I am using foobar2000, running the foo_input_vgmstream component that I maintain, to play back a file of the ADX2/HCA format, which is MCDT based. This particular build of the component is automatically built using Github Actions, with x87 based math, 32 bit, like the main app.
When attempting to play back this file (Sidhe Priest cover of The Healer Stalks from Doom 2), using either the ReplayGain scanner or the decoding speed test, it is apparent that the HCA decoder is only running at about 2.75x real time on my M1 Mac mini. In my native player Cog, even when forced to run as Rosetta, it runs almost 50 times as fast as this. Under Parallels 16.5, running Windows on ARM, running the same foobar2000 configuration, it also decodes at over 80x real time.
Is this a bug in Rosetta 2, or a limitation in CrossOver's 32on64 wrapper?
Please Note: This Forum is for non-application specific questions relating to installation/configuration of CrossOver. All application-specific posts to this Forum will be moved to their appropriate Compatibility Center Forum.
CrossOver Forums: the place to discuss running Windows applications on Mac and Linux
CodeWeavers or its third-party tools process personal data (e.g. browsing data or IP addresses) and use cookies or other identifiers, which are necessary for its functioning and required to achieve the purposes illustrated in our Privacy Policy. You accept the use of cookies or other identifiers by clicking the Acknowledge button.
