From 76ece92cfd47002f09e2aad3bfd9e059ee384856 Mon Sep 17 00:00:00 2001 From: Andrew Stoltz Date: Fri, 24 Apr 2026 23:37:30 -0500 Subject: [PATCH] fc-ttsreader: enable real Whisper alignment via fc-speech-align MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Flips Speech__Alignment__Enabled=true and points BaseUrl at the new BLUEJAY-WS podman quadlet running fc-speech-align (faster- whisper, /align contract). When Lane 1δ's /api/v1/voices/preview-with-timings runs after this lands, the alignment.source field flips from 'estimated' to 'whisper' and the per-word timings come from real audio analysis instead of uniform-spacing estimates. No image rebuild — the Lane 1α DI registration already routes IWhisperAlignmentClient to FasterWhisperAlignmentClient when Speech:Alignment:Enabled is true. Companion firewall rule from FlowerCore.Puppet@bbc02ea + @05504ed (whisper_align_enabled flag on bluejay-ws-linux Hiera) opens port 9200 to RKE2 pod CIDR durably. --- apps/fc-ttsreader/fc-ttsreader.yaml | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/apps/fc-ttsreader/fc-ttsreader.yaml b/apps/fc-ttsreader/fc-ttsreader.yaml index 9ab1646..cc024db 100644 --- a/apps/fc-ttsreader/fc-ttsreader.yaml +++ b/apps/fc-ttsreader/fc-ttsreader.yaml @@ -172,6 +172,12 @@ spec: value: "http://10.0.56.20:10401" - name: TtsReader__Kokoro__TimeoutSeconds value: "120" + - name: Speech__Alignment__Enabled + value: "true" + - name: Speech__Alignment__BaseUrl + value: "http://10.0.56.20:9200" + - name: Speech__Alignment__TimeoutSeconds + value: "120" - name: TtsReader__Ollama__BaseUrl value: "http://10.0.57.17:11434" - name: TtsReader__Ollama__DefaultModel