previous next

Chapter 3: Producing Audio

RealSystem can stream many types of audio in addition to the RealAudio format. Not all audio files stream equally well, however. This chapter describes the types of audio files you can stream, explaining how to prepare or encode your files. It also provides tips for capturing high quality audio source.

Recording High Quality Audio

The following are tips for creating high quality audio source files. Although geared for RealAudio, these guidelines will help you no matter which audio format you stream.

Additional Information
For pointers on video, see "Producing High Quality Video".

Use Professional Recording Equipment

Every piece of equipment in the audio chain, from the microphone, to the sound card, to the sound editing software, affects sound quality. If you intend to provide commercial audio content, invest in professional audio equipment and software. Poor quality equipment can add hiss and distortion, degrading sound clarity.

Use Quality Source Files

If you are not recording your own sound, be sure to use high-quality audio source files. Use sources from CD or DAT, for example.

Digitize Sound before Encoding

If you are not broadcasting live, capture or "digitize" the sound to a supported file format such as a WAV, QuickTime, or AIFF whenever possible. Digitizing the sound before encoding the file allows you to use a sound editor to adjust the signal amplitude to maximize the available dynamic range. If you do not adjust the signal, the resulting streamed files may sound flat.

Minimize Source File Size

Keep the audio source file as small as possible. This makes it easier to encode the file in a streaming format. Cut any unnecessarily long silences from the beginning or end of the file to conserve space.

Set Input Levels Correctly

Setting correct input levels is crucial. All audio equipment has a signal-to-noise ratio, a ratio between the loudest possible sound the equipment can reproduce without distortion and its inherent noise. To work with the loudest input sounds possible, set the input level to use the full range of available amplitude without distortion. This distortion is known as "clipping," and is audible as a high frequency crackling noise.

When digitizing with a sound card, do several test runs and adjust levels on the mixer page of your sound card utilities so the input approaches but does not exceed the maximum. Most mixer pages graphically display input. Make sure there are no peaks above maximum. Be conservative, though. You never know when someone will get excited and speak much louder, or when a crowd at a sports event will roar.

Sound files that do not use the full amplitude range will produce low quality streaming files. If the amplitude range is too low, use your sound editor's Increase Amplitude or Increase Volume command to adjust the range before encoding the file. Most sound editors have a Normalize function that maximizes levels automatically. However, you get better quality if you set the levels correctly when recording.

Eliminate DC Offset

Eliminate DC offset either while recording content or later with an sound editor. This removes low frequency noise.

Equalize Frequencies

Equalization (EQ) changes the tone of the incoming signal by "boosting" (turning up) or "cutting" (turning down) certain frequencies. Using EQ, you can emphasize frequencies you want and cut frequencies that contain noise or unwanted sound. In addition, EQ can compensate for RealAudio Codecs that do not have flat frequency responses (that is, Codecs for which certain frequencies are not as loud after encoding).

Normalize Audio Files

As the last step before encoding the file, normalize the source file to 95% of the maximum sound volume. This lets you feed your encoding tool the loudest distortion-free files possible. If your machine's normalization option does not let you specify a percentage, turn down the overall volume after you normalize by using your software's Volume or Amplify option.

Prevent Clipping

If your original audio file signal exceeds the acceptable amplitude range, the file may be "clipped." Clipping can give rise to clicks or pops on playback. If your source file contains a clipped signal, your streaming file may have high-frequency background noise or static. Lowering the input volume helps reduce clipping.

Prepare Volume Levels for Live Broadcasts

When broadcasting live audio, you have less opportunity to manipulate the input signal. Be sure that volume levels are prepared and tested before encoding live input.

Producing RealAudio

RealNetworks pioneered streaming audio with RealAudio, the first streaming media product for the Internet. Since its debut in 1995, RealAudio has become the standard for network audio, delivering stereo sound over 28.8 Kbps modems, with near-CD quality sound at ISDN and LAN speeds. RealAudio files use the file extension .ra.

Audio Input Formats

RealAudio is a compressed format suitable for streaming over low to high network speeds. Because RealAudio is compressed, you typically start with a sound file in a digitized, uncompressed format such as WAV or AIFF. You then create a RealAudio file from this source file through an encoding tool. Your encoding tool should be able to accept some or all of these input formats:

Using RealAudio Codecs

RealAudio uses a "lossy" compression scheme that disregards parts of the audio source file to achieve a highly reduced file size. A RealAudio file encoded from a WAV file, for example, is typically smaller than the WAV by a factor of ten or more. Although discarding audio information during encoding lowers the file's frequency response and dynamic range, carefully choosing Codecs minimizes the impact of compression.

An encoding tool uses a Codec to compress the original sound file and create a RealAudio file. RealPlayer uses the same Codec to decompress the streamed RealAudio file for playback. When you encode a RealAudio file, you choose a Codec (or series of Codecs) based on two criteria:

  1. Bandwidth

    As "Targeting Bandwidth" explains, you need to decide how much bandwidth each part of your presentation will consume. When you have a bandwidth target for your audio component, you can choose a Codec that encodes RealAudio at or below that target.

  2. Audio Content

    RealNetworks provides different Codecs for music and spoken voice. Voice Codecs, for example, focus on the standard frequency range of the human voice. Music Codecs have broader frequency ranges to capture more of the high and low frequencies.

Choosing a RealAudio Codec

When you encode your audio file with a RealNetworks encoding tool, the tool selects the Codec or Codecs automatically. You simply set your bandwidth targets and some other parameters, such as whether you want earlier versions of RealPlayer to be able to play your presentation. The tool then selects the best Codec or Codecs to use based on your input.

