On Nov 18, 2014, at 04:06 35, Alessio Elmi <[email protected]> wrote:
> Mp3 outperforms mp2 at the same bitrate, This is a great over simplification, one that glosses over a lot of context that is necessary in order to choose the correct codec for a particular job. “Outperforms” it how? In complexity? Latency? Perceived quality? Robustness? The tradeoffs between the various layers are such that each are optimized for particular roles. Layer 2 is optimized for higher bit rates with moderate encoder complexity, reasonably low latency and good resistance to cascading codecs, which makes it a good choice for local data storage (and hence its use in Rivendell). Layer 3 OTOH is optimized for low bit rates with much higher encoder complexity and significantly greater latency. It also tends to degrade more quickly when cascaded. Thus, it is well suited for applications where cost of transmission is an overriding factor —e.g. Internet applications, but is a poor choice for primary storage in content delivery systems. > so from a theoretical point of view your imported > mp3@256kbps will probably sound even "better" than imported > mp2@384kbps. Not so, not for any modern MPEG encoder of which I’m aware. Layer 2 consistently turns in better ratings over Layer 3 in double-blind listening tests for bit rates greater than 128 kB/sec or so. Layer 3 wins for lower bit rates. It’s a widespread misconception that Layer 3 constitutes some sort of ‘upgrade’ over Layer 2. This is simply wrong. Both layers are codified in the same ISO standard and each is optimized to work best in particular situations. One is not automatically ‘better’ than the other. They are simply different things. Cheers! |----------------------------------------------------------------------| | Frederick F. Gleason, Jr. | Chief Developer | | | Paravel Systems | |-----------------------------------------------------------------------| | The real problem with hunting elephants is carrying the decoys. | | -- Anonymous | |----------------------------------------------------------------------| _______________________________________________ Rivendell-dev mailing list [email protected] http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
