I did some searching on this and did not find the exact answer I was looking for. If I take a 320 kbs file, decode it into a wav and then reencode into 192, will the quality be the same as if it were directly encoded initially into 192? I'm not looking for mathematical exactness, just to the listening ear.

Short answer, no.

Barely longer answer: It might.
Even longer, but not by much: The quality will be definitely degraded, but whether you hear it or not, depends on your hearing and equipment.

Clearly 'no', but this does lead to a question I have been pondering for a while: just how bad is transcoding?

Here's my point. Let's say I want to have my music as, say, mpc on my main computer, to have the best possible lossy format (lossless just isn't feasible due to the large size of my collection). But I also have a portable player that only plays mp3, and I listen to most of my music on this. The question is: just how bad would the mpc -> mp3 transcoded files be? My suspicion is that they would be good enough, and possibly even fairly transparent in many cases.

The point is, if transcoding once from mpc or aac (or whatever) to lame -aps mp3 results in a high enough quality file then I'll start ripping to mpc (or whatever) instead of mp3.

In any case, have there been any tests done on the effects of one transcode like this? [I have searched HA and not found anything] Does anyone else think such tests would be worthwhile??

.dd.

Truth is, there is no reliable data regarding how bad is the effect of transcoding on quality. A test was attempted once transcoding to --preset medium from different sources, but for some reason it didn't get enough participation to get statistically valid results and therefore was indefinitely suspended. Too bad, since I do think it would be worthwhile to know for sure just how much of its terrible reputation is real and how much is myth.

I tested several samples in the test and even though I never found out what my results were, there were always one or two codecs that I couldn't ABX from the direct transcode. The strongest candidates in that test were WavPack lossy and MPC insane, so I guess it was probably them as sources which I couldn't ABX. I don't claim to have exceptional hearing nor equipment, but I wouldn't say I am any less than average, and if I couldn't ABX some transcodes, then they can't be so bad. I heard kinda extraofficially from Steve (sthayashi) that in his insufficient and preliminary results MPC insane was standing pretty good.

Hmmm... you'r right. The transcoding test idea does have some history, which I should have remembered

Is it worth resurrecting Steve's test?