Azure-Samples/Cognitive-Speech-TTS

Pronunciation Error: Spanish MX Neural Voice - Jorge

inglesuniversal opened this issue · 4 comments

inglesuniversal commented

I would like to report that the way the word VIDEO is pronounced for the neural voice named Jorge is not totally accurate.

Your system stresses the word VIDEO sounding like ----> VEE - de - oh (on the first syllable)

instead of ----> vi - DEH - oh (on the second syllable)

If you require an audio sample, I would be more than happy to share it with you.

szhaomsft commented

thanks for reporting.

we will take a look

ForrestGumb commented

This issue is fixed. Please have a try.

  • 👍1
inglesuniversal commented

Awesome... I just checked and YES in deed... It works beautifully! Thanks for following up on my request. Good luck!

CORRECTED DEMO AUDIO SAMPLE:

synth_fixed_es-mx_Jorge.mp3

Jun711 commented

@inglesuniversal @szhaomsft @ForrestGumb
Could you share an example on how to use Spanish neural voices to read accented words? Thank you
The audio I get doesn't pronounce accented characters.

This is the issue that I opened.
#225