Skip to main content
Version: v2

Ingest protocols

The Dolby OptiView Live platform supports low latency live streaming through two ingest protocols: RTMP and SRT.

RTMP supports both pull and push mode out of the box. SRT works out of the box in pull mode only, where the platform connects to your encoder to pull the stream. If you need SRT push mode, contact your account representative to have it enabled.

RTMP — Real-Time Messaging Protocol​

RTMP is a widely adopted protocol originally developed by Adobe for transmitting audio, video, and data over the internet. It remains one of the most commonly supported protocols across hardware and software encoders.

  • Pros
    • Broad encoder support — Nearly all hardware and software encoders support RTMP out of the box.
    • Simple setup — Requires minimal configuration, typically just a server URL and stream key.
    • Low latency — Provides low latency contribution suitable for most live streaming use cases.
  • Cons
    • Container limitation — RTMP uses the FLV container, which only supports H.264 video and AAC audio. Newer codecs such as H.265/HEVC are not supported.
    • Single audio track — RTMP is limited to a single audio track per stream, making it unsuitable for multi-audio use cases.

SRT — Secure Reliable Transport​

SRT is an open-source protocol designed for low latency, secure, and reliable video transport across unpredictable networks. It is increasingly adopted as a modern alternative to RTMP.

  • Pros
    • Resilient on unreliable networks — SRT uses UDP with built-in error correction (ARQ), making it well suited for contribution over the public internet or unstable connections.
    • Multi-audio support — SRT uses the MPEG-TS container, which supports multiple audio tracks within a single stream.
    • Closed captions and metadata — The MPEG-TS container also supports embedded closed captions (CEA-608/708) and SEI metadata pass-through.
    • Codec flexibility — MPEG-TS supports a broader range of codecs compared to FLV.
  • Cons
    • Less universal encoder support — While adoption is growing, not all encoders support SRT, especially older hardware models.
    • More complex configuration — SRT may require tuning parameters such as latency, overhead bandwidth, and encryption settings for optimal performance.

SRT passphrase​

In pull mode, you can secure the connection by including the passphrase directly in the SRT URL that you configure on your channel (e.g. srt://encoder-host:port?passphrase=my-secret). The passphrase is write-only.

In push mode, the Dolby team configures the passphrase for you and provides an SRT URL with the passphrase included.

Credentials in pull source URLs​

Credentials in the source url of a pull ingest (srt-pull, rtmp-pull or rtsp-pull) are write-only. These include SRT passphrases and stream IDs, RTMP stream keys in the last path segment, and usernames or passwords in the URL's user info.

Neither the API nor the dashboard returns these credentials after you submit them, including in create and update responses. The API sanitizes the url as follows:

  • For all three (SRT, RTMP and RTSP): user info, query and fragment are removed.
  • For RTMP also: the last path segment (the stream key) is removed.

The entire SRT query is removed, including non-secret options such as latency or mode, not just the passphrase or stream ID. When anything is removed, the response includes credentialHidden: true. Otherwise, the URL is returned unchanged and credentialHidden is absent.

Keep the full source URL securely: to change the source, send the full URL, including credentials, again. Do not reuse the sanitized response URL as the source URL in an update. An update that omits url keeps the stored URL and its credentials.