This wouldn't work as the codecs work with human psychoacoustics to leave out sound data that humans are less sensitive to. They are also tuned for 'normal' music in the same way that voice codecs are tuned for human voices. It follows that the only true test is humans listening to music and rating subjectively.
An analogy: JPEG compression discards more information about the colour blue as humans are less sensitive to it. It also focuses on photographic images to the detriment of GIF style logos or POV-RAY type images. If you mechanically compared a JPEG with the uncompressed source then the blue section would show the most difference yet this would be less noticeable to a human than the loss of a similar amount of red or green info. Or if you JPEG compressed a 2 colour GIF logo it would look horrible. Again it follows that they only viable test relies on the subjective evaluation of the target image type by a human being.
sorry, that last comment was out of order. Pretend I said this instead.
Please, never encode files with a lossy codec that are already encoded with a losssy codec. Even if you don't mind wasting your own time to create a file with worse quality than you started with, think of the ogg vorbis developers. They have given you an open source codec of incredible quality and anyone who listens to your music files will think it sucks ass.
Also, try AAC, similar quality to vorbis, and supported by Quicktime and probably iTunes and iPod soon. Unfortunately, not free.
Slashdot Mirror
uptime
This wouldn't work as the codecs work with human psychoacoustics to leave out sound data that humans are less sensitive to. They are also tuned for 'normal' music in the same way that voice codecs are tuned for human voices. It follows that the only true test is humans listening to music and rating subjectively.
An analogy: JPEG compression discards more information about the colour blue as humans are less sensitive to it. It also focuses on photographic images to the detriment of GIF style logos or POV-RAY type images. If you mechanically compared a JPEG with the uncompressed source then the blue section would show the most difference yet this would be less noticeable to a human than the loss of a similar amount of red or green info. Or if you JPEG compressed a 2 colour GIF logo it would look horrible. Again it follows that they only viable test relies on the subjective evaluation of the target image type by a human being.
sorry, that last comment was out of order. Pretend I said this instead.
Please, never encode files with a lossy codec that are already encoded with a losssy codec. Even if you don't mind wasting your own time to create a file with worse quality than you started with, think of the ogg vorbis developers. They have given you an open source codec of incredible quality and anyone who listens to your music files will think it sucks ass.
Also, try AAC, similar quality to vorbis, and supported by Quicktime and probably iTunes and iPod soon. Unfortunately, not free.