I'll probably never buy music from iTunes, or any other online music store, because they all use constant bit rate audio encoding formats. Once I heard the incredible difference in fidelity between variable bit rate (VBR) and constant bit rate (CBR) encoding, I can never go back. And if I'm spending my own money to "own" this music, why pay for the crappy encoded version anyway? I'd rather buy the CD with the raw, uncompressed versions of the music and rip it myself.
Why don’t you use Ogg Vorbis? A song encoded in Ogg Vorbis quality level 3 (nominal bitrate of 112 kbit/sec) sounds better than one encoded as a 128 kbit/sec MP3. Personally, I can’t hear any differences at quality level (nominal bitrate of 128 kbit/sec) compared with the original, but that might be caused by my equipment.
Ogg Vorbis quality settings go from -1 (45 kbit/sec) to 10 (500 kbit/sec). It is always VBR. Even if you select quality 10, the actual bitrate will, most of the time, always be lower than 500 kbit/sec - it doesn’t know what to do with all those ‘unnecessary’ bits!
Anyway, I was telling this because you mentioned you wanted the most bang for your byte.
Also, I think there’s something inherently wrong about MP3 VBR. Why does the encoder limit itself to multiples of 32 kbit? I haven’t seen the Vorbis encoder do that.
I agree that the main advantage of a superior encoding format is better bang-to-byte ratio. So if I was shooting for 160 kbps average using MP3 VBR, I could perhaps do 128 kbps average with Ogg. The results in the listening test study were decisive in favor of Ogg and MPC:
Personally, I can’t hear any differences at quality level (nominal bitrate of 128 kbit/sec) compared with the original, but that might be caused by my equipment
You should try some A/B testing with the uncompressed WAV files and Ogg files of various encoding levels to see what your threshold is. Mine is definitely 160 kbps, and I hear a HUGE difference in dynamism betwen CBR and VBR at that bitrate. The music is much more “alive” with VBR.
Heh, a man after my own heart, thanks for spreading the word.
The strange thing is that I happened to be ripping old Robert Plant CDs to VBR at the very moment I read this.
I just do 128KB/s avg for various reasons:
It’s still way ahead of most of the retail encoding.
I rarely get the luxury of headphones or close listening these days (now that I’m a boring responsible grownup who needs to hear the phone).
Being a coder, 128 is just a nice familiar number (am I the only one who rounds to binary in real life?). It also helps preserve the 1min==1MB exchange rate, which is very handy for estimating capacities.
Matthijs –
Here’s why I don’t use Ogg: there are hardly any portable/car players which support it (or there were last I shopped ~1yr ago). Otherwise, yes it’s great.
I would suggest migrating away from Audiograbber to Exact Audio Copy… the newest version is amazing. Audiograbber hasn’t been truly updated in ages-- I was surprised to hear it was even still around. I remember doing some work for the author at least 5 years ago, and getting a free registration code in reply.
Also, I would check out hydrogenaudio.org’s forums for tips on command lines and what builds of LAME are the best ones. there is also “ABR”, which is somewhat like VBR… but yeah, you can read up on it
You might want to try a couple of things; first, use the --preset switches with LAME, even if it means encoding from command line, and second, ditch Audiograbber. Real Men ™ use ExactAudioCopy (http://www.exactaudiocopy.de/). Not only can it flawlessly rip audio even from scratched CD’s, it can also burn your rips and it’s quite possible you’ll be able to make byte-perfect copies since EAC knows about read/write offset correction.
You might hear the difference when you burn the 96 kbps back to cdrom and ask somebody else to do a blind test with you. I’ve got no doubt that you’ll instantly hear which CD is the original and which was degraded to 96 kbps. You computer speakers probably aren’t a good indicator of sound quality.
Of course if you only ever listen to your MP3s on a computer 96 kbps might be OK, but me being an audio junkie I can’t stand the idea of losing that much quality.
EAC all the previous posts are right, you are using a vastly inferior ripping application.
I have over 250 gigs encoded in FLAC, and while it is mostly for archival purposes, you just haven’t lived until you listen to your MFSL albums encoded in FLAC for year, after year, after year in all their pristine glory without any degredation. Furthermore, with storage space as cheap as it is nowadays I just don’t see the need for lossy audio compression.
I guess you are all super hero’s because I can’t tell the difference between a CD and the same CD encoded in 96Kbps and my ears are perfectly fine. I have made a lot of research in this area some time ago before I ripped all my music to mp3 so I could get rid of my cds that were taking up to much room. I decided to encode them in 96Kbps because I’ve read report where it said that the average human can’t tell the difference at above 80Kbps with pop/rock songs, so I went for sure with 96Kbps, in case I had above average ears. Most people do not believe this, so I even tested it with some and they really could not tell the difference when they did not know at which compression rate it was encoded.
Back in 2001 when I was seriously testing this stuff (with a then-relatively-high-end Xitel Storm Platinum, Aureal 2-based audio card and Sennheiser HD600 headphones through an Akai AA-R30 amp) I could tell the differences between the original and even 320kbps CBR encode with LAME. There was a slight degradation in the dynamic range, particularly the bass which was, for lack of a better word, flatter.
I’ve used LAME --r3mix for years, but recently (a couple of years) moved to Vorbis - particularly the oggencgt3b1 (IIRC) encoder fed through Exact Audio Copy (which comes highly recommended). I did some serious pseudo-blind listening tests (i.e. listening to the same pieces of the same track on a lame --r3mix encode, oggenc -q 6 encode and the uncompressed WAV in a looped, shuffled playlist for about 40 minutes) and found Vorbis to be somewhat higher in quality while slightly lower in disk space.
Add to all of the above the fact that Vorbis is genuinely patent-free (no royalties to Franhaufer by anyone) and you’ve got a winner. Most portable equipment (including no-name flash players) supports it nowadays, I just wish it were easier to find car electronics that does (there’s only the Kenwood Music Keg and Yakumo Hypersound Car, both very difficult to get).
My ultimate ripping goal is to have a complete lossless archive of all my CDs (probably using EAC) and then have the ability to batch convert it to the “best” compressed format du jour. I’d probably just play the lossless one on my home stereo (if there is no stuttering over the Wi-Fi net) and then one fairly well compressed one for portable audio players. Today that might be 160k CBR or a medium quality VBR and in a couple of years it might be something else and eventually it might be lossless as the memory density of the players increases.
The issues right now for me are that I use an iPod Nano so iPod/iTunes compatiblity is important for me, and I haven’t found a good way to preserve all the tagging from the lossless format to compressed formats - especially if you take advantage of “special” tags like iTunes ratings.
Hey dont give up on web sites just yet. I know a site that offers downloadable music which is encoded on-the-fly, set to your parameters. Plus, they charge you buy the megabite (not the song) so its much cheaper that itunes. Some formats they carry are: Mp3, ACC, OGG, FLACK, Lossless and they support VARIABLE BIT RATE ENCODING.
I cant put up the the link but if you email me i’ll give it to you.
