I have several plug-ins that are VST2 and VST3. I tested a limiter VST2 vs. VST3 version to see if there was any discernible difference in sound.
There was.
I mixed a track with both the VST2 and VST3 version, and then compared the two by inverting the phase of one, and then mixing them together.
It was down there about -46 db, but upon increasing the gain, I saw a large differential below about 200 Hz.
REALLY interesting result, in my opinion. Upon listening to each separately, the VST3 seemed to have more liveliness and sparkle - for lack of better words.
Has anyone else noticed this?

This is not good news. What of the AAX and AU versions? Are they also different in how they will produce output?
 
Is there a reason the plugin in question is not named? 

I hope CW implement all the features of VST3 in new version...

Did you test it twice with just VST2 (or VST3) to ensure that they nulled completely before comparing the two?
 
Processors often have some randomness that shows up like this in a null test, even things you wouldn't expect to.

"Interesting" indeed. Given that most plugins' VST2 and VST3 versions share a common code base, the audio-processing portions should be literally identical.
 
Remember that VST is an interface specification. The difference from v. 1 to 2 to 3 is in how plugins communicate with the host, and has little to do with what the plugins actually do to audio. If a given plugin sounds noticeably different between VST2 and 3, it's probably a peculiarity of that particular plugin and not something you can generalize about VST2 versus VST3.

bitflipper
"Interesting" indeed. Given that most plugins' VST2 and VST3 versions share a common code base, the audio-processing portions should be literally identical.
 
Remember that VST is an interface specification. The difference from v. 1 to 2 to 3 is in how plugins communicate with the host, and has little to do with what the plugins actually do to audio. If a given plugin sounds noticeably different between VST2 and 3, it's probably a peculiarity of that particular plugin and not something you can generalize about VST2 versus VST3.

+1

There would need to be far more detailed information (is what plugins test positive and what plugins test negative). If there's just one plugin comparison that compares OK then it isn't a Sonar issue.

CakeAlexS
There would need to be far more detailed information (is what plugins test positive and what plugins test negative). If there's just one plugin comparison that compares OK then it isn't a Sonar issue.

not necessarily, but repeating the test in other daws would rule sonar out

I don't believe so. Other DAW's may send out false flags themselves.. Or some of the plugin's code may only be executed in particular DAWs but not in Sonar (or vice versa).

Which plug-ins? Would like to try and reproduce.