Hi @Paz Elisha,
Thank you for your follow up query.
To ensure Hebrew TTS pronounces English letters correctly without switching voices, you can use SSML phonemes to specify how each letter should sound, keeping the same Hebrew voice. For example, <phoneme alphabet="ipa" ph="piː">P</phoneme>
ensures "P" is pronounced as "Pee." Alternatively, you can use pauses between letters with <break>
tags or adjust the speech rate using <prosody>
to emphasize each letter. If these approaches don't fully meet production quality, consider training a custom neural voice or post-processing the audio for seamless output.
I hope you understand! Thank you.