Meridian America Link/Customer Controlled Routing manual Recording and trimming voice segments

530Chapter 18: Acceptance testing

Recording and trimming voice segments

In Edit Voice, each voice segment has a title and text associated with the recording. In general, you will want to create the new file, and then create the segments in the order you plan to use them.

When creating voice segments, enter the title and text before you start recording. You can then record with the text on-screen in front of you, so you are sure that you wonÕt forget any part of your planned prompts.

ItÕs almost impossible to record without leaving some silence at the front and end of the recording, so you would normally use the trimsilence command to trim the ends of a recorded digit, thus creating a more natural- sounding message.

Use the Edit Voice Trim Silence facility to cut blank space from prompts that will be concatenated. For example, if you want to record the digits ÒoneÓ and ÒtwoÓ for playback in a sequence of digits, you will record Ò...one...Ó and then record Ò...two....Ó When you hear digits recorded this

way played back in sequence, it sounds like Ò...one......two....Ó However,

after using the Trim Silence feature, you would hear Òone..twoÓ instead. It is best to run Edit Voice in a quiet environment, since background noise will prevent you from using Trim Silence.

Creating a header file

If you define the segments the same way both in the application and in the voice segment file, you will not need a header record.

When you save the file, you can optionally use the editvoice command to write a ÒCÓ language header file to associate the prompt titles you created with the segment number they are recorded in. This file may be moved to tape with the backfiles command and transferred to your host computer from the tape if your host is compatible with the Motorola UNIX System V bru tape format.

553-3202-210 Standard October 1998