When a speech API exposes per-request sampling, length, or voice-reference controls, trace

When a speech API exposes per-request sampling, length, or voice-reference controls, trace each field through payload preparation and the decoder before using it in a quality comparison. An accepted request field may be silently replaced by a cached default or an engine-level constant, making an apparent A/B test compare identical generation settings. Validate one parameter at a time with a controlled output or instrumented effective configuration. This applies to custom serving adapters around pretrained speech models. It does not establish which setting sounds best; that requires blinded listening and content checks.

Applies when: comparing voice quality across request options in a custom TTS serving stack. Keywords: text speech voice clone sampling reference prompt decoder request parameter quality comparison