bx_limiter & freeze/commit/bounce; anybody willing to look into this?

[alfst28]

Hello everybody,

I've stumbled across quite incredible behavior of bx_limiter and some other Brainworks plugins in regard to freeze, commit and bounce functions. I froze bx_limiter and some other plugins on a drum machine closed hat track in a sparse arrangement and I noticed it's transient is not what it was before the "freeze". Elimination and null tests led me to bx_limiter.

Next I created a blank session, imported a single hit of the high hat, and did the exercise again. The same result; committed track doesn't cancel the track with bx_limiter working on line. The same with freeze and even with imported bounced track. I have the latest built of bx_limiter and I've found similar behavior with bx_digital V3 as well. Yes, I have PT 12.7.0 on osx 10.12.5, which is not certified by Avid, but Plugin-Alliance's (PA) tech support team and a friend of mine with PT 12.8.0 on osx 10.12.5 confirmed null test failures too.

PA's representative has proposed a workaround that is supposed to be working on his side, but it doesn't work for me. He's also suggested that this may actually be a Pro Tools' problem so I'm posting this here, if anybody would be willing to spend couple of minutes to see if they have an idea about what goes on here. I've tested quite some other manufacturers' plugins and they freeze just fine. Even the latest built of bx_dynEQ works fine (while the previous one didn't).

So, below is a Dropbox link with a small .zip file of my test session. If anybody is willing to do it, please, do the following:

1. Commit up to bx_limiter, flip the phase in Trim on the committed track and listen. If the two tracks subtract, then it must be that you won't reproduce my problem on your system and freeze and bounce probably wouldn't give you a different result.
2. If they don't cancel out, you can try freeze and bounce (and import) as well. On my system, they all fail the null test.
3. PA's rep proposed to copy (with all group activated) some silence from beyond the last bar of the song and, with shuffle mode on, paste it to the beginning of the session. His take on my problem was that there is something going on in the first half a second of my session that is messing with the process. So please do that and repeat the commit/freeze/bounce exercise.
   In my case, commit and freeze still fail the null test, after I'd pasted one bar of silence to the beginning as suggested. Bounced track on the other hand cancels the original.

If you get the same results as I do, does anybody have an idea, whether there is something in the session itself causing this?

I'd really appreciate any suggestions. Thanks, Aljoša.

https://www.dropbox.com/s/k87gc6o2px...miter.zip?dl=0
------------------
iMac 27-inch, 3.2 GHz Intel Core i5
32 GB 1600 MHz DDR3
OS Sierra 10.12.5
PT 12.7.0
UAD Apollo twin duo