Technical Helpweb

- more articles

VOX and VAP files

VOX files are audio files used in speech applications. Usually they contain pre-recorded voice prompts. Because voice networks have relatively low bandwidth the audio used can be sampled at a fairly low rate (an ISDN "B" channel carries only 8kHz voice). Consequently, VOX files are often 8kHz (although 6kHz is also possible).

VOX files can contain audio in different formats. There are at least four different formats:

• Raw A-law
• Raw mu-law
• 8kHz 4-bit ADPCM (also known as IMA/OKI ADPCM)
• 6kHz 4-bit ADPCM

There is no file header inside a VOX file, only audio samples, so there is no way to know which format a particular file contains. If you play it and it sounds bad (i.e. distorted, wrong speed/pitch, or just ‘noise’), then you are not playing it back in the same way that it was recorded. 

VOX files have been long associated with the Dialogic^® DMV and JCT Media Boards, and customers often have collections of recorded prompts stored in "VOX format".  People get confused about this and often ask "how to I play/convert a prompt in VOX format?", but in fact VOX is not really a format, but simply a dump of sound samples that has been given a ".vox" extension.  An application cannot guess exactly what codec was used to record the samples, because that information is not stored in a file header, as it would be for a .WAV or .AU file.

VOX files can be played using the Dialogic^® Global Call / R4 family of APIs or by the Dialogic^® Diva^® Software Development Kit (SDK), but in both cases the application needs to specify the format and sample rate in order for playback on the line to be correct.  If you have old ".vox" files that need to be converted for a new application, then the first step could be to use a sound editor to read the file in, so that you can try to determine the codec that was used to encode the sound samples.  Many sound editors have an "Import Audio" feature that allows you to read in raw data from a file, while applying different parameters, for example: sample rate, sample size and codec.  Once you have found parameters that recover the files so that they "sound right", you can add these parameters to your application.

Other formats

Envox (a product of the Envox Group) uses VOX files, and also uses a variation called a VAP file. A VAP file is a number of 8kHz ADPCM fragments, packed into a single file with an index. This allows a number of related voice prompts (such as prompts in a particular language) to be kept in a file together and played back using the index.