| TalkBank | Audio Digitizing Process |
|
|
|
Warning: Please avoid use of compressed digital formats such as MP3, since it is impossible to obtain good linkage of compressed audio formats to transcripts. What you will need:
Connecting the output source to the computer: The first step in digitizing a sound file is to connect the audio source (DAT, casette, minidisk, or reel-to-reel) to the computer. We will refer to the input audio source as a tape recorder. However, DAT, casette, minidisk, and reel-to-reel all work in essentially the same way. The connection is usually done by connecting a mini audio cable from the headphones output on the tape recorder to one of the input lines on the computer or sound card. For better sound quality we used to recommend going through the line output and not the headphones. However, newer sound cards make it possible to use the headphone output. If you are able to get a good level of audio input from the headphones jack, you can skip reading the next section. Using a mixer: Some combinations of older hardware (often reel-to-reel and sometimes DAT) and older sound cards cannot make good use of the output from the headphones jack of your audio source. In these cases, you need to use the line output from the audio source. The problem with using the line output is that you often cannot control output level or volume (although some audio sources have a control for this). If the output level of the line output of your machine does not match that required by the computer, you will overdrive the input and get bad results. Newer sound cards seem to have solved this problem and it is usually possible to use line input for recording. If this situation arises, you will need to use a device to control line output level. Another reason to have a device to control line output level is that, if the tapes are of poor quality or contain a large range of volume levels you will need to adjust the sound manually at various points in the tape in order to prevent clipping. The best solution is to use a mixer from places like Radio Shack or Full Compass Audio (800 356-5844). You can get miniature preamps, some with mikes built-in from www.microphonemadness.com. Audio technicians tell us that mixers are better than amplifiers. However, we have not noticed an audible difference. Running the tape recorder through an amplifier allows volume control while maintaining the sound quality. Setting up your connections in the simplest case
Setting up your connections when you have an amplifer or a mixer
Sound Settings Sounds settings for your card change with each version of the operating system, so we cannot give instructions that will be valid for all users. Basically, you just have to read your computer manual for this. Make sure that you have the input and output settings adjusted to tape input from the external audio source and to play back to speakers or headphones if you want to monitor the recording. Using Sound Edit 16 and CoolEdit We describe the recording process here for SoundEdit 16. CoolEdit and Sound Studio work in much the same way. For CoolEdit, try to ignore the complex interface and just focus on the recording buttons down at the bottom left of the screen. For SoundEdit 16, recording will default to 44,100 Hz, 16 bits, mono, with no compression. This is OK for recording. However, you will probably want to save the file at 22,050. Please use the WAV format for compression. This is the standard for CHILDES and TalkBank media. AIFF and AIFC are also acceptable. MP3 is not acceptable, because compressed audio files cannot be accurately linked to transcripts. The Controls window The Controls window in Sound Edit 16 is used to record sound, play a sound file and stop or pause the recording. You can also use the controls pull down menu at the top of the screen for the same options.
Monitoring Sound levels The Levels window allows you to monitor and control the recording and playback levels in Sound Edit 16. For the best results, in the recording window, make sure to set the L/R volume to 16. This setting is recommended by Sound Edit for obtaining the best over all sound quality.
If a red dot or red block appears in the recording levels area the sound is being clipped. Clipping is defined by Sound Edit as the amplitude of a sample exceeding the quantization range. For example, if a child is sitting to close to the microphone and starts to scream the top and bottom of the wave form will be cut off. This results in a very poor digitized sample. In order to control for clipping, adjust the volume of the source (i.e., amplifier, mixer, or tapecorder) until the red light or dot no longer appears in the Recording level window. You can also control the volume using the input or volume controls in the Recording level window. Sound Edit does not recommend this method because the sound quality may be sacrificed. Creating a new sound file
|