In case of mp3, m4a, ALAC, AAC, WAV, AIFF, CA formats, files are decoded also by Core Audio.
Files are decoded by their native decoders and Core Audio (Mac OS X audio engine) is used for playback. So as for technical information - the system is playing audio for us. And I am a sound engineer with 12 years of experience in live sound/studio. When data are sent to device which does not support 32-bit float sample format internally, these data will have to be converted to a supported sample format (typically, 24 -bit integer), and that's where bad things may come into play, however, that's only my assumption, for I am unable to hear the difference. That'd be bit-perfect playback up to 24 bits. If re-sampling required - we do it very gently, using 32-point sinc interpolation, and that may be avoided by Sample Rate Sync option. There's no hidden EQ or enhancing, we just decode file to 32-bit floating - point data and send it to system's device. Otherwise than this, it tries to keep audio path as clean and processing - free as possible. The answer is it does, but only if You enable it. Your question - does VOX introduce EFX/EQ. I asked my contact at Coppertino about the technical details of Vox and here is the relevant part of his response:
