Topic
Nkomo Engineering
Choosing a Mandarin Voice: Xiaoxiao Is a Default, Not the Whole Catalogue
Xiaoxiao is a sensible Mandarin default, but a default should not erase the catalogue. Microsoft documents multiple Simplified Chinese speakers and styles…
Mandarin TTS Beyond the Demo: Names, Numbers and Polyphonic Characters
Mandarin neural voices can sound excellent on curated prose and still fail on the details users notice immediately. Polyphonic characters, personal names…
Choosing a Hindi TTS Voice: Swara, Madhur and the Audience
Voice selection is more than choosing “male” or “female.” A Hindi narration voice needs to fit the audience, format and duration. A lively short-form…
Hindi TTS and Hinglish: Why Mixed Script Changes the Result
Hindi text-to-speech comparisons often use clean Devanagari prose. Real messages are less tidy. Product names, addresses, acronyms and whole clauses…
Testing Arabic TTS: The Prompts That Reveal a Weak Voice
A polished Arabic demo sentence proves very little. Production text contains undiacritized words, personal names, currencies, telephone numbers…
Arabic TTS Voice Choice: One Arabic Menu Is Not Enough
A single option labelled “Arabic” hides the decision listeners care about most: which Arabic will they hear? Microsoft’s speech catalogue separates…
Mandarin TTS in Nkomo: Why Xiaoxiao Moves Ahead of WaveNet
Nkomo previously sent Mandarin synthesis to Google `cmn-CN-Wavenet-A` before trying Microsoft's `zh-CN-XiaoxiaoNeural`. We have reversed that order…
Hindi TTS in Nkomo: Why Google Neural2 Still Leads Swara
For Hindi, Nkomo is deliberately not making the same change as Arabic and Mandarin. Google `hi-IN-Neural2-A` remains first. Microsoft `hi-IN-SwaraNeural`…
Arabic TTS in Nkomo: Why Salma Now Speaks Before Google WaveNet
Nkomo has two materially different Arabic voice choices. Google supplies `ar-XA-Wavenet-A`, a broad Arabic WaveNet voice. Microsoft supplies…