more a sound sample is compressed, the more that is missed from the sample we started from, so a compromise has to be made. For example, digital telephony uses very low bitrate codecs, thus resulting in the bluntness of all voices when heard via a mobile telephone. On the other hand, human perception can be deceived, which means that the human brain can throw in some missing bits when hearing a compressed sample.
So the result of audio compression depends on two factors, the method employed and the level of compression, and mp3, as a compression method, gives results according to this rule.
Higher bitrates give almost perfect results, depending on the equipment used to reproduce sound. Lower bitrates, on the other hand, leave much to be missed, but can be usefull when the critical factor is the size of the resulting file (or the bandwidth available when the file has to be transmitted in real time).
This means that mp3 encoding will cut a lot of information of the original sound file, and this cannot be changed. Maybe the difference cannot be heard on cheap speakers/sound cards, but it exists no matter what.
The most common bitrate for mp3 is 64Kbps per channel, which means 128Kbps for stereo sound. When using a decent encoder, this can be enough for the vast majority of people, depending on the kind of music and the speakers used of course.
The mp3 algorithm is unique, but various implementations can give varying results, because of doing a better job in deceiving the human brain. The other end of the equation is the reverse process, decoding, where a similar situation is met: Depending on the decoding algorithm, the result can be impressive or disappointing.
The trick has mostly to do with equalization, but this is not the point. The point is that a specific decoding algorithm may seem better, but this is because it changes the sound to seem more vibrant and accurate, and this is done by altering the audio information which is supposed to be contained in the original mp3 file.
However, the original mp3 file decoded by such a decoder is severely changed. Though the sound seems better, something has changed, and this is extremely important if you have reason to want the best accuracy.
What makes things worst is the fact that the process of decoding cannot be reverted, so audiophiles should copy and transfer the original mp3 files instead of decoding them and burning them on an audio cd.
The bottom line is that converting mp3 files
to audio cd can be a risky process when the best level of accuracy is
needed, and the final audio cd is not a replacement for the original mp3
files.