Step 3: Transcription - How do I enable Transcription for my Service using the new UI?

In this step, you will configure the transcription settings for your Service.  This determines how your media will be transcribed. For a list of all steps see our Service Wizard Steps below.

Transcription Type: This determines how your media is transcribed during the Service setup process. You can choose between two options: Automated (AI) or 608 Connection (Human), depending on your workflow and captioning requirements. 

Automated (AI)

Choose the Automated (AI) option to enable AI transcription for your Service. This option enables AI-driven speech recognition to automatically generate transcriptions from your media.

Source Language: Select the language spoken in your media from the dropdown list. This ensures that the transcription engine correctly interprets and transcribes the audio content.

Language Variant: Choose the language variant (example, English US or English UK) to specify the variation of the language used to transcribe your  media.

Speech Engine: Select the Speech Engine that will be used to process the transcription.
There are 4 options to choose from the dropdown menu - Speechmatics (default), Soniox, Amazon and Deepgram.

Dictionary: Optionally, you can select a custom dictionary to improve recognition of specialized terms, names, or industry-specific vocabulary

Advanced Settings: Additional configuration options are available in the Advanced Settings section. Details about these optional settings are listed below.

Navigation Options

At the bottom of the screen, you will see the following actions:

• Cancel - Cancels the creation of your Service and exits the wizard.
• Reset - Clears all fields and settings in the Wizard so you can start over.
• Step Back - Returns to the previous step if you need to modify earlier settings.
• Continue - Moves to the next step of the wizard once the required fields are completed.

Advanced Settings - Automated (AI)

To access the Advanced Settings, click the Advanced Settings dropdown on the page. This will expand the section and display additional configuration options.

Max Delay: Specifies the approximate time (in seconds) the AI engine can take to process and generate transcription. Increasing this value allows more time for complex or lengthy speech segments but it may also increase overall transcription latency.

Filter Disfluency: Enable this option to automatically remove filler words such as “uh” and “um” from the transcript for cleaner output.

Filter Profanity: Select "remove" from the dropdown to identify and filter out profanities from the transcription output. Selecting "keep" would not remove any content. 

Prefer Current Speaker: This setting prioritizes assigning speech to the current speaker, minimizing unnecessary creation of new speaker labels.

Diarization: Enable Diarization to insert speaker change labels in the transcript. Speaker changes will be indicated using chevrons (>>), making transcripts easier to follow.

Punctuation Sensitivity: Controls how frequently punctuation marks are added to the transcription output. This setting accepts a value between 0 and 1, with 0.5 as the default. Lower values (closer to 0) reduce punctuation; higher values (closer to 1) add more punctuation marks.

608 Connection (Human)

Choose the 608 Connection option to use Human generated transcription for your Service. 608 Connection allows you to receive captions from a professional captioner or external captioning service via the CEA-608 protocol instead of AI-based recognition.

Source Language: Choose the language of the captions embedded in your media. This ensures downstream processes, such as translation or dubbing, are aligned with the correct language context.

Timing Offset: The Timing Offset setting help adjust the synchronization between your caption source and the media stream. You can enter a value (in seconds) to shift captions forward or backward if they appear too early or late during playback. Only positive whole numbers are allowed.

Disable Timeout: Enable Disable Timeout to prevent the 608 connection from disconnecting automatically during idle periods. This is useful when your source feed may temporarily pause but should maintain an active session.

Advanced Settings: Additional configuration options are available in the Advanced Settings section. Details about these optional settings are listed below.

Navigation Options

At the bottom of the screen, you will see the following actions:

• Cancel - Cancels the creation of your Service and exits the wizard.
• Reset - Clears all fields and settings in the Wizard so you can start over.
• Step Back - Returns to the previous step if you need to modify earlier settings.
• Continue - Moves to the next step of the wizard once the required fields are completed.

Advanced Settings - 608 Connection

To access the Advanced Settings, click the Advanced Settings dropdown on the page. This will expand the section and display additional configuration options.

Use TLS: Enable Use TLS to establish an encrypted connection for your 608 data stream.
TLS (Transport Layer Security) ensures that all caption data transmitted between your encoder or source and the Service remains secure and protected from unauthorized access.

Convert Upper to Mixed Case: Enable Convert Upper to Mixed Case to automatically convert all-uppercase captions into sentence or mixed case formatting. This makes captions easier to read and visually cleaner for viewers, especially in multilingual or modern display contexts.

Service Wizard Steps