This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(105.60000000000001, 19%, 20%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ESH%3C/text%3E%3C/svg%3E)
When synthesizing Speech Synthesis Markup Language (SSML) text explicitly using the Iraqi locale (xml:lang="ar-IQ") and designated Iraqi neural voice models (ar-IQ-BasselNeural and ar-IQ-RanaNeural), the system is failing to produce the correct regional dialect. Instead of outputting an Iraqi accent, the neural engine defaults/falls back to a Cairo-centric Egyptian accent (ar-EG).
Steps to Reproduce:
<voice name="ar-IQ-BasselNeural">.ar-EG (Egyptian) model instead of ar-IQ.Expected Behavior: The audio output should strictly utilize the native phonetics, intonation, and speech patterns of the deployed Iraqi (ar-IQ) neural models.
An Azure service that integrates speech processing into apps and services.
This kind of support is not provided on the free tier. I only have access to a chatbot which does not help and also does not accept support documents and sample files.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(230.39999999999998, 48%, 32%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EEW%3C/text%3E%3C/svg%3E)
I would recommend opening a Speech support ticket and providing a minimal SSML sample together with the generated audio output. According to Microsoft's language and voice support documentation, the ar-IQ locale is supported and Iraqi voices are published as distinct voice models. If the rendered output does not match the expected Iraqi dialect, that could indicate a model-quality issue rather than an SSML configuration problem. Including the Speech resource region, SSML payload, and sample audio files in the support request should help the Speech team determine whether this is a service issue or expected behavior of the current voice model.