Metadata extractors and embedders are just a specialist form of transform. The targetMediaType in the T-Engine engine-config.json is set to "alfresco-metadata-extract" or "alfresco-metadata-embed" the following is a snippet from the tika_engine_config.json.
{
"transformerName": "TikaAudioMetadataExtractor",
"supportedSourceAndTargetList": [
{"sourceMediaType": "video/x-m4v", "targetMediaType": "alfresco-metadata-extract"},
{"sourceMediaType": "audio/x-oggflac", "targetMediaType": "alfresco-metadata-extract"},
{"sourceMediaType": "application/mp4", "targetMediaType": "alfresco-metadata-extract"},
{"sourceMediaType": "audio/vorbis", "targetMediaType": "alfresco-metadata-extract"},
{"sourceMediaType": "video/3gpp", "targetMediaType": "alfresco-metadata-extract"},
{"sourceMediaType": "audio/x-flac", "targetMediaType": "alfresco-metadata-extract"},
{"sourceMediaType": "video/3gpp2", "targetMediaType": "alfresco-metadata-extract"},
{"sourceMediaType": "video/quicktime", "targetMediaType": "alfresco-metadata-extract"},
{"sourceMediaType": "audio/mp4", "targetMediaType": "alfresco-metadata-extract"},
{"sourceMediaType": "video/mp4", "targetMediaType": "alfresco-metadata-extract"}
],
"transformOptions": [
"metadataOptions"
]
},
If a T-Engine definition says it supports a metadata extract or embed, it will be used in preference to any extractor or embedder using the deprecated frameworks in the content repository.