Forum Discussion
Week 1 Videos
Generating voice-overs with AI has definitely been a game-changer for my team. One constant challenge I've run into is adjusting specific pronunciation (e.g., acronyms, industry jargon, etc.). In Rise, I find myself adjusting the source text to spell out certain words phonetically or even re-writing the script to get something that works. Are SSML tags supported in Rise? Any other tips would be welcome.
Hey MattRecio-a4144 thanks for calling this out.
When someone runs into a mispronounced word, I generally recommend starting by spelling the word phonetically, as you mentioned.
You can also try the V3 voice model, which is our latest model and offers the most accurate pronunciation. We made V3 the default yesterday, so if you tested text to speech before that change, you were likely using the previous model.
