Hello,
While waiting for the next version of Maciulxochitl
try “REspeak”.
cz.martblek.REspeak_v01.xrnx (4.1 KB)
It is a tool for creating samples using the TTS engine Espeak.
Tested on Linux.
But perhaps it will also work on windows if espeak is installed in the path.
I don’t use Mbrola phonemes yet.
I wanted something offline and I don’t have Google TTS paid, we’ll see in time 
Got this when trying to generate a .wav in windows (without espeak installed I guess? I don’t see an option to specify a path in the tool gui, so when you say “the path” above, it is in the particular tool folder?)
"‘C:\Users\pluge\AppData\Roaming\Renoise\V3.4.0\Scripts\Tools\cz.martblek.REspeak.xrnx\main.lua’ failed in one of its notifiers.
Please contact the author (…)) for assistance…
std::logic_error: ‘trying to load sample data from a non existing file: ‘result.wav’’
stack traceback:
[C]: in function ‘load_from’
main.lua:459: in function ‘draw’
main.lua:317: in function main.lua:316"
I put the espeak windows version commandline.exe in the tools folder and have been able to generate sound, though immediately followed by this message;
"‘C:\Users\pluge\AppData\Roaming\Renoise\V3.4.0\Scripts\Tools\cz.martblek.REspeak.xrnx\main.lua’ failed in one of its notifiers.
Please contact the author (…)) for assistance…
std::logic_error: ‘trying to load sample data from a non existing file: ‘result.wav’’
stack traceback:
[C]: in function ‘load_from’
main.lua:459: in function ‘draw’
main.lua:317: in function main.lua:316"
No sample generated back into a renoise instrument.
edit:
also the tool gui’s parameters don’t seem to influence the previewed resulting sound. Dunno if I need to copy more then just the commandline.exe to the tools folder, also the config files etc?
hi,
“path” i mean set windows path variable. Path from where is programs executed.
viz here
I didn’t know if I could distribute the program in binary form
Ah ok, I’ve once had to do this for the CDP tool, but don’t think I’ll mess with these kind of settings anymore. Putting the commandline in the tools folder does let the tool process sound though, perhaps without setting the path variable it can still work?
hi,
I just looked into the downloaded installation for win.
There will be more problems.
I’m using linux so the paths and program names are written in the tool.
I’ll try to add an executable selection field.
The CDP & batchrender tools also make use of commandline executables & let you specify its whereabouts during install or in the gui, these tools also don’t come distributed with the commandline stuff, the user has to direct the tool to the correct path. Perhaps you can make this similar with REspeak? Would be cool to have this working
Nice. But I still prefer Vstspeek, where Talkany doesn’t fit.
@martblek This is really cool! I appreciate that you’re also using TTS that is not going to cause any issues with commercial usage. Another set of voices that are also available to use with no commercial restrictions (as far as I can tell) are MaryTTS voices.
Also PRAAT!
You’re coming up with some really neat tools these days, gotta say, I’m impressed.
Also DECTalk: https://dectalk.nitrix.me/
@Neuro_No_Neuro i moved tool to right forum.
DecTalk seems to be good.I want to try SSML with espeak,
but it needs few more knowledge about creating phonemes.
In the future i want implement creating your own voice into gui.
MaryTTS dont work for me coz some java errors 
As soon as this’ll work on Mac/Win, I’d love to know. I’m on Mac here. Would love to use something like this instead of just importing TTS .WAV files into Renoise.
Good luck and thanks for working on stuff like this!
Not working on Mac too ?
Yeah, same issue as others. I even installed eSpeak for Mac (using HomeBrew).
Error Message:
std::logic_error: ‘trying to load sample data from a non existing file: ‘result.wav’’
stack traceback:
[C]: in function ‘load_from’
main.lua:459: in function ‘draw’
main.lua:317: in function <main.lua:316