I see many people thinking that eSpeakNG only supports the default male voice, but it actually supports a multitude of voices accessible via the --voice parameter. In my opinion, with custom voices, it is quite capable of sounding clear and readable. It'll never be as natural as Google's cloud TTS, but I think it's a wonderful middlepoint. The default appears to be m1, which is of rather low quality among the preset voices IMO. Female voices are supported such as f1. There are also other non-gender-categorized novelty voices like announcer, robosoft, and zac which have various fun-sounding speech. There's plenty of other voices that aren't mentioned in here but can be found in the voices directory of espeakNG.
The above mentioned voices are just presets, though. More importantly, the voices are stored as a human-readable text file which can be easily understood. The format is here. I think that this could be used to allow users to personalize their own TTS by uploading a custom voice configuration file.
I see many people thinking that eSpeakNG only supports the default male voice, but it actually supports a multitude of voices accessible via the --voice parameter. In my opinion, with custom voices, it is quite capable of sounding clear and readable. It'll never be as natural as Google's cloud TTS, but I think it's a wonderful middlepoint. The default appears to be
m1, which is of rather low quality among the preset voices IMO. Female voices are supported such asf1. There are also other non-gender-categorized novelty voices likeannouncer,robosoft, andzacwhich have various fun-sounding speech. There's plenty of other voices that aren't mentioned in here but can be found in the voices directory of espeakNG.The above mentioned voices are just presets, though. More importantly, the voices are stored as a human-readable text file which can be easily understood. The format is here. I think that this could be used to allow users to personalize their own TTS by uploading a custom voice configuration file.