Transcript: Support set the video codec parameters. v5.14.11

Author: winlinvip

DEVELOPER.md

@@ -1151,6 +1151,7 @@ The following are the update records for the SRS Stack server.
     * Dubbing: Refine download button with comments. v5.14.8
     * Room: AI-Talk support post processing. v5.14.9
     * Website: Support setting title for popout. v5.14.10
+    * Transcript: Support set the video codec parameters. [v5.14.11](https://github.com/ossrs/srs-stack/releases/tag/v5.14.11)
 * v5.13:
     * Fix bug for vlive and transcript. v5.13.1
     * Support AWS Lightsail install script. v5.13.2

platform/transcript.go

@@ -1018,11 +1018,13 @@ type TranscriptConfig struct {
 	Language string `json:"lang"`
 	// The force_style for overlay subtitle.
 	ForceStyle string `json:"forceStyle"`
+	// The video codec parameters.
+	VideoCodecParams string `json:"videoCodecParams"`
 }
 
 func (v TranscriptConfig) String() string {
-	return fmt.Sprintf("all=%v, key=%vB, organization=%v, base=%v, lang=%v, forceStyle=%v",
-		v.All, len(v.SecretKey), v.Organization, v.BaseURL, v.Language, v.ForceStyle)
+	return fmt.Sprintf("all=%v, key=%vB, organization=%v, base=%v, lang=%v, forceStyle=%v, videoCodecParams=%v",
+		v.All, len(v.SecretKey), v.Organization, v.BaseURL, v.Language, v.ForceStyle, v.VideoCodecParams)
 }
 
 func (v *TranscriptConfig) Load(ctx context.Context) error {
@@ -1758,11 +1760,15 @@ func (v *TranscriptTask) DriveFixQueue(ctx context.Context) error {
 			}...)
 		}
 	}
+	// Generate the video parameters.
+	videoCodecParams := "-c:v libx264 -profile:v main -preset:v medium -tune zerolatency -bf 0"
+	if v.config.VideoCodecParams != "" {
+		videoCodecParams = v.config.VideoCodecParams
+	}
+	args = append(args, strings.Split(videoCodecParams, " ")...)
+	// Generate other parameters for FFmpeg.
 	args = append(args, []string{
-		"-vcodec", "libx264", "-profile:v", "main", "-preset:v", "medium",
-		"-tune", "zerolatency", // Low latency mode.
-		"-bf", "0", // Disable B frame for WebRTC.
-		"-acodec", "aac",
+		"-c:a", "aac",
 		"-copyts", // To keep the pts not changed.
 		"-y", overlayFile.File,
 	}...)

ui/src/pages/ScenarioTranscript.js

@@ -53,6 +53,7 @@ function ScenarioTranscriptImpl({activeKey, defaultEnabled, defaultConf, default
   const [baseURL, setBaseURL] = React.useState(defaultConf.baseURL || (language === 'zh' ? '' : 'https://api.openai.com/v1'));
   const [targetLanguage, setTargetLanguage] = React.useState(defaultConf.lang || language);
   const [forceStyle, setForceStyle] = React.useState(defaultConf.forceStyle || 'Alignment=2,MarginV=20');
+  const [videoCodecParams, setVideoCodecParams] = React.useState(defaultConf.videoCodecParams || '-c:v libx264 -profile:v main -preset:v medium -tune zerolatency -bf 0');
 
   const [liveQueue, setLiveQueue] = React.useState();
   const [asrQueue, setAsrQueue] = React.useState();
@@ -89,15 +90,15 @@ function ScenarioTranscriptImpl({activeKey, defaultEnabled, defaultConf, default
     if (!baseURL) return alert(`Invalid base url ${baseURL}`);
 
     axios.post('/terraform/v1/ai/transcript/apply', {
-      uuid, all: !!enabled, secretKey, organization, baseURL, lang: targetLanguage, forceStyle,
+      uuid, all: !!enabled, secretKey, organization, baseURL, lang: targetLanguage, forceStyle, videoCodecParams,
     }, {
       headers: Token.loadBearerHeader(),
     }).then(res => {
       alert(t('helper.setOk'));
       console.log(`Transcript: Apply config ok, uuid=${uuid}.`);
       success && success();
     }).catch(handleError);
-  }, [t, handleError, secretKey, baseURL, targetLanguage, forceStyle, uuid, organization]);
+  }, [t, handleError, secretKey, baseURL, targetLanguage, forceStyle, videoCodecParams, uuid, organization]);
 
   const resetTask = React.useCallback(() => {
     setOperating(true);
@@ -307,6 +308,9 @@ function ScenarioTranscriptImpl({activeKey, defaultEnabled, defaultConf, default
                   <Nav.Item>
                     <Nav.Link href="#overlay" onClick={(e) => changeConfigItem(e, 'overlay')}>{t('transcript.overlay2')}</Nav.Link>
                   </Nav.Item>
+                  <Nav.Item>
+                    <Nav.Link href="#transcode" onClick={(e) => changeConfigItem(e, 'transcode')}>{t('transcript.transcode')}</Nav.Link>
+                  </Nav.Item>
                 </Nav>
               </Card.Header>
               {configItem === 'provider' && <Card.Body>
@@ -334,6 +338,15 @@ function ScenarioTranscriptImpl({activeKey, defaultEnabled, defaultConf, default
                   <Form.Control as="input" defaultValue={forceStyle} onChange={(e) => setForceStyle(e.target.value)} />
                 </Form.Group>
               </Card.Body>}
+              {configItem === 'transcode' && <Card.Body>
+                <Form.Group className="mb-3">
+                  <Form.Label>{t('transcript.trans0')}</Form.Label>
+                  <Form.Text> * {t('transcript.trans1')}. &nbsp;
+                    {t('helper.see')} <a href={t('transcript.trans2')} target='_blank' rel='noreferrer'>FFmpeg: video codec</a>.
+                  </Form.Text>
+                  <Form.Control as="input" defaultValue={videoCodecParams} onChange={(e) => setVideoCodecParams(e.target.value)} />
+                </Form.Group>
+              </Card.Body>}
             </Card>
             <p></p>
             <Button ariant="primary" type="submit" onClick={(e) => {

ui/src/resources/locale.json

@@ -379,6 +379,10 @@
         "res": "最后收到的响应"
       },
       "transcript": {
+        "trans2": "https://ossrs.net/lts/zh-cn/faq-srs-stack#how-to-setup-the-video-codec-parameters-for-ai-transcript",
+        "trans1": "自定义视频转码参数",
+        "trans0": "视频转码参数",
+        "transcode": "转码设置",
         "fstyle3": "https://ossrs.net/lts/zh-cn/faq-srs-stack#how-to-setup-the-font-style-for-ai-transcript",
         "fstyle2": "设置字幕的样式",
         "fstyle": "字幕样式",
@@ -870,6 +874,10 @@
         "setupOk": "Set record rule ok"
       },
       "transcript": {
+        "trans2": "https://ossrs.io/lts/en-us/faq-srs-stack#how-to-setup-the-video-codec-parameters-for-ai-transcript",
+        "trans1": "Customize video transcode parameters",
+        "trans0": "Video Transcode Parameters",
+        "transcode": "Transcode Settings",
         "fstyle3": "https://ossrs.io/lts/en-us/faq-srs-stack#how-to-setup-the-font-style-for-ai-transcript",
         "fstyle2": "Force the style of the subtitle",
         "fstyle": "Force Style",