Our convolver was completely rewritten a few months ago and in our benchmarks even outperformed Apple's own reference convolver. The convolver code is shared between MIR and Synchron Player, so convolution performance is on par for both.I also don't believe that MIR's convolver is less efficient under ceteris paribus conditions, but that's something I never tested myself, admittedly.
Still, if you are using multiple outputs in MIR (for example for surround or ATMOS) complexity will add up.
Last edited: