TextToSpeech?

I am going to enter the Compo. If I write the dialog, is it OK to generate voices using software or do I need to do all the voices myself?