DisplayPort MST Audio to Multiple Virtual Devices

DisplayPort Multi-Stream Transport (MST) supports sending independent audio streams to multiple displays connected via a single DisplayPort output. Because MST treats each connected display as a separate downstream node with its own dedicated bandwidth allocation, the operating system can expose each display as an independent audio endpoint. With the appropriate operating system support, drivers, and routing software, users can route distinct audio sources to separate virtual and physical audio devices across an MST topology.

How DisplayPort MST Handles Audio

DisplayPort transmits audio embedded within the video stream using Secondary Data Packets. Under an MST configuration, a single physical connection multiplexes multiple independent data streams into a single high-bandwidth link.

Each display connected to an MST hub or daisy-chain topology is allocated its own virtual channel. Just as video pixels are routed specifically to targeted monitors, the accompanying audio packets are routed to the specific target monitor's DisplayPort receiver. As long as the sink device (monitor or TV) reports audio capabilities in its Extended Display Identification Data (EDID), the graphics card can allocate an independent audio pipe to that display.

OS and Driver Detection

When multiple MST displays with integrated speakers or audio pass-through jacks are connected, the graphics driver (NVIDIA, AMD, or Intel) registers multiple High Definition Audio (HDA) endpoints with the host operating system.

Instead of showing a single audio device for the entire physical DisplayPort connector, the system detects each active audio-capable screen individually. In environments like Windows, macOS, or Linux, these monitors appear as distinct hardware playback devices (e.g., "Display 1 - NVIDIA High Definition Audio", "Display 2 - NVIDIA High Definition Audio").

Routing Audio to Multiple Virtual Devices

Because the hardware exposes these independent endpoints, audio streams can be mapped directly or through virtual devices:

  • Native Application Routing: Modern operating systems allow users to assign different applications to different output devices natively. For example, a web browser can output sound to Monitor A while a media player outputs sound to Monitor B.
  • Virtual Audio Cables and Mixers: Virtual audio software (such as VoiceMeeter on Windows or PipeWire/PulseAudio configurations on Linux) can create virtual input devices that route and split audio signals across multiple physical DisplayPort endpoints simultaneously.
  • Virtual Machines: When using virtualization (such as KVM/QEMU), individual DisplayPort audio endpoints or virtualized audio devices mapped to those endpoints can be passed through directly to separate guest operating systems.

Hardware and Bandwidth Limitations

While DisplayPort MST architecture natively accommodates multiple audio streams, real-world execution depends on specific hardware constraints:

  1. Audio Codec Endpoints: Graphics cards have a fixed number of internal digital audio codecs (often limited to 3 or 4 concurrent streams). If the number of connected displays exceeds the GPU's internal audio endpoints, some displays cannot receive an audio stream.
  2. MST Hub Support: The MST hub or daisy-chaining monitor must properly forward secondary audio packets according to the DisplayPort 1.2 (or higher) specification. Cheap or non-compliant splitters may duplicate audio across all outputs rather than presenting distinct endpoints.
  3. Total Link Bandwidth: While uncompressed audio consumes minimal bandwidth compared to high-resolution, high-refresh-rate video, the total aggregated bandwidth of all video and audio streams must remain within the limits of the source DisplayPort standard (e.g., DP 1.4 or DP 2.1).