Are you sure there's no echo or reverb processing? Also, I'm with Kingsmix. WDM/KS has worked for me, but ASIO has always been better so that's what I use. RME has the recipe down for ASIO drivers. Give it a try, it won't break anything and you can always switch back.
 

SonicExplorer
Help me understand..... why would I change over to ASIO if WDM have been working fine for so long?  What are the advantages or additional features?  Or rather is there some drawback/risk to using WDM over ASIO that I'm maybe not aware of?

Most consumer and pro Audio Interfaces makes their own drivers for their products (or outsource it).
RME makes their drivers in house, and are rated (by many) as the Benchmark for other to chase.
You will end up with lower latency, better stability. They support older devices, and OS back to XP.
As mentioned, if it doesn't work, you can always go back.
Driver are found here.
Sorry if I came across as an arse earlier 
All the best.

ASIO allows Sonar to automatically compensate for input timing regardless of your latency settings.
New takes will always be " in sync".
With WDM you have to set the latency manually, it will be different for every latency setting.
I suspect you sometimes use two interfaces?
That would explain the ASIO drivers not " working" on your system.
T

Yes, I use two interfaces.  The RME for primary DAW work and another interface (sometimes on-board) for conventional uses like listening to music, reference checking mixes, system sounds, etc.  I pipe that second interface (via cable) back into the primary interface's monitor input so I can essentially have everything still come out of the same speakers and be controlled by the primary interface, but can do so while keeping things separate in terms making sure the actual DAW work is still isolated & stable.
 
I also need to work on W2K at times, so maybe that was another reason I didn't use ASIO. ??
 
Latency isn't an issue, I can get down into single digit latency with RME's WDM drivers, that's plenty fine for me.  And I'm not even using a blazing system - just a 3GHz Intel Dual Core Pentium.
 
Sonic

Why would the interface driver type have anything to do with the meter behavior described in the OP?
 
The only post-playback activity I've ever noticed with SD2 has been a reflection of cymbal tails fading out. I haven't been using SD3 long enough to notice its meters. If this is SD3 you're referring to, it could be some residual noise in the samples. Given their obsession with realism and not caring about sample size with this version, it wouldn't surprise me if some samples are longer than really necessary.

Agreed Bitflipper, while possible, I highly doubt this problem has anything to do with driver type.
 
Correct, it is Superior 2.

This is why we use ASIO when at all possible. Listen to what Tom said above , this is important, unless you want all your tracks out of sync. Below is a loopback test to show how different drivers re-record an audio track and how Sonar works with the driver to compensate for latency. Only ASIO will report correctly to Sonar, Other driver modes, including WASAPI usually are out of sync unless you manually adjust the offset. 
 

 
But that said, it probably has nothing to do with your issue, Sorry I can't test it, SD 2 was like 8 years ago for me. I think I'd have to install Sonar 7 or something.. This is why we upgrade.. 

<boom> 
 
That was my head exploding.  What are you guys talking about with latency alignment issues?? I never heard of that before, never noticed any problems either.  <confused>   I see what is being said, but it's just not registering to me.  I use MIDI tracks to trigger drum samples, and the rest is audio tracks, always recorded one track at a time.  So what would be out of sync using WDM?  The audio tracks against the drum samples?  Would all the audio tracks be out of sync equally?  
 
FWIW, I've used some plug-ins that have some serious latency and never noticed those even causing a problem, Sonar obviously knew how to handle things just fine.
 
Anyway, I just spent an hour experimenting with SD2 trying to see if I could figure out anything causing the latent meter readings after playback stops.  So far nothing obvious.   There are no delay FX anywhere, I just created a simple test project with a tiny kit.  Nothing special.  Really stumped on this one....

This is going to drive me insane. I've now confirmed this on two different systems: One XP and one W2K. And using two different patches of Superior 2 on each - (2.20 vs. 2.30).  And using two different sound cards with two different driver types.  And two different versions of Sonar (on both systems).   All that is needed to replicate is open a new project, put the Superior 2 synth in the rack, select a kick drum, and simply hit it.  Then watch the meter down below.....
 
Baffled how this can happen, nothing is even feeding Superior 2 but itself in this instance, and no other patching or FX inside Superior 2.  So how in the world is the meter taking forever to settle after a strike of the drum??
 
Anybody able to check this?  I'm pretty sure what I'm experiencing is not normal but it would be nice to hear a confirmation or two to be certain....   It just might be normal after all?
 
Sonic

It works as expected on my laptop, using WASAPI Shared mode.