Below the script I used to generate a bunch of wav files with different text to speech applications.

#!/bin/bash

string="This is HPR episode 2792 entitled "Playing around with text to speech synthesis on Linux" and is part of the series "Sound Scapes". It is hosted by Yeroon Bahten and is about 20 minutes long and carries a clean flag."

echo "${string}" > text.txt

espeak -w espeak.wav "${string}"

espeak -w espeak-ng-v-mb-us1.wav -v mb-us1 "${string}"

espeak -w espeak-ng-v-mb-us2.wav -v mb-us2 "${string}"

espeak -w espeak-ng-v-mb-us3.wav -v mb-us3 "${string}"

espeak-ng "${string}"

espeak-ng -v en-gb "${string}"

espeak-ng -w espeak-ng-en-gb-scotland.wav -v en-gb-scotland "${string}"

espeak-ng -w espeak-ng-en-us.wav -v en-us "${string}"

flite -o flite-voice-cmu_us_slt.wav -voice cmu_us_slt "${string}"

echo "${string}"| festival --language english --tts # same as next line

echo "${string}"| text2wave --language british_english --tts -o festival_british_english.wav

text2wave -o festival_british_english.wav text.txt

for voice in don_diphone kal_diphone ked_diphone rab_diphone

do

text2wave -o festival_voice_${voice}.wav -eval "(voice_${voice} )" text.txt

done

# Gnustep say, recorded with audio recorder.

say "${string}"

text2wave -o festival_voice_cmu_us_slt_arctic_hts.wav -eval "(voice_cmu_us_slt_arctic_hts )" text.txt

# merlin https://github.com/CSTR-Edinburgh/merlin

# marytts: https://github.com/marytts