Streaming Media Bible - Excerpts from Chapter 8

Excerpt from Chapter 8 - Optimizing Your Audio Files


	Table of Contents

	Chapter Excerpt

	Updates

	Order the Book

	Resources

	Tips & Tricks

	Streaming Magazine Column

	Feedback

	Author

	Home

This chapter takes a closer look at audio files and discusses a few simple techniques to make them sound much better. If you've been careful up to this point, you should now have a system set up to record and capture good, clean audio signals. But perhaps upon playing them back, you felt that they still lacked something -- that little extra oomph that broadcast feeds have. With any luck, you're about to give your files that extra polish that will make them sound as professional as your local radio station.

As an extra-added benefit, the steps you're taking now to produce professional sounding audio are very beneficial when you get to the encoding stage. As powerful as encoders are, they can only work with what they are given. The higher quality the input, the higher quality streaming file you'll end up with.

Signal Processing

One of the wonderful things about working with audio is that it is very malleable. Through the judicious use of equalization (EQ), compression, noise reduction, and a host of other tricks, you can make an audio file recorded in a small closet sound as if it was produced in a professional broadcast studio. These audio tricks are generically referred to as signal processing and are employed by audio engineers worldwide.

Earlier in Chapters 4 and 7, the importance of setting correct input and output levels for each piece of equipment in your audio chain was discussed. This is done to maximize the performance of each piece of equipment and to prevent them from adding any noise to the sound. Now that you've recorded a clean audio file, you're going to be actively molding the sound to achieve the desired result: professional-sounding audio.

Cross Reference

If you'd like to review or find out more about setting the input and output levels of your equipment, please refer to Chapter 4. For a discussion about good recording practice, please refer to Chapter 7.

At its most basic, all audio signal processing essentially boils down to one thing: volume, or what audio engineers refer to as level. In dynamics processing (the most common being compression), you'll be adjusting the overall level of the entire audio file. With equalization, you're going to be boosting or attenuating the levels of certain frequencies in your sound files. At each stage of signal processing, you have to be careful with your levels so as not to exceed the maximum level allowed and produce distortion. You also have to make sure that each step is an improvement and that the resulting file is faithful to the original author's intent. Signal processing is extremely powerful, and should be used with caution.

Why Signal Processing Is Important

Signal processing is important because to stand out in a sea of content, your programming has to sound as good or better than your competition. Commercial television and radio stations invest millions of dollars to ensure that their programming meets rigidly defined standards, giving you the quality you expect. On the Internet, this is not the case. There are no standards, and quality varies dramatically.

One of the most powerful features of the Internet is that it makes it possible for just about anyone to become a broadcaster. But this blessing is also a curse: There are already thousands upon thousands of people creating programming for the Internet every day. Ignoring the actual content for a moment, the sonic quality of these programs ranges from excellent to unbearable.

No doubt you've stumbled upon a Web site featuring your favorite music or sports team. You click with baited breath only to find that the audio quality is terrible. Maybe you've been startled out of your chair when a new stream is twice as loud as the last thing you listened to. The problem is that many people who are creating streaming media for the Internet have little or no previous experience with audio or video.

The good news is that because audio is so easy to work with, there is no reason why your audio should not sound as good as your favorite local FM radio station. Although they have no doubt invested a lot of money to ensure that their sound quality is top notch, there is no reason that you can't achieve the same high standard with a minimal investment of time and money. By producing your content to a high standard, you give your programming an edge over your competitors.

All the signal processing techniques described in this chapter are based on solid audio engineering practice. These techniques will make your programming sound better regardless of the eventual streaming format, whether it is an audio-only radio station, a soundtrack to a video presentation, or your latest musical composition available for download.

[ . . . ]

Compression

Compression is the most common form of dynamics processing (don't confuse audio compression with file size compression; that comes later in the encoding section). Think of it as an automatic volume control that operates according to rules you can specify. In most cases, you want to turn the loudest sections of the audio down. You can do this by specifying a "threshold" above which you want the volume attenuated. When you do this you are compressing the dynamic range.

The dynamic range of a program is the difference in volume between the loudest and quietest parts of your audio (see Figure 8-4). Because you are turning down the loudest sections, the measured dynamic range is now smaller. It has been compressed.

Figure 8-4: The dynamic range and headroom of a file

After the dynamic range has been compressed by turning the loud sections down, you are left with extra headroom, the space between the maximum allowable volume before distortion and the loudest sound in your file. When you initially record your file, you try to leave about 3 dB of headroom. After compression, you should turn the overall volume of the file back up to restore the original headroom.

Why Compression Is Important

Compression is important for a number of reasons. First, compression is a good preventative tool because, by turning down the loudest sections of your file, you are taking steps to ensure that the signal level never adds distortion.

Secondly, audio equipment doesn't usually perform optimally at the extremes of its dynamic range. This is particularly true for multimedia systems. Between cheap soundcards and the fidelity of most multimedia speakers, you're operating with a fairly limited dynamic range. It is therefore necessary to compress the natural dynamic range of recordings in order for them to sound better on playback.

The last reason is harder to quantify, but essentially it boils down to this: All broadcast media is compressed to some degree, and most people tend to think of things that sound this way as sounding "professional." Commercial FM radio is the most obvious culprit. Try this: Turn your radio on, listen to a commercial music station, and then switch to a local community station. The main difference between the huge, full (sometimes crowded) sound of a commercial FM radio station and a small local station is the amount (and type) of compression used.

Compressed audio tends to sound louder and more present than uncompressed audio (see the "Why television commercials are so darn loud" sidebar). Since the hunter-gatherer days, people have been conditioned to respond to loud things and are easily fooled into thinking that louder is better. Ask any stereo salesman. The plain fact of the matter is that if someone else's programming sounds louder than yours, the listener usually thinks it sounds better.

All this adds up to one bold statement: audio compression is probably the single most important step you can take to improve the quality of your streaming content, audio or video. With the quality of audio codecs today, the proliferation of good audio software and reasonably priced audio hardware, there is no reason your audio should not sound as good or better than your favorite local radio station.