Equalization


Equalization (mentioned briefly in Chapter 2) is a process that allows you to change the overall sound of your audio file by raising or lowering one or more frequencies anywhere along the frequency spectrum of your audio. Equalizers come in a variety of flavors, including the familiar "bass" and "treble" tone controls found on a typical consumer stereo system.

Equalization's primary purpose in the streaming audio environment is to help your audio translate more clearly after being encoded. By adjusting small amounts of boost or cut, you can change the character of the sound in a way that allows the encoder to do a more efficient job. Increased encoder efficiency generally translates into increased sonic legibility.

When encoding to low bit rates, equalization can be especially useful. The lossy perceptual encoding process functions by throwing away portions of the overall sound that a codec's engineers have decided the human ear isn't likely to miss. The lower the bit rate, the more sound that is permanently thrown out (hence the term lossy). The encoder analyzes the audio from a technical standpoint, so it can't make qualitative judgments based on the audio's actual content. You know more about what frequencies might be unnecessary than the encoder does. By minimizing specific frequencies before handing your source audio to the encoder, you make a decision based on a human assessment. For example, if you're encoding a classical cello suite, the lower frequencies are going to be far more important than the high end. Conversely, a voice-only interview doesn't need the low frequencies. Without optimization on your part, the encoder makes a "best guess" and discards what it thinks isn't necessary.

Think about the application of equalization as a process of taking away what you don't want (or can't use) as opposed to adding more of what you think you need.

You can perform these adjustments using the basic built-in equalization functions in your waveform editor, higher-quality software plug-ins, or a standalone hardware equalizer. Equalizer interfaces are typically one of two types. A graphic equalizer has individual sliders at multiple frequencies, allowing for overall shaping of the signal across the full frequency spectrum. A parametric equalizer allows for fine-tuning a handful of configurable areas within the frequency spectrum. Graphic equalizers are easier for beginners to use, and parametric equalizers are very effective when controlled by advanced users. A Web search on equalizer or equalizer plug-ins will detail a wide array of both hardware and software products.

A first simple step when encoding to streams less than 64Kbps is to remove both the very low (40Hz and below) and very high (15kHz and above) frequencies. Most of your low bit rate listeners are probably using inexpensive computer speakers and likely won't miss these frequencies. When encoding for 56K modem users (and below), experiment with a slight boost in mid frequencies (around 2.5kHz) to compensate. Spoken word encoding at any bit rate can especially benefit from this treatment.

For low bit rate encoded voice-only files try this starter setting: Cut everything below 100Hz and carefully boost frequencies between 1 4kHz.

Here's a reminder about high bit rate encoding: Usually little (or even no) equalization is necessary. Recent codecs do a pretty good job of translating the sound quality of source audio when encoding at high bit rates. Heavy equalization of professionally mastered audio (such as major label pop music CD releases) can easily do more damage than the encoding process.

Of course, strange things do happen in the encoding process at bit rates below 16Kbps. You might find that you need to boost certain frequencies that aren't obvious in the source audio. These results will vary on a codec to codec basis. Experiment!



Streaming Audio. The FezGuys' Guide
Streaming Audio: The FezGuys Guide
ISBN: B000H2N1T8
EAN: N/A
Year: 2001
Pages: 119

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net