1These wav files show how Sonic performs at increasing speech rates. All sound 2sampels are in the public domain. 3 4sonic.wav 5This is a sonic 2X sped-up version of a public domain librivox.org recording, from 6the audiobook "Princess of Mars". 7 8soundtouch.wav 9This is the same recording as sonic.wav, but sped up using soundtouch, which 10uses WSOLA rather than the sonic algorithm. Even at 2X speed up, you should be 11able to hear the characteristic WSOLA distortion relative to the sonic version. 12 13talking.wav 14This is my father talking, using a decent microphone and 44KHz sample rate. 15 16talking_2x.wav 17This is his voice sped up by 2X using Sonic. 18 19espeak_s450.wav 20Sonic also performs well at increasing the speed of synthesized speech. 21espeak_s450.wav was generated using 'espeak -s450 -f test1.txt -w 22espeak_s450.wav'. This is the highest speed currently supported by espeak, 23though Sonic can speed up espeak to much faster rates. 24 25espeak_sonic.wav 26This was generated with 'espeak -f test1.txt -w out.wav; 27sonic 2.6 out.wav espeak_sonic.wav'. Sonic sped it up 2.6X, which is about the 28same speed as espeak at -s450. I personally feel that the sonic sped up sample 29sounds better than espeak at -s450. 30 31twosineperiods.wav 32This is just two sine periods, which is too short to hear. However, it's 33useful for making sure the flush function works correctly. A 2-X speedup should 34result in one sine period with no distortion. 35