fix: correct TTS voice parameter name in asset_based pipeline

[hobostay]

Summary

• asset_based.py line 585 passes voice_id=config.voice_id to self.core.tts()
• TTSService.__call__ (tts_service.py:73) accepts voice, not voice_id
• The voice_id kwarg was silently captured by **params and ignored by the local Edge TTS code path
• This caused the default voice to always be used regardless of user's voice configuration

Before

await self.core.tts(
    text=narration_text,
    output_path=str(audio_path),
    voice_id=config.voice_id,  # ← wrong param name, silently ignored
    speed=config.tts_speed
)

After

await self.core.tts(
    text=narration_text,
    output_path=str(audio_path),
    voice=config.voice_id,  # ← matches TTSService.__call__ signature
    speed=config.tts_speed
)

Test plan

• Verified TTSService.__call__ parameter name is voice (tts_service.py:73)
• Verified voice_id goes to **params and is not used by _call_local_tts
• No other callers of self.core.tts() use voice_id= incorrectly

🤖 Generated with Claude Code

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.

---

Test User seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}