This is a crosspost of mine from the BBC Backstage list, felt it was relevant to the earlier discussion about AAC/AAC+ and World Service material so posting for people to see what I was on about.
==================== > Chris, > Can't speak for my colleagues elsewhere in radio, but WS doesn't > transcode between codecs anywhere in our longform workflow and never > has done. I did a quick bit of testing and it appears that WS Listen Again material is only available through /iplayer as 64kbps AAC+, so it all sounds a bit sub-par compared with domestic channels' 128kbps AAC. Perhaps it's ingested in a different way or do you encode internally then deliver to the iPlayer team? I have noticed the PIDs for WS material have a different range (p*** as opposed to b***)... To show you what I mean about the MP3 vs AAC quality difference, here's a quick quality comparison (randomly chose an episode of The Archers, from Radio 4 the other day). The first time is the MP3 encode, the second is the AAC encode (served by default through the Flash player): http://bit.ly/bbciprtest1al (~3.8MB) Even on average speakers you should be able to hear a difference - the MP3 is "rumblier", warbly and speech is distinctly less clear with noticeable distortion under the main frequency of the speaker's voice. If you use headphones or good monitors you should be able to clearly hear the inferior quality of the MP3 version. Comparing the two clips spectrally also shows a visible difference, there's less 'cohesion' in the MP3 clip, what appears to be double-encoded noise and the frequency ranges containing the speech energy are less distinct. Neither speech nor musical content comes off well in the MP3 versions - either the iPlayer's using an *AWFUL* MP3 codec (because both the AAC and MP3 files are 128kbps) or the MP3 version is being transcoded from the original AAC source, which would explain a lot. _______________________________________________ get_iplayer mailing list [email protected] http://lists.infradead.org/mailman/listinfo/get_iplayer