Additional Information
Refer to the manual or online help for your RealNetworks encoding tool for step-by-step instructions on how to encode a RealAudio file.

The following table lists all RealAudio Codecs. It is provided as a general reference. Check your RealAudio encoding tool for the Codecs it supports. See below for information about interpreting the table.

Audio Codec654321Sample RateResp.Comments
Low Bandwidth Audio Codecs
5 Kbps Voice X X - - - - 8, 16, or 32 kHz 4 kHz
6.5 Kbps Voice X X X - - - 8, 16, or 32 kHz 4 kHz
8 Kbps Voice X X X X X X 8, 16, or 32 kHz 4 kHz Superseded by 8.5Kbps Voice Codec.
8 Kbps Music-G2 X - - - - - 8, 16, or 32 kHz 4 kHz Generation 2 Codec. Use with multiply encoded files.
8 Kbps Music X X X - - - 8, 16, or 32 kHz 4 kHz DolbyNet Codec.
8.5 Kbps Voice X X X - - - 8, 16, or 32 kHz 4 kHz
11 Kbps Music-G2 X - - - - - 11.025, 22.05, or 44.1 kHz 5 kHz Generation 2 Codec. Use with multiply encoded files.
12 Kbps Music X X X - - - 8, 16, or 32 kHz 4 kHz DolbyNet Codec.
Medium Bandwidth Audio Codecs
15.2 Kbps Voice-RealAudio 2.0 Mono X X X X X - 8, 16, or 32 kHz 4 kHz Superseded by 16 Kbps Voice Codec.
16 Kbps Voice-Mono Wideband X X - - - - 16 or 32 kHz 8 kHz Highest bit rate Codec for voice.
16 Kbps Music-G2 Mono X - - - - - 22.05 or 44.1 kHz 8 kHz Generation 2 Codec. Use with multiply encoded files.
16 Kbps Music-Mono Low Response X X X - - - 8, 16, or 32 kHz 4 kHz DolbyNet Codec.
16 Kbps Music-Mono Medium Response X X X - - - 11.025, 22.05, or 44.1 kHz 4.7 kHz Suitable for pop/rock music. DolbyNet Codec.
16 Kbps Music-Mono High Response X X X - - - 11.025, 22.05, or 44.1 kHz 5.5 kHz Suitable for classical music. DolbyNet Codec.
20 Kbps Music-G2 Mono X - - - - - 22.05 or 44.1 kHz 10 kHz Generation 2 Codec. Use with multiply encoded files.
20 Kbps Music-Stereo X X X X - - 8, 16, or 32 kHz 4 kHz DolbyNet Codec.
High Bandwidth Audio Codecs
32 Kbps Music-G2 Mono X - - - - - 44.1 kHz 16 kHz Generation 2 Codec. Use with multiply encoded files.
32 Kbps Music-Mono X X X - - - 11.025, 22.05, or 44.1 kHz 8 kHz DolbyNet Codec.
32 Kbps Music-Stereo X X X - - - 8, 16, or 32 kHz 5.5 kHz DolbyNet Codec.
40 Kbps Music-Mono X X X X - - 11.025, 22.05, or 44.1 kHz 11 kHz DolbyNet Codec.
40 Kbps Music-Stereo X X X X - - 8, 16, or 32 kHz 8 kHz DolbyNet Codec.
80 Kbps Music-Mono X X X X - - 11.025, 22.05, or 44.1 kHz 20 kHz DolbyNet Codec.
80 Kbps Music-Stereo X X X X - - 8, 16, or 32 kHz 16 kHz DolbyNet Codec.

The preceding table includes the following information:

Notes on RealAudio Encoding

The following sections give tips for encoding RealAudio files.

Multiple Encoding in a Single File

You can create a single file encoded for multiple bandwidths only with the Codecs introduced in RealSystem G2. In the preceding table, these Codecs are marked as playable only by RealPlayer 6.0. To support multiple bandwidths when encoding with other Codecs, you must encode a separate file for each Codec. You then use a SMIL file to specify bandwidth choices.

Additional Information
For more on bandwidth selection through SMIL, see "Setting Bandwidth Choices".

Backwards Compatibility with Earlier Versions of RealPlayer

When you use a RealNetworks encoding tool to encode a file for multiple bandwidths, you can specify backwards compatibility with earlier versions of RealPlayer. The tool encodes the file for your selected bandwidths with the G2 Codecs. It also includes in the file an encoding at your lowest bandwidth selection that uses an older Codec.

For example, you can encode a single file at 8, 16, and 32 Kbps using the G2 Codecs. In the encoding tool, you can choose backwards compatibility to create an additional 8 Kbps stream with an older Codec. Depending on its connection speed, RealPlayer 6.0 receives the 8, 16, or 32 Kbps G2 stream. Earlier versions of RealPlayer receive the 8 Kbps stream encoded with the older Codec regardless of their connection speeds.

Streaming Other Audio Formats

RealSystem can stream audio formats other than RealAudio. Because these formats may not be as highly compressed as RealAudio, they may not be good for low bandwidth connections. The following sections give you guidelines for preparing files in these formats. This information will help you decide if you should stream from the native format or convert the file to RealAudio.

Note
The following sections will be added later.

AU

WAV


Copyright © 1998 RealNetworks
For technical support on RealNetworks products, visit http://service.real.com/.
Comments on this document? Contact techpubs@real.com.
This file last updated on 05/20/98 at 16:51:54.
previous next