Williams CVSD Speech Info

I have found two datasheets that deal with the CVSD process.

There is software that will take the raw CVSD data and convert it into many types of output files. One is called AudioToolbox and you can check that out at http://vinfo.com/ToolBox.htm

The AudioToolbox actually works but it is rusty. Or maybe Im just rusty and don't know how to use it yet. Anyway, I took some speech data from Jungle Lord and converted it to .wav. It did work but is very hard to understand if you don't know what it is saying. There are two samples below that were converted into .wav files. I messed around with AudioToolbox for awhile and found that using 15KHz for both the input and output files worked the best so far. It is still pretty rough around the egdes.

Im wondering some things about Williams Early CVSD data..

What sample rate is most of the speech data? It seems like 24KHz is a Standard but Im not sure, Im using 15KHz since it seems to give the best quality so far.
Will reading the WMS speech data in at 24KHz affect the quality of the speech or just make it higher in pitch? Im getting some weird reverb or delay in my decodes.

If you are interested in this or have any info, please drop me an e-mail