Main Site Navigation

KeyboardMag.com >> This Month >> Dumbed-down Audio
Images
External Weblinks

The Soft Studio

Dumbed-Down Audio

| September, 2006

Whether you’re exchanging files with other bandmates, posting tunes on a website to promote your band, offering digital downloads for sale, or transferring your songs to an iPod or other portable player, data-compressed files save memory while maintaining “reasonable” fidelity. However, conversion is not quite as simple as just pushing a button.

DATA COMPRESSION TYPES

The most popular audio data compression methods are:

MP3: High quality modes sound quite good, but as with all data compression algorithms, better sound requires a bigger file. MP3 is the most common audio file protocol for the web, and there are many free or shareware MP3 players and encoders. Most digital audio editors, and even some sequencers, include a “Save as MP3” command or accommodate an MP3 encoding plug-in.

AAC: As the iPod’s native file format, this format is pretty popular — and to most people, AAC sounds better than MP3 given the same file size. Apple iTunes, a free download for Windows or Mac, allows converting to AAC.

Windows Media Audio. This is Windows’ native compressed format, and while not as common as MP3, is popular among Windows users. Microsoft no longer supports a media player for the Mac, but Flip4Mac allows playing Windows Media audio and video formats, and exports in several Windows-centric formats. As with AAC, most people think WMA sounds better than MP3.

FORGET THE DETAILS

Data compression is not about subtlety. Start with a signal that has a high average level (that doesn’t mean it needs to be heavily squashed; a little peak limiting does most of what you need). Before encoding, remove any DC offset, then apply one of the many tools for creating a hot level (e.g., Waves’ level maximizing plug-ins, iZotope’s Ozone, etc.). Any of these increases the perceived loudness; just don’t go too nuts, and please, don’t master your CDs this way.

CONVERTING WITH iTUNES

In iTunes’ menu bar, go to iTunes > Preferences > Advanced, then click on the Importing tab. Under “Import Using,” you can choose among AAC, Apple Lossless Compression, MP3, AIFF, and WAV. The latter two are non-compressed formats.

Under the “Setting” drop-down menu, click on “Custom.” You can choose from Bit Rate (for MP3, 16kbps to 320kbps), Sample Rate (8kHz to 48kHz and Auto, which uses the same sample rate as the original file), Channels (mono or stereo), and Stereo Mode. For Channels, Mono gives better quality sound than stereo because the bandwidth is “spent” on only one channel, whereas with stereo, that bandwidth is split over two channels with lower quality.

Stereo Mode is mostly appropriate for bit rates under 128kHz, which you don’t really want to use anyway.

Checking “Variable Bit Rates” can improve a file’s quality without increasing file size, by using a higher bit rate only when needed. For the most universal compatibility, leave this unchecked. Still, most MP3 players can handle variable bit rate files.

I leave “Smart Encoding Adjustments” unchecked; this analyzes your music and changes the settings based on what it thinks is best, and I generally prefer to customize settings. Finally, always check “Filter Frequencies Below 10Hz.” You won’t hear them, and they just take up bandwidth.

TIME TO EXPORT

Now that you’ve set up your compression parameters, it’s time to export. Select your file in iTunes, then go to Advanced > Convert Selection to MP3 (or whichever protocol you chose). This creates a new entry in iTunes. Drag it to the desktop, then do what you want with the file. That’s all there is to it!

Jargon Jockey


. Data Compression: : Audio requires over 10MB per minute for CD-quality sound (stereo, 44.1kHz, 16-bit). To accommodate bandwidth and memory limitations, data compression, more accurately called “data omission,” algorithms drastically reduce the amount of data needed to convey music. (Note that this is different from dynamic range compression, i.e. when you run an audio signal through a compressor.) If the sound levels are very high, the algorithm might assume you can’t hear lower-level material, and decide that those sections need only about 24dB of dynamic range. This requires 4 bits of resolution, which uses only 25 per cent of the storage space required for 16-bit resolution. Different data compression algorithms have different sonic characteristics, and different tradeoffs of quality vs. file size.

 

Keyboard Magazine is part of the Music Player Network.