Re-ripping audio CDs

March 30, 2006 12:04 PM

Ripping tracks from a audio CD-R...what bitrate to use?

I have a bunch of music CDs I burned in the past that I would like to rip back to my computer. My problem is that I don't know what bitrate they originally were in. All of them are at least 128, but that is all I know. Since I burned them as albums, every song on each CD should have the same bitrate, but each CD varies. Is there a simple way to find out? The reason I would like to know is because I'd like to re-rip the CDs at the highest bitrate possible without compromising the song quality. I also use Nero generally for these kinds of tasks, if you know how I could do it with that. Also, if there's something wrong with my line of reasoning please let me know, I am not too familiar with bitrate and quality in general.

posted by apple scruff to computers & internet (16 comments total)

If you're in XP you can show the bitrate info in the "details" view of a folder. right click the info bar under the address pane in the window to get a drop down menu of such options.
posted by dong_resin at 12:14 PM on March 30, 2006

As for re-ripping, once encoded at a given bit rate the data is compressed and the loss in quality is permanent.
posted by dong_resin at 12:16 PM on March 30, 2006

Basically, the quality of the source music should have been preserved at whatever relative bitrate it had before you burned it. Ripping it back, just choose a bitrate that represents the highest bitrate you likely would have used before.

The audio on the CD is preserved as a normal waveform, and uncompressed. Except that it was generated from compressed data. As long as you set the ceiling high enough (I'm thinking around 192kbps, as it's a good balance of CD-like quality and space), you shouldn't further diminish the quality of your music to any noticeable degree. If you had ripped them all at 320kbps, and you re-rip them at 128kbps, then the quality will suffer.

There's no easy way for you to determine the absolute bitrate of the source material, since it exists in a technically uncompressed form on the CD. In fact, it's bitrate would be that of any regular audio CD, or "high." (I'm too lazy to look it up.) That said, just keep a consistent decent ceiling, and you'll be fine. You honestly won't notice any major differences form the burned CD to your freshly re-ripped @ 192kbps MP3. Keep in mind, some audiophiles will. You are (likely) not one of them.
posted by disillusioned at 12:19 PM on March 30, 2006

Since WAV and CDDA data have no way of documenting the bitrate of the source used to generate them, you won't be able to get that information from the CDs you have.

Whether or not you get the same bitrate as before, there will be quality degradation. I might be wrong about this though. MP3 encoding is about throwing away extraneous information, so I suppose it is possible that decoding and re-encoding wouldn't have much of an effect. Realistically though, I would expect to take a quality hit no matter what you do.

Are you worried about not having enough disk space? You can always just pick a bitrate higher than what they were encoded at. Rarely does anyone go higher than 192kbs. That might be a good target.

Try a disk or two and see what it sounds like. You might not even be able to tell the difference.
posted by joegester at 12:25 PM on March 30, 2006

Sorry I didn't mention this before, but the CDs were compiled of downloaded mp3s...I did not copy an actual audio CD. But I'm not an audiophile by any stretch of the word so I guess I won't really have a problem.
posted by apple scruff at 12:27 PM on March 30, 2006

Am I understanding this right:

1. You had a bunch of MP3s of various bitrates
2. You burned them to Audio CD? (or was it a data CD?)
3. Now you want them back as MP3 files on your hard disk?

If you burned them to audio CD you are going to loose quality by ripping them again unless you rip to a lossless format. However, lossless formats take a lot more space and can be a problem on portable devices for various reasons (including filesizes in the 20-30MB range). If you have to rip back to MP3s I would probably go with a high bitrate, like 256 or 320kbps.

If they are just MP3 files that you burned to a data CD then you should just copy them off to your hard disk. Don't rencode them, because quality will just drop.
posted by Good Brain at 12:29 PM on March 30, 2006

Yes, let's say I downloaded an album via BitTorrent, where all the tracks were present as mp3s at 192kbps (or at least all the same bitrate). I burned them as an audio CD. Now I would like to rip that CD back to my computer, except in this case I don't know what the bitrate of the tracks was.

