I'm planning to build music software for macOS that primarily handles MIDI data and plays audio with low latency. It would be more MIDI-focused than audio-focused, with no real-time audio processing, sample stretching, VST support, or third-party instruments. Are there any important limitations or performance concerns I should expect if I use Swift?
3 Answers
It should work, provided you design the performance-critical parts carefully. Avoid allocating memory during playback, reuse buffers where possible, and keep blocking operations, locks, and unpredictable work away from the real-time audio thread. Since your project is mostly MIDI and straightforward playback rather than heavy DSP, Swift is less likely to become a limitation.
Swift should be capable of handling this kind of application, especially since it’s compiled and doesn’t use a traditional garbage collector. The main thing is to be very careful about allocations, reference counting, and anything that could introduce unpredictable pauses on real-time audio or MIDI threads. For a small or medium-sized project, Swift is a reasonable choice if you already know it well. If you’re choosing a language specifically for maximum control and future scalability, C++, C, or Rust would give you more control over memory and runtime overhead.
The language itself probably won’t be the deciding factor here. Even higher-level languages can work when allocations are minimized and data is handled efficiently. Swift gives you good macOS and Apple-platform integration, while C++ or Rust may be better if you eventually add complex audio processing, plug-in hosting, or other components that require very fine-grained control.
Thanks, that helps. I’ll pay close attention to allocations and real-time thread behavior.