In this lesson, you will learn how to distinguish between the legacy Sound Cue system and the modern, sample-accurate architecture of MetaSounds.
The Sound Cue Legacy
Sound Cues are the "high-level" legacy way to handle audio in Unreal Engine. Architecturally, they function as a node-based graph that acts as a wrapper for your audio assets. When you trigger a Sound Cue, the engine processes the graph on the Audio Thread. Because it relies on high-level commands, Sound Cues are often subject to thread latency; you cannot perform frame-perfect logic or complex sample-rate manipulations because the system was designed for simple playback, layering, and randomization.
The MetaSound Revolution
MetaSounds operate on a fundamentally different, "low-level" architecture. Unlike Sound Cues, a MetaSound is a node-based digital signal processor (DSP). It doesn't just play an asset; it compiles into a graph that generates or processes audio on the Audio Render Thread.
Because MetaSounds run at the sample level rather than the control level, they are sample-accurate. You can manipulate audio input, create procedural synthesizers, and drive complex logic—like dynamically changing a filter based on a UI slider value—without the "chunky" latency associated with legacy cues.
Consider the difference in execution:
- Sound Cues: "Play this wave file at 0.5 volume." (Command-based).
- MetaSounds:
[Wave Asset] -> [Envelope] -> [Low Pass Filter] -> [Output]. (Signal-flow based).
In practice, if you are binding audio to a UI event—such as a slider movement you’ve implemented using your MVVM patterns—MetaSounds allow you to send continuous float values that the graph processes in real-time, resulting in smooth, reactive audio that feels integrated into the engine rather than "tacked on." While Sound Cues are still useful for simple "fire and forget" sounds, MetaSounds provide the professional-grade control required for dynamic, data-driven audio environments.
Key Takeaway: Sound Cues are high-level command-based wrappers, while MetaSounds are low-level, sample-accurate signal processors that move audio logic from the Audio Thread to the Audio Render Thread.