When the default text-to-speech program (spd-say) speaks it is accompanied by noise/crackling/buzzing (and some echo), how to fix?
Solution 1:
It seems to be a bug in the combination of speech-dispatcher with pulseaudio. As a workaround you can set
AudioOutputMethod "libao"
in /etc/speech-dispatcher/speechd.conf.
(Source: https://bbs.archlinux.org/viewtopic.php?id=215987)