Good Brain writes "If you have to rip back to MP3s I would probably go with a high bitrate, like 256 or 320kbps."

But if the tracks were originally little more than half that quality, wouldn't that cause major distortion?
posted by apple scruff at 12:33 PM on March 30, 2006

wouldn't that cause major distortion?

No. The audio on the CDs are uncompressed versions of the compressed MP3s. For all intents and purposes, they sound exactly the same as those compressed MP3s. If you want to maintain the sound quality of the originals, you'll need to rip then at a higher rate so you avoid the problem of compounding losses due to compression.
posted by pmbuko at 12:41 PM on March 30, 2006

avoid = minimize
posted by pmbuko at 12:41 PM on March 30, 2006

"But if the tracks were originally little more than half that quality, wouldn't that cause major distortion?"

No. I'll admit I don't know all the ins and outs of audio codecs, but the only problem with ripping at a needlessly high bitrate here is wasted hard-drive space. Putting anything through two rounds of lossy encoding is obviously not ideal, but I believe the worst thing that will happen here is that your high-bitrate MP3 will be trying to preserve acoustic details that had been stripped out on the previous run.

Here's a visual analogy you can try with an image editor:

Take a high-quality photo file. Save it as a JPEG with aggressive compression: this is equivalent to your original MP3. Next, save that as a TIFF: this is equivalent to your burned CD. Next, save that TIFF as a JPEG with very light compression: this is equivalent to your re-ripped MP3. Compare that to the first JPEG. You will be able to observe differences, but they won't be dramatic.
posted by adamrice at 12:47 PM on March 30, 2006

wouldn't that cause major distortion?

Like pmbuko said, no. If you want to hear for yourself, I've run a generational loss test of a number of audio tracks, and posted the results on my site.

Yes, there will be generational loss from transcoding. But in one generation, unless you're using low bitrates for all the steps, it's not what I'd call "major".
posted by hades at 12:49 PM on March 30, 2006

adamrice: JPEG has the property where if you decode and and re-encode it at the same settings, you will get the same file back. mp3 however is not like this. Decoding and re-encoding will hurt the quality.

My advice is: Try it at a few settings (128, 168, 192, whatever) and keep the smallest one that doesn't sound bad to you.
posted by aubilenon at 4:40 PM on March 30, 2006

I'd recommend the same thing I'd recommend for regular audio: the LAME --alt-preset standard or -V 2 presets.

If you're using EAC, find and download LAME (too lazy to find link at the moment) and enter as the command line in EAC:

LAME.EXE --vbr-new -V 2 %s %d

(if you're using an older version use --alt preset standard %s %d instead)
posted by neckro23 at 11:59 PM on March 30, 2006

For info on EAC (a ripper) and LAME (an MP3 encoder), or to generally geek out on audio encoding, visit Hydrogen Audio.

Another possible way to minimize re-encoding artifacts is to transcode to a codec other than MP3, such as Vorbis (.ogg) or AAC (.m4a).
posted by unmake at 2:06 PM on March 31, 2006

"Another possible way to minimize re-encoding artifacts is to transcode to a codec other than MP3, such as Vorbis (.ogg) or AAC (.m4a)."

I think that would make it worse, since each codec is looking for different acoustical cues. If AAC is looking for cues that were stripped out by MP3, you're getting the worst of both worlds.
posted by adamrice at 6:59 AM on April 4, 2006

The first relevant test I pulled up from HydrogenAudio -
Short re-encoding blind listening test, wavpack - mp3 - mpc - aac - vorbis - deals with transcoding in the opposite direction (non-mp3 source converted to mp3), but concludes with what I said earlier: mp3->mp3 seems to create the worst result.

My earlier statement was essentially a paraphrase of the general consensus at HA. I'm not in a position to give a clear explanation why, other than that it's because different codecs use different acoustical cues (or rather, alternate means of analyzing a waveform, breaking it up into spectral regions, and lossily encoding said data) that transcoding from one codec to another is preferable from same codec to same codec.
posted by unmake at 4:00 PM on April 4, 2006