2017.05 - seriously loud artifacts when using LP MB, LP EQ or Adaptive Limiter with AD2

I came across this when testing for phase issues somebody else reported (which I remembered having myself at some point), so I set up a very simple project using a drum clip from AD2 which I sent to a drum bus and from there to a parallel bus for compression. 
 
I could not (generally) reproduce the phasing problem in this simple project, but I enountered sound artefacts sounding like a computer game from the 1980s (when using Adaptive Limiter) and extremely loud sound which required immediate closing of Sonar (when using LP MB or LP EQ).
 
Projects can be downloaded from here - only using Sonar Platinum stock contect.
 
https://www.dropbox.com/sh/x972u1yg3vh5rn1/AAAJ2X6X70l-M6pvqxVHJiTYa?dl=0
 
BE AWARE THAT THESE PROJECTS CAN CAUSE EXTREMELY LOUD BUZZING NOISE!! TURN DOWN SPEAKERS AND HEADPHONES WHEN TRYING !!!
 
Please note that I tested the same setup with boost 11, Cakewalk CA-2A, ProChannel compressors, Nomad compressors, FF MB, FF C, ... all worked fine. No artefacts, no phasing, no problems.
 
Phasing issues I could only obtain with IK TRacks CS Quad Comp, so I assume that plug may not be reporting latency correctly ???
 

msorrels
I didn't have any problems with these projects at 48khz or 96khz (or maybe I'm not looking at the right thing, I just loaded them and hit play and listened). 
 
So a couple of questions, do you have the "64-bit Double Precision Engine" enabled?  (on the Audio->Driver Settings panel).  It causes so many problems I've written it off as a bad option. 

 
you nailed it right there. "64-bit Double Precision Engine" is indeed checked when hitting play causes that loud beeeeep noise. if unchecked it works.
 
NOW here it gets weird: if I check "64-bit Double Precision Engine" after playing the project for a few seconds with the option unchecked first, it also works. so what is this? sloppy coding/uninitialized variables? LP EQ is Cakewalk in-house plugin, not even 3rd party.
 
If the "64-bit Double Precision Engine" has so many issues (now that you mentioned it, I remember it coming up again and again through the years), why is it still an active (probably even default) option that goes without proper fixing???
 

msorrels 
Do you have AD2 set to upsample on playback/render?  Do you have 2x upsampling enabled at all?  Both those options cause lots of issues as well.

no. i have never used that since I do basically all my work in 96 kHz. the "2x" button is off.

abacab
Apparently a long history, based on these search results ... "64-bit Double Precision Engine" ...

very true. unfortunately problems occur so sporadic (yet so severe) when you have forgotten about the troublemaker and you need to be reminded here in the forum.
 
hence I say fix or remove. maybe it's become irrelevant considering there are other options these days ????

Here is a good description of what happens with 64-bit floating point math in the DAW.
http://www.audiorecording.me/advantages-of-64-bit-daw-over-32-bit-float-digital-audio-workstation.html
 

At this point, it would be nice to hear some input on this issue from the Cake staff!