Back To The Main Page Why you should avoid mp3s and other lossy formats
 
1. What do the "lossy" and "lossless" mean?
There are many ways to compress a wav file - basically, there are methods:
-
lossless compression methods - methods that completely maintain the original quality of the wav file.
  A drawback of keeping a full sound is the rather small compression ratio (large files). 
-
lossy compression methods - highly effective methods where high compression distorts the original sound.
  Here, the sound is damaged once and forever and there is no way to restore it back to the original.


2. Related software and file extensions (links)
a) lossless:
software(link) extension features
MKW .shn  An external checksum allows identiication of  the particular recording before  trading or download. Highly used!
MONKEY AUDIO .ape  A good compression ratio and a  friendly software frontend. An internal checksum.
FLAC .flac  An average compression ratio and an awful frontend, but many people have got used to this one. An internal checksum.
TTA .tta  A new one, still under development. It is said to have an extremely high compression ratio - up the 30% of the original

b) lossy:
software extension features
a variety .mp3  The most common format as you know it.
"musepack" .mpc  Sneaky one! the files are really large, but the sound is still poor. What the f**k is this one for?
no idea .vqf  Not much to say

c) other recommended software:

auCDtect

The guys who came up with TTA also wrote this little utility for recognizing lossy / lossless-based wav files.This is a priceless software for any audio collector - it can recognize the origins of the wav files on your HDD. It is sometimes necessary to analyse a recording you have downloaded or traded and you are not surewhetherf the music is an authentic lossless audio or just mp3s or stuff. Also, my friend Sasa Kanocz made a  frontend for it.

Visit the auCDtect authors' web: True Audio Software
 
download auCDtect and the frontend from this site (zipped package 109 kB)
   
preview

NOTE: You will need Microsoft .NET Framework to run the frontend.
 

    

3. A brief description how the lossy compression damages the sound quality

The lossy methods take advantage of the fact that the human ear does not have the same sensitivity throughout the whole frequency spectrum. The point is that the frequencies people can't hear properly are completely removed. This results mainly in colouration of the sound (based on the the well known fact that the colour of a sound is created with help of particular frequency-components of particular intensity).

4. Examples of the sound-distortion
Since the frequency spectrum changes with time, I took a snapshot at one particular moment. The links below show how the frequency spectrum changes after both lossless and lossy compression (mp3-320kbps and mp3-160kbps). Note the changes in the upper part of the frequency spectrum (frequencies for which the human ear has relatively little sensitivity were removed by the mp3 algorithm).

Click on the following links to see the results:

Pink Floyd: Hey You  ...

1979, a relatively new studio recording

a spectral analyser

Eric Clapton: Layla  ... 1975, a decent audience recording
Pink Floyd: Green Is The Colour  ... 1968, a very old audience recording
Robert Johnson: Ramblin' On My Mind  ... 1936, a very old studio recording

5. Conclusion
Of course it always depends on the quality of a person's ears and the degree of compression used, but at any rate - the file is damaged and there is no way back. Consider a scratched postage stamp - it can have just a small scratch somewhere in the corner and its value is gone - although, it still looks quite good.

6. Similar articles as i have found them on web
http://www.pwarcham.cjb.net/
http://www.geocities.com/radioheadbootleg/index.html

thanks to Roy and Kate for help and comments :))