Hello,

This my first post here and Im relatively new to the world of good audio. I am wondering how much quality will I lose if i change formats of my music from AAC to MP3 or vice versa? Is there any particular advantage to one over the other? I use a mac and have it hooked up my iPod but recently I feel in love with the sound of the Sansa fuze hence my question. Thanks in advance and hope this makes sense.

Yes, there will be quality loss. As to how much, it is dependent on your original codec+bitrate and the target codec+bitrate. Please read the page on transcoding in the knowledge base:

http://wiki.hydrogenaudio.org/index.php?title=Transcoding

Here is a listening test on transcoding to mp3:

http://www.hydrogenaudio.org/forums/index....c=32440&hl=

Transcoding from lossy to lossy is generally a bad idea. Best to rerip and reencode from a lossless source (such as CD).

Thanks a lot Ive been looking for a site like that for the last couple of days.

Never transcode lossy to lossy. Imagine one algorithm throwing out 80+% of the music and than another taking that and throwing more away in it's own process. Always have a copy of the lossless files around to experiment with lossy encoding.

People can transcode between lossy formats so long as they don't hear the differences. After all, if they can't hear the difference, what exactly are they missing out on. I do think that they need to understand the possible flaws of lossy-to-lossy transcoding but it should ultimately be decided by their ears. I don't think that someone should say "never transcode lossy to lossy."

I'm a bit interested to know what it is about the sound of the Sansa fuze what would cause people to fall in love to the point that they are willing to lose quality due to lossy to lossy transcoding.

I'm just wondering, since you're considering switching for the sound, if you've experimented with different headphones with your iPod. You'll get a dramatic improvement in sound quality just by switching from the stock buds.

It's not quite that bad, because a lot of that 80% is "nothing useful".

Here are a couple of analogies that don't involve lossy compression... Let's say you have a 44.1kHz, 16bit file ripped from a CD. Let's say you resample it to 96kHz, 24-bit. That new file is about 3 times as big, and it contains 3 times as much data, but there's no more useful audio data in the bigger file. Now, let's take that 96/24 file and downsample back to 44.1/16... We can throw-away 2/3rds of the data without loosing any real audio information.

Or, say you have a 100Hz tone sampled at 44.1kHz. If you resample to 4kHz, you are throwing-away 90 percent of the data, but you have lost no audio information... You can still perfectly reproduce the 100Hz tone!

Back to lossy... Both MP3 and AAC are trying to throw-away the least audible information. If you transcode (or encode a 2nd time), the encoder doesn't "look for" more audio to throw-away. There isn't as much real audio data in the file the 2nd time, and it doesn't need to throw-out anything (assuming the same bitrate).

Of course MP3 and AAC are not identical (they throw-away different information) and the encoders are not perfect, so there will be some additional audio data loss. But at a high-enough bitrate, the additional loss might not be noticeable.

That's an excellent rule-of-thumb, but sometimes we don't have an uncompressed original... You may have purchased an MP3 from Amazon, or an AAC from iTunes, or ripped an AC3 file from a DVD, etc.

Hello,

It has been stated numerous times before, but it all depends on your ears.

At the ripe age of 43, I find it impossible to differentiate a 192 kbps AAC from the original WAV. I still insist on ripping all my music at 256, although I could probably get away with a much lower bitrate and still not being able to hear any difference whatsoever...

I have a large number of MP3's @320, @256 and even @192, and I have them all transcoded to AAC 256 VBR in ITunes (yeah, for simplicity's sake).

ABX'ing the original MP3's and the the transcoded AAC's, I could not hear any difference between the two sound files, as long as they both had higher rates than 192 kbps. Something nasty happens when you move closer to 128 kbps though - it is not just the frequency loss, but an unpleasant wishy-washy effect that starts creeping in. This effect is in turn greatly magnified in the transcoding phase, so I wouldn't recommend it. But as long as you stay above 192, you can happily get away with a one generation transcoding. At least if you are older than 15 and have the hearing of a bat.

//Tom