Here are tips on how to set the variables when converting via iTunes.
- Stereo Bit Rate (data rate). This is variable from 16 kbps to 320 kbps. The higher the rate, the higher the fidelity and the more space taken up by the file.
- Use Variable Bit Rate Encoding. This varies the number of bits used to store the file, based on the needs of the program material. Although it can create smaller file sizes, VBR files are not compatible with every single MP3 player, so it’s probably best to leave this unchecked. If you do select this, another pop-up menu lets you specify the level of quality.
- Sample Rate. Selecting Auto chooses the same sample rate as the source material, which is usually the best choice. Choosing a lower sample rate than the source creates a smaller file size with the tradeoff being reduced fidelity; choosing a higher sample rate than the source creates a bigger file size, but gives no audible benefit.
- Channels. Auto will create a mono file from a mono source, and a stereo file from a stereo source, so this is usually the best option. If you want to halve a stereo file’s file size, you can choose mono although of course, you’ll lose any stereo effects. Conversely, a mono file at the same bit rate as a stereo file will sound better (but of course it won't be stereo).
- Stereo Mode. This is available only if you choose Stereo for channels. At bit rates under 160kbps, the Joint Stereo option can improve sound quality by not devoting unneeded bandwidth to redundant material.
- Smart Encoding Adjustments. This causes iTunes to analyze the encoding settings and source material and make the appropriate adjustments. Uncheck this if you’re going to do custom settings.
- Filter Frequencies Below 10 Hz. I recommend always leaving this on, because even if your source material does contain frequencies below 10 Hz, very few transducers can play back frequencies that low. Therefore, there’s no need to waste bandwidth on encoding what are essentially sub-sonic frequencies.
As far as the file to be converted is concerned,
beware - encoding to MP3 can create peak levels higher than your master (and trivia fans, this can happen with Acidized files as well). The Sonnox Pro Codec mentioned previously, which is to MP3 converters as "War and Peace" is to a comic book, measures and automatically reduces the level of the incoming signal to avoid overloads.
When MP3 conversion falls down the most is high frequencies. Therefore, you might as well roll them off before converting so the MP3 converter can devote its attention to frequencies it knows how to handle. Ditto low frequencies most speakers can't reproduce.
At my seminars, I often play examples of
only the signal that's removed during the data omission process (it's not really data
compression, data is thrown away to create the smaller file size). As you can imagine, it's not that bad at 320 kpbs. At lower bit rates AAC and MP3 can be very different. In all cases, though, what you tend to lose are the percussive qualities of audio.
Interestingly, if you isolate what an MP3 takes out and compare it to what an Aural Exciter adds, they're very similar...
Remember too that Ogg Vorbis is open source, free of licensing fees, and most will tell you the quality is better than MP3 for equivalent bit rates. SONAR can import and export Ogg Vorbis. If you're compressing for your own use, use it instead of MP3. Also, for lower bandwidths, IMHO WMA outperforms AAC although it seems Microsoft has pretty much abandoned the WMA format.
Finally, garbage in = garbage out. The better your mixing and mastering, the less violence MP3 conversion will do to the signal.