Why is the average bitrate of ogg files encoded with quality setting always considerably lower than target? A dumb question maybe, but all my files are like that. For instance, a solo piano piece encoded at -q7 (~224kbps) results in an ogg file with an average bitrate (according to ogginfo) of only ~180kbps. It never peaks above ~194. Why so much smaller? Ok, solo instrument you say. But the average is not much higher with symphonic tracks. Admittedly, it's sort of a moot point because I can't tell -q7 from lossless, but I've since gone to -q8 for a number of reasons, one of which being that I can't get comfortable with seeing the bitrate under 200. I guess the question is why claim a given target bitrate for a given quality setting when the averages never come close? Does it all come down to the comparison with mp3 vbr settings, where a smaller ogg file sounds as good as a vbr encoded mp3 at the next level up (i.e. -q6 ~ V 2, etc.)?

Put simply, it's just that classical music is generally more easily compressed than some others. The bitrate equivalent indicated for a particular quality setting is estimated on an 'average' musical content. As you've seen, classical often requires fewer bits while heavily synthesised music usually requires more.

You should not concern yourself with the actual bitrate as the encoder will allocate what it takes to encode the content at the specified quality setting. You will find a very similar situation with mp3 encoders, and the rest.

Most of my music is classical, and I find it quite satisfying see a medium bitrate and at the same time know I'm getting quality 0.5 when I use Nero. Excellent all round. I see the bitrate in iTunes.

When I encode to Vorbis (using Quality 5) I simply don't know what bitrate I'm getting, since the format doesn't import (natively) into iTunes and Rhythmbox on Linux simply claims the bitrate is always 160kbps. I haven't bothered to look for anything that will tell me the accurate bitrate. But I can feel satisfied at the filesizes. With classical guitar or songs with piano accompaniment Nero often gives bitrates as low as 100--110, and I should imagine (from the filesizes) I'm getting very similar bitrates with the Ogg Vorbis encoder.

Encode a Metallica song (see the bitrate go way up). And stop looking at the dynamic bitrate while listening to music.

Interesting. That makes sense. But I guess what is still confusing me is that I was expecting to see bitrate spikes approaching or even exceeding the target on more musically complex passages, and drops on silence and pauses and 'thin' textures, averaging out to whatever. It's just so much lower than I thought it would be. I'm not forgetting, either, the fact that ogg uses different psychoacoustic models/algorithms which are more efficient (smaller files), and that the point of q settings is to achieve a certain (subjective) sound 'quality' level which cannot be accurately gauged solely in terms of bitrates.

Like Nick E, I was thinking the average must be close, even though Amarok and Rhythmbox only show the nominal rate like he says. Even when playing them back in a player showing the dynamic bitrate jumping around I thought it must be close. But then I checked MediaInfo and ogginfo.

Oh well, like I said it's no biggie--ogg at the higher levels might as well be flac for all I know. Seems its time to skim back over my digital audio handbook.



(John33--Forgot to add: nice spit (Mk IX?) in your picture. That's my all-time fav aircraft. Nothing like the sound of a merlin. Just wish I could track down more good-quality audio downloads of airshow flybys of merlin-engined spits. I can listen to that engine for hours!)

I'm not sure which model it is TBH! The sound of the Merlin is quite magical but I also have to say the Griffin sounds amazing too. I was at the Battle of Britain Show at Duxford in 2000(?) and there were many, many WWII aircraft flying, but the Spits were certainly my favourites, they are so much more elegant than the Hurricanes.

You're lucky...all we see over here in the us is mustangs, which is fine because they have merlin engines too! Griffon engined spits sound pretty amazing and are certainly more powerful, but the merlin just has that je ne sais quoi...That's a IX model or later spit in your photo--note the pointy, deep-chord tail, 4-blade prop and larger airscoop under the nose. R. Mitchell really produced a work of art when he designed that plane back in the '30's.