Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Is it ok to convert a MP3 file from 128kbps to 80kbps vbr? I mean, will there be a difference if I convert 1411kbps WAV to 80kbps MP3 vbr? How much will be that difference?

share|improve this question
Do several sample conversions and listen to them, if they sound good enough for you, then its ok. – Moab Apr 8 '12 at 1:10
Yes actually they both sound good to me. I guess I just wanted to know the difference in quality anyway. – user103241 Apr 8 '12 at 1:22
Speech is a different case from music - while the below answers are all correct (you will lose quite a bit of quality if you re-encode lossy audio), for audiobooks it usually does not matter. I regularly use 64kbps for audiobooks and could probably go even lower without significant loss of quality. There are also codecs (e.g. Speex) that are specifically designed for speech, and these can reduce the file size even further (but your iPod won't play them, unless it supports Rockbox). – user55325 Apr 8 '12 at 2:14
@user103241 I always judged audio quality with my own ears, since that is all that matters to me. Maybe you need to define "quality". – Moab Apr 8 '12 at 14:49
By 'quality' I mean the 'difference' as I asked in question. – user103241 Apr 8 '12 at 20:20
up vote 6 down vote accepted

When converting from lossy format to lossy format (i.e. reencoding from 128Kbps to 80Kbps) you will have a difference. You already are starting with a file where data has been removed, and you are removing still more data.

If you convert from a lossless format (e.g. WAV) to a lossy one (MP3), you will also lose data (indeed, this is the entire point of lossy compression: removing unneeded data in order to get a reasonable file size).

As for the differences, often they are subjective: some people can't tell the difference between lossless files and 128Kbps files, some people can. You will have to test them experimentally and find which parameters give you a good quality/file size proportion.

share|improve this answer

Well, you're going about VBR wrong to start with. The idea behind VBR isn't just to get a low bitrate - its to get a good mix of quality and low bitrate. The same source file with maxiumum different bitrates may have same minimum and maximum bitrates, and filesize... but if you need a higher bitrate its better to have it. 128kbps vbr is a good compromise.

Firstly, re-converting a lossy file will result in more quality loss than converting a lossless file. Its a little like how a photocopy of a photocopy is worse than a photocopy. While most algorithms try to throw away stuff we normally won't pay attention to (psycoacoustics) they arn't perfect.

In addition, the it also depends on the source. A good quality recording would 'suffer' more from downsampling than say, a lecture recording off the crappy condenser mic on a voice recorder of some sort.

share|improve this answer
It's a file from audiobook. I'm converting constant 128kbps to 80kbps VBR to reduce some the size of all the files to fit them on iPod. isn't 80kbps more than enough for speech? Should I reduce more quality and size? – user103241 Apr 8 '12 at 0:53
@user103241 as already said, this is subjective: you will have to try a few choices with your own ears. – Renan Apr 8 '12 at 1:09
Ok, I understand I guess the point is that it may sound different to different people. – user103241 Apr 8 '12 at 1:28
@user103241 pretty much so. – Renan Apr 8 '12 at 2:47

There will be a difference. The re-encoded mp3 will suffer the degradation of the lossy copying process twice, just like quality is lost with each generation of a photocopy or audio cassette.

How much is hard to tell. Most people already find it difficult to distinguish a properly encoded mp3-file at 128 kbps from its lossless source, let alone be bothered by the inferior quality. The differences that are audible can largely be attributed to the properties of the mp3 format itself, rather than the encoding process, especially at lower bitrates. To re-encode the audio once would not affect the perceived quality by that much.

If you have access to the lossless files, encoding to mp3 from those is recommended, as this would lead to a better representation of the audio at the same file size, but it may be insignificant to your needs; if 80 kbps vbr mp3 will suffice, chances are converting from lossy will too.

share|improve this answer

WAV and AIFF audio files are uncompressed-- this means that for every audio sample, there is a corresponding unit of data (like 1 byte). For this reason raw audio files are large. E.g., CD audio, which, like WAV, is just a PCM audio stream, typically takes up the entire CD: ~600MB for about 60 minutes of audio. This is why the "bitrate" of WAV files is very high; it cannot be compared directly to MP3 bitrate without taking MP3's compression ratio into account.

MP3 addresses this file size problem in two ways: by "compressing" the data, and also by discarding some of the data before the compression routine is run. This combination of techniques is often called "lossy compression", and for things like audio files, the loss of some information isn't a problem. What you're asking is: at which point does it become a problem?

Unfortunately, there's no easy way to answer that, because the way that MP3 VBR encoding technique "loses" data depends on something called the psychoacoustic model. The basic idea is like this: while the human auditory system can typically perceive sounds across the entire 44kHz range of a typical recording, not all frequencies are perceived equally. For instance, most humans cannot hear below 10Hz (you can FEEL it, though), and many people, especially older adults, cannot hear at the high end either. In between, your auditory system hears some frequencies better than others, with great sensitivity near 22kHz (if I recall correctly), which is around the timbre of a human voice. In all the areas where your hearing is not sensitive, you can throw those sounds away if they are not above a given loudness threshold, because you will not hear them.

This is all a long way of saying: you really need to try encoding it at a given bitrate and listen to it. If you can't tell the difference, there is nothing to worry about. One thing you should be aware of is: if you are re-encoding a file that was already encoded using a lossy codec, you will likely have unsatisfactory results given the file size.

share|improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .