Tuesday, July 28, 2015

i should point out that i've been able to determine that i'm *not* hearing things.

i had all of the initial renders. i created a few new renders. now, if the gear was working correctly and i was hearing things, those files should be identical. you can test to see if that's the case via phase inversion. if they're the same, you should get a flat signal when you paste the inversion on top of the original. we then say that the two signals null.

the renders i created today did not null with the renders i created a few weeks ago. the cubase projects have not been modified in that period, i can be sure of that by checking the dates on the project files.

that should not make sense. they should null.

so, what that indicates is that the signal that was coming out of the mixer a few weeks ago was different than the signal that is coming out of the mixer today. that can only be true if something was broken at that time. so, that's a mathematical proof that the files are in fact different and i'm not tricking myself.

it's sort of strange, though. because in some cases the drums actually do null. the bass does not come close to nulling. it's normal for otherwise identical files to not null over reverb, and so i might expect some bleeding over the amp simulation. but, it's far too profound for that. the guitars also tend to roughly null, except at the high frequencies. put together, this is consistent with what i was experiencing - a floor in the sub-bass along with a cut in the high end.

i've been fairly certain that what i'm getting out today ought to be the clean, correct version. but, i'm not entirely sure if i mixed some of these things before or after it broke. the first track actually nulled perfectly. the next couple of tracks were mixed on the same day, but i re-rendered some of them. the result is that i'm lacking a control.

if i was absolutely certain the device is stable, i could start over, but i'm really not.

so, i guess i should take a step back and realize that i've built up a level of complexity that i'm better off finding a way to jettison. that's what i ought to be focusing on - ensuring clean drivers and starting over.