Hi Marie
Welcome to Microsoft Q&A and Thank you for reaching out.
The “Alias” control in Speech Studio (used to override pronunciation) can disappear or be disabled for a few specific reasons. Here’s a targeted checklist to get you unstuck and an immediate workaround if the UI won’t let you “lock” the adjustment.
Alias checkmark can be missing
- Feature context & tool mismatch The Alias function is available in specific Speech Studio experiences (e.g., Audio Content Creation and some Custom Voice tuning panes). If you’re in a pane that doesn’t support per‑token aliasing, the UI will not render the “✔” apply/lock icon. Double‑check you are in Speech Studio › Audio Content Creation or in a Custom Neural Voice tuning screen, not a different tool like Pronunciation Assessment (which evaluates, not overrides).
- Docs for TTS & supported locales/features: Language & voice support • Text‑to‑Speech REST reference [Language s...soft Learn | Learn.Microsoft.com] [Text to sp...soft Learn | Learn.Microsoft.com]
- Voice type or locale limitations Some combinations of voice style/locale may not expose the Alias UI. Try switching to another voice within the same locale, or confirm the locale supports the features you need. See: Language & voice support [Language s...soft Learn | Learn.Microsoft.com]
- Project permissions/role If your account has Reader or a custom role without edit rights, Speech Studio can load the page but hide “apply/save” controls (including the Alias checkmark). Ensure your identity has Contributor (or higher) on the Speech resource and the Speech Studio project. (General troubleshooting guidance for Speech auth/config: Troubleshoot the Speech SDK) [Troublesho...soft Learn | Learn.Microsoft.com]
- Browser/UI issues
- Try Edge/Chrome with no extensions; some ad‑blockers or script blockers suppress inline controls.
- Reset zoom to 100%—at unusual zoom levels the button can be off‑screen or overlapped.
- Clear cache or open in InPrivate/Incognito.
- If you’re on a small viewport, widen the window; Alias confirmation sits within an inline chip that can collapse.
- Environment differences Your colleague’s success may be because they’re using a different tenant, resource, or voice. Confirm you’re using the same Speech resource, region, voice, and project workspace. The per‑region capabilities (and occasionally preview flags) can differ. Reference endpoints & regions: TTS REST API regions [Text to sp...soft Learn | Learn.Microsoft.com]
Immediate workaround (even if the Alias UI won’t lock)
If the UI control is missing, you can still override pronunciation in synthesis using SSML:
<``speak version="1.0" xml:lang="en-US">
``<voice name="en-US-JennyNeural">
``<!-- Option A: Provide phonetic pronunciation -->
``<phoneme alphabet="ipa" ph="ˈæː.li.əs">alias</phoneme>
``<!-- Option B: Substitute display with alias word form -->
``<sub alias="ay-lee-us">alias</sub>
``</voice>
</``speak>
``
- Use
<phoneme>with IPA (or sapi alphabet) for exact phonetics. - Use
<sub alias="...">to map a written form to the spoken form. - Send this SSML in your TTS REST or SDK call; it bypasses the Studio UI. API details: Text‑to‑Speech REST [Text to sp...soft Learn | Learn.Microsoft.com
Note: For batches or reusable overrides, keep a small SSML library/file in your app. If you later want Studio-managed overrides, you can paste the same phoneme/sub entries into the UI when it’s available.
If this answers your query, do click Accept Answer and Yes for was this answer helpful.
Thank you