Originally Posted by BIslander
Originally Posted by m_vanmeter
yes, there is a difference - "lossy" digital audio is compressed, lossless is not.
Both lossy and lossless codecs are compressed. That's the whole point actually. They remove some of the data during the encoding process to save space and put the data back in the decoding step. Here's the difference: The lossy codecs use so much compression that some of the original data cannot be restored while the lossless ones use less compression and all of the data is restored.
Exactly. As pointed out above there are two kinds of compression - lossless compression which is usually limited to about 1/3 to 1/2, and lossy compression which can, at a variable cost in sound quality, range up to 20:1 ( reduction of file size to 5 % or less). What may not be obvious is that various schemes for lossy compression and various kinds of music can produce differing results. Some place around 10:1 audible degradation can become very noticable, but that partially depends on the musical work and the exact compression algorithm.
Lossless compression can produce ideal and perfect reconstruction of the source music file.
There are far more different lossy compression schemes than lossy decompression or reconstitution schemes. A file can be compressed using one of 10 different schemes but decompressed using just one and the same program. Most lossy compression schemes can compress a file by varying amounts ranging over a 10:1 or larger range. Some schemes will analyze the file in sgements, estimate the maximum amount of compression for a given level of sound quality, and actually change the amount of compression of that file segment to optimize file size and sound quality. This will be repeated again, different amounts of compression will be chosen for each file segment, and this will be repeated again and again over the duration of the file.
The idea that expensive equipment is required to effectively hear the adverse effects of excess lossy compression fails because problematical situations can be often detected with headphones costing $50 and portable digital players in the same price range. Simply picking a different lossy compression methodology or degree of compresion can correct the situation with the same music and equipment. Good quality speakers costing less than $200 a pair can be useful tools because most lossy compression artifacts affect the upper frequency ranges so mongo bass response is not usually needed for critical listening. Generally, a good quality pair of headphones or earphones can make any adverse effects due to lossy compression more noticeable than a far more expensive pair of speakers.
A key point is that the methodology of lossy compression and the kind of music are critical. For example MP3 files have been around for about 15 years. 10 years ago it was far easier to detect problems when they existed for a given amount of compression.
The correct methodology for detecting problems with lossy compression is a double blind test using music known to be especially demanding. Modest amounts of degradation may only be noticeable with direct comparisons unless specific artifacts can be heard.