eSpeak uses an ASCII representation of phoneme names which is loosely based on the Kirshenbaum system.