Replies: 1 comment 4 replies
-
if __name__ == "__main__":
from RealtimeTTS import CoquiEngine, TextToAudioStream
def dummy_generator():
yield "Hey guys! These here are realtime spoken sentences based on local text synthesis. "
yield "With a local, neuronal, cloned voice. So every spoken sentence sounds unique."
source_voice = "my_voice_reference.wav"
coqui_engine = CoquiEngine(voice=source_voice)
stream = TextToAudioStream(coqui_engine)
stream.feed(dummy_generator())
stream.play(output_wavfile=stream.engine.engine_name + "_output.wav")
coqui_engine.shutdown()my_voice_reference.wav should be 22050 Hz sample rate, mono and 16 bit. |
Beta Was this translation helpful? Give feedback.
4 replies
-
|
It worked perfectly, it works much better for realtts, the way it was created makes it easier to implement, thank you for your help and succes |
Beta Was this translation helpful? Give feedback.
-
|
@KoljaB thanks for sharing the example code! Unfortunately I get an error saying it can't find a valid output device: I tweaked it slightly to enable logging and point to my voice file/output paths, so this is the script I'm running: if __name__ == "__main__":
import logging
from RealtimeTTS import CoquiEngine, TextToAudioStream
def dummy_generator():
yield "Hey guys! These here are realtime spoken sentences based on local text synthesis. "
yield "With a local, neuronal, cloned voice. So every spoken sentence sounds unique."
level = logging.DEBUG
debug = True
print("Starting the test script")
source_voice = "voices/female_01.wav"
coqui_engine = CoquiEngine(voice=source_voice, level=level)
stream = TextToAudioStream(coqui_engine, level=level)
stream.feed(dummy_generator())
stream.play(output_wavfile="outputs/" + stream.engine.engine_name + "_output.wav", debug=debug)
coqui_engine.shutdown()I'm running it on WSL2 (Ubuntu), so it makes sense there's no audio output devices available. But when it's outputting to a file it shouldn't need an output device right? Is there some parameter we could pass to disable this check? Thanks again! |
Beta Was this translation helpful? Give feedback.
-
|
List all devices using this script. Check the output for a suitable device and use its index as the stream = TextToAudioStream(
coqui_engine,
level=level,
output_device_index=<your_output_device_index> # Replace <your_output_device_index> with the actual index
)If you want to skip playback entirely and avoid opening an output stream, set stream = TextToAudioStream(
coqui_engine,
level=level,
muted=True,
)
stream.feed(dummy_generator())
stream.play(
output_wavfile="outputs/" + stream.engine.engine_name + "_output.wav",
debug=debug,
muted=True
) |
Beta Was this translation helpful? Give feedback.
-
|
Thanks so much for your super quick reply! Setting |
Beta Was this translation helpful? Give feedback.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hello, the documentation states that it is possible to upload an audio of 5 to 30 seconds to use as a model during a generation of tts to wav but how do I send this to coqui? using realtts?
I tried to use some coqui variables in example-fast_api but I was unsuccessful, do you have a simple example? for me to test the source file of the text voice, generate a new file with the voice of the source file
Beta Was this translation helpful? Give feedback.
All reactions