Help - Search - Members - Calendar
Full Version: Encoder efficiency by genre
Hydrogenaudio Forums > Hydrogenaudio Forum > General Audio
atici
Ok I know it might sound ridiculous at first. But is there any comparison of various encoders (LAME, Vorbis, MPC...) with respect to the genre? It is possible that a particular lossy format or encoder might be using an acoustic model that reproduces music of a particular genre better than the others (or similarly produce more artifacts when it comes to music of particular genre). For instance human voice, a song of Carcass and a Bach harpsichord concerto sound like each constitutes different set of signals. I might be too naive to think so but isn't that possible? In particular I have a big classical music collection. Is there any encoder that produces better quality with less information when it comes to classical music?
Thanks.
Mr. Mulder
Check this:
http://www.hydrogenaudio.org/forums/index....f=16&t=1532&hl=
SometimesWarrior
Well, you could probably make one statement about encoder efficiency as far as bits/quality: MP3 is inefficient for metal (or any other music with lots of fuzzy distorted guitars) because of the MP3 format's missing scalefactor band for frequencies >16KHz, which causes bitrates to bloat when there's lots of high-frequency information. Mr. Mulder's link addresses the rest of the post.
Garf
Well, it depends really. If you can live with a lowpass at 16Khz, MP3 will not do worse on metal than on other music.

I don't think a 16kHz lowpass is very bad.

--
GCP
SometimesWarrior
QUOTE
Originally posted by Garf
Well, it depends really. If you can live with a lowpass at 16Khz, MP3 will not do worse on metal than on other music.

I don't think a 16kHz lowpass is very bad.
That's true; I was going to say something about that, and the -Y switch, but then I fell asleep wink.gif

(A bit OT: I don't think this forum ever had a survey regarding what level lowpass people can detect in music; there was only that tonesweep test.)

So doesn't MP3 handle classical music pretty well? I'm guessing that pre-echo doesn't play much of a role, so MP3's biggest deficiency isn't exploited. the alt-presets give about the same bitrate for classical as they do for everything else, but I've heard that for portable listening, MP3 can encode classical music at a lower bitrate than other genres and still sound pretty good.

I don't have golden ears, so the only alt-preset insane clips I can ABX are sharp transients. I don't know what other kinds of MP3 artifacts are still hanging around at that bitrate, but I haven't noticed them.

So does classical music stay transparent at high bitrates much more often than other music genres that have sharp attacks or lots of high-frequency content (that the user wants to keep)? Or does classical music have just as many "killer clips" as any other genre? (Well, Dibrom's intelligent techno is in its own class...)
Dibrom
QUOTE
Originally posted by SometimesWarrior
So doesn't MP3 handle classical music pretty well? I'm guessing that pre-echo doesn't play much of a role, so MP3's biggest deficiency isn't exploited. the alt-presets give about the same bitrate for classical as they do for everything else, but I've heard that for portable listening, MP3 can encode classical music at a lower bitrate than other genres and still sound pretty good.


It depends really. LAME has had some problems in the past with improper ath adjustments and classical music. For the most part, they don't really seem to be a problem anymore, except for the fact that sometimes you can get noise pumping still. This is usually caused by not encoding high frequency content up to a high enough level with regularity. Again, for the most part --aps doesn't really suffer from this, but LAME on it's own, usually does.

Aside from that, it depends on the specific piece as to whether or not there will be problems with pre-echo. Instruments like harpsichords, dulcimers, classical guitar (think Bernd Steidl), etc, can all cause pre-echo too.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2008 Invision Power Services, Inc.