Audio Workroom

Processing Audio

I have said "the aim (of compression) is to reduce the size of the audio file without affecting the 'quality' of the audio signal..." The caveat here is 'quality'. What is meant by this? Is it an absolute or a relative term?

Before I expand on this, it is necessary for you to have a basic grasp of what is meant by the term CODEC.

What is a CODEC?

Not only do we need a means of compressing the audio, but we also need a means of seamlessly decoding the file as it is played back. This is the job of the CODEC or Coder/Decoder.

Audio Compression 'Classes'

If one plays back audio on small bud headphones associated with many portable media players one just cannot hear the subtle differences and nuances in music. Full bass sounds just will not be produced.

What if one could compress the audio signal in such a way that the parts we could not hear, were removed altogether? The music would still sound OK but the file size would be much reduced... wouldn't it?

Conversely playback on a Studio Monitor system will reveal every difference and nuance. Any missing audio information will be noticed.

It becomes apparent two 'classes' of compression schemes are needed.

Lossless Compression

Using a lossless CODEC for audio compression, is similar to using ZIP for normal files, in that audio is compressed without any loss in quality. Lossless Compression gives you perfect, bit-for-bit copies of your music. That means it always sounds perfect - exactly the same as the original.

In this category are the APE and FLAC codecs.

Lossy Compression

During the process of Lossy compression the original audio file is first analysed (by using an Fast Fourier Transformation routine). Then frequencies the human ear cannot hear because other louder frequencies dominate and mask them, are removed. Lossy compression removes the part you are not supposed to be able to hear...

In this category are the MP3 and OGG codecs.

archive your CD collection

Audio CD's are not as popular now as they once were. How do you archive your CD collection as lossless FLAC audio files for storage on a  HDD or the WWW and have the ability to burn an audio CD from them if required?

Outline of the Task...

Ripping a CD to one large 'CD Image' file makes sense for achieving  purposes as it is easier to deal with one file (one album) and a CUE sheet holding the track information, rather than a collection of tracks. If one wanted a particular track it could be extracted from the album  length file using the cue sheet. If one wanted an exact copy of the CD  one just loaded the CUE and burnt a CD.

Why is the cue file important?

The cue file contains Artist, Album and Track information. More  importantly it holds the pre-gap timings - that is - the length of the  silence between each track. With this information an exact copy of the  original disc can be made.

What are Meta-Tags?

Embedded within each music file are 'Meta Tags' which contain Artist,  Track and Album information and well as technical details about the  encoder and the quality of the encoding. Album cover artwork can also  be embedded into the music file (though I prefer to add separate files  for the artwork and booklet).

How do I use FLAC+CUE files on a MAC?

As far as I know, there is NO application for the MAC OS that can directly burn an audio CD from a FLAC file using a CUE file to supply the necessary track information.

You can however, make use of the CUE indirectly...

The CUE file is used to split the one large album length FLAC 'CD  Image' into individual track length FLACs. During the the splitting  process you may be able to add the 'Pre-gap' time to the end of each track. You then set the inter-track gaps in the burning program to 0, load the tracks into your CD burning program and start the burning process. You end up with a copy of the original CD when the audio CD is burnt.

Note:
Depending on the burning software the individual track files may need to be converted from FLAC to the WAV format before an Audio CD can be burnt.

'track' centric or 'album' centric?

For most music styles 'track' centric is OK, but it does not work very well for Classical music. Here it well and truly sucks. With classical music, single tracks can result in extremely long file names. Not only this, the 'flow' of the piece may be affected on play back, as the silence between the tracks (the pre-gaps) are no longer as the conductor and/or producer intended.

On the other hand, having single tracks from the outset means that it is a little easier to make a CD from various artists, a 'best off' or the like.

Horses for courses I suppose!