Using the CopiaFacts Voice Algorithm Files
CopiaFacts comes with sample voice algorithms for speaking values either as discrete digits or as currency amounts in various languages. There are also examples of algorithms to speak time and date values in English. In all cases, we leave the actual voice phrases for you to speak using a voice which matches your other voice messages. The sample algorithm source files include a list of suggested voice filenames with the text of the word or phrase that needs to be spoken for each file.
Once you have recorded all the required words and phrases into separate voice files, the algorithm compiler places them into one combined multi-segment file, together with the compiled algorithm. Each combined file therefore carries with it the knowledge of how any value passed to it is to be interpreted and how the resulting sentence is built. All you have to do in a question-box IIF is specify the value and the filename with a $play_var command. This command can be used instead of $image_desc in an image IIF or in a type decision or voice IIF to play any variable.
You can also replace the standard FAXFACTS.DIG and FAXFACTS.AMT files, for example, to incorporate a different algorithm for a language other than English. Normally, the you have entered... message is spoken before the digits themselves, but CopiaFacts does not need to be changed or reconfigured in any way if you replace FAXFACTS.DIG with the German version which has the verb at the end of the sentence. This detail is held in the file itself.
1.The procedure for creating a new multi-segment voice file is as follows:
2.Select the algorithm source file that most closely matches your requirements.
3.Create a working directory and copy the algorithm source file to it.
4.Modify the algorithm if necessary.
5.Run FFAMOUNT -t sourcefile to test the algorithm with various different inputs. This simply displays the text from the source file as it would be spoken. Once you are satisfied that the output is correct, it only remains to record voice segments matching the texts and assign the correct filenames to them.
6.Print out a copy of the algorithm source file so that you have a reference list of all the voice segments and their contents.
7.Prepare all the individual voice segment files using the CopiaFacts Recording Studio feature. Use a command $type studio amount with the name of the working directory and algorithm file. The phrases to speak will be automatically displayed on the screen as prompts.
8.Run FFAMOUNT -c sourcefile to compile the algorithm and merge the component voice files. The output file name will be as specified in the algorithm source file.