Brett, thank You for Your question.
Short answer is, it's all about gain structure. Vox plays file
"as is", AU's are simply inserted in digital signal chain with no
intentional gain change. After all, Vox is a player, not a DAW.
"Bit perfect" is often misinterpreted term.
Longer answer.
First note is, once You plug in any effect (EQ, AU, VST,
whatever) it's not "bit perfect" anymore per se - in a sense that
You've just brought in additional computations between the file and
output affecting the signal. Then, other problem of audio
processing often comes in - an additional gain that without
appropriate gain reduction causes clipping.
There are different approaches to what "bit perfect" is. Vox is
bit perfect (when there is no effect in chain! and volume is 100%!)
in a sense that it tries not to change the incoming audio data by
multiplication or other related processing. Given that most music
tracks go out excessively limited to the maximum possible
magnitude, being "bit perfect" means, playing at maximum digital
levels just prior to clipping. Then analog volume control is used
to reduce volume. It's ok with just =playing= already mastered
music.
What that means is, if You have added some more gain/volume in
Your Audio Units in Vox, You will get unwanted distortion. So You
need to be knowing what You are doing. That's partly a reason why
Audio Units is somewhat hidden feature in Vox - it's too easy to
spoil sound by unqualified use. For example, desktop Vox uses
"gain" slider in EQ to compensate the gain the user might dial in
on EQ itself, to avoid clipping. You need to adjust this slider
Yourself. In Audio Units there are also "gain" controls for the
same purpose. Vox does not do this for You automatically - it does
not try to compensate for levels or shift dynamic range. After all,
it's for playing what has been already mastered and maxed, not for
mastering and maximizing.
On the other side, Logic has to deal with audio processing a
lot, it's a professional software for audio and that brings up a
common issue with audio editing software - for example, to be able
to amplify some portion of spectrum, in audio where everything is
already amplified, someone (including software) needs first to
lower the total gain of the signal or/and use wider internal
dynamic range in digital domain. Then, given You use floating-point
sample format, You are able to do Your math/processing, without
user even noticing the gain change.
In other words, Logic is much more complex software, so don't
expect Vox to behave the same. Probably that's why You paid for
Logic much more than for Vox :)
In my opinion, music production software cannot be "bit perfect"
per se - it's for processing, not for playback!
Therefore, the problem is in Your levels - You need to lower
Audio Units gain in Vox to achieve the same result You get in
Logic. Logic just may use different (more complicated) gain
structure.
Vox uses BASS audio engine internally in a 32-bit floating point
mode, so You may find information about it.
Sorry for long text, but the question is serious!
Please feel free to ask more questions and supply
recommendations!
