I was wondering what people do when deciding which bitrate to use when taking in consideration differences between genres of music.

For example, I normally encode all my music to AAC using Nero's encoder with the -q 0.35 switch. But if I suddenly encode a different genre, is it enough for that type of music? This would be more applicable for the practice of trying to get the lowest possible bitrate threshold, but I'm curious.

Good VBR should be constant quality. Q 0.5 is defaulted so I take that as some rough guide to high quality across the board.

I don't chose bitrate by genre. I use Ogg Vorbis and it is smart enough to chose bitrate by itself. I just select quality level for car audio and portable (lower), speech (lowest) and home use (higher).

Nero's AAC encoder should be more than adequate at determining the proper bitrate regardless of genre. I have encoded some songs at -q 0.45 with various metal genres, industrial, and hard rock music. Nu-metal tends to come out at around 160kbps, heavy/thrash/core/metalcore/hardcore comes out at 170kbps, hard rock at 150kbps, and most of the industrial songs have a bitrate of around 130kbps. So you can see that Nero's AAC encoder can properly adjust itself to the music just fine.

I would be a little worried about the iTunes AAC encoder since it isn't true VBR. Instead you pick a bitrate and enable VBR encoding, it will then make sure that no frames are encoded below the bitrate you chose and some frames can be encoded a little higher. When I select 128kbps VBR, most songs tend to come out at 130kbps or 140kbps but nothing higher than that. That is why I would like to use Nero's AAC encoder but it doesn't work with my 4GB Creative Zen. I simply trust Nero's VBR encoding schemes more so than iTunes and I can easily use Nero's AAC encoder with secure/accurate rippers.

I use the latest Ogg Vorbis aoTuV build at -q5 (~160kbps nominal) for all my music across all genres to which I listen on a regular basis: rock, pop, classical, jazz, R&B, electronica and techno. It's slight overkill, as -q4 (~128kbps) or even -q3 (~96kbps) would be fine for my ears in almost all cases. But the slightly larger average file size of -q5 is quite acceptable based on the storage limitations of my playback devices.

If I used a quality setting much closer to my perceptual threshold (i.e., something around -q3), then I might use different settings for different genres. However, to really justify such an approach, and to make it as accurate as possible, I'd have to ABX several tracks from each genre to test each quality setting against my own abilities of audio perception. That would be worthwhile if I were really crammed for storage space, but since I'm not, Vorbis @ -q5 seems the most reasonable option for me.

Occasionally I'll encode tracks in MP3 (if I get a player that's not Vorbis-friendly, for instance), in which cases I still use the old LAME 3.90.3 with --alt-preset standard, following the philosophy, "if it ain't broke, don't fix it." I've never been able to differentiate anything I encoded with that encoder version and setting from a lossless reference, so there's never been a significant reason for me to "upgrade" to anything newer.

And, more rarely, when I encode to formats other than MP3 or Vorbis (for experimentation or platform testing), then I look up in our FAQs here which encoders and settings to use, trying to estimate the most efficient settings based on written recommendations and on what I know as my own perceptual threshold. For example, if there is a recommended quality setting of 0.45 for a particular AAC encoder, then I'd start with 0.25 and adjust it from there as necessary.

But with any encoding format, I would still seek a quality setting or average bitrate that would work acceptably well for all my musical genres.

Thanks for the replies. I suppose most encoders including Nero's do process files based on their complex, so a given parameter would give very different results depending on the file.