I was having a discussion on another thread about audio clipping and the idea that clipping only happens at the interface level and NOT inside of sonar at all.  A couple of people brought up clipping fx in a bin and that to me seems like something I've never been able to do in the box.  Sure, it sounds like it when you play it back and the little red level indicators light up - but it's all an illusion at the end of the day.  
I wanted an easy way to illustrate this, so I thought of a little test I'd share. 
 
PLEASE FOR THE LOVE OF GOD TURN YOUR VOLUME ON YOUR INTERFACE WAY DOWN!! 
 
start with a blank project.  
pull in a short audio file - one of sonar's included loops or something
if it is a loop, unloop it
normalize it to 0db
in the fx bin of the audio track add th-2 producer with an empty preset and turn the output volume all the way up
play back the clip - sounds distorted right??
insert an empty fx chain on the track and turn up the input and output gain on the chain all the way
copy this fx chain 3 times in the bin
play it back - KEEP INTERFACE OR COMPUTER VOLUME LOW! 
sounds just crushed to heck right?
you should be seeing volume of roughly 60db
 
ok - it gets even more fun here!
export the project as a 32bit wav file.  
import the 32bit file into sonar on a new track
the wavform looks totally clipped right?
solo the 32bit file import
play it back - total mayhem
 
drop the volume on the track to -60db  (or whatevery your volume was on the track with effects) and play it back and let me know what you get!
 
Hope some of you have fun with this!!
 

This is roughly what the end result should look like...

I guess I wasn't clear. Some FX just can't handle a too hot input. That was the point I was trying to make. One would have to test every one of the FX to note which ones are susceptible to being over driven. Some wont have a problem others will. I simply don't see the point in pushing things too much. In a way before Sonar had a floating point audio engine it was a easy to know when to back off. 
 
Also I really hate a loss of dynamics in a music. 

I think the idea I was trying to illustrate is those dynamics aren't *really* lost - they're still there, just hidden.  I'm not going to try out every plugin I have to see which ones are susceptible to too hot of an input - I just haven't found any that are yet.  
I totally agree before the digital world a lot of this was easier to judge - and people were forced to use their ears much more.  Part of this is just for fun and experimentation.  
The real neat part is when you flip the phase on the 32 bit file after lowering the volume and play it back with the original clip it perfectly nulls.  That to me was a mind=blown moment a few years back.
I was watching some of Pensado's vids on youtube the other day and he got me thinking about this again with this vid - Digital clipping is your friend

I think over-driving plugins does distort your sound. I definitely work to try to keep from over-driving the PC.

Right, the raw floating point signal path in SONAR can tolerate just about anything, but many FX are sensitive to input levels, often deliberately so (what good would a tube amp sim be if it didn't respond to input level?). You can sum a couple hundred full scale signals, and so long as you pull the level of the last bus down to get it under 0 dBFS before it gets to the D/A converters it'll sound (and export ) nice and clean. But if you want to be sure that FX are all working as intended it's best practice to keep both input and output levels of every individual track, bus and plugin under 0dB.
 
 

Yeah, but it's not happening in the plugin usually - it's happening in the d/a conversion.  Any distortion I've ever gotten on a track by feeding an effect too hot can always be removed by lowering the volume fader for the track.  
 

brundlefly
(what good would a tube amp sim be if it didn't respond to input level?). 

 
Great point Dave!
That's the crux of the idea - those plugins have been designed to act that way.  In other words in a behavior that is unnatural in digital audio processing in an attempt to model real-world analog gear.  
 
I guess I'm just weird that I like having wav files that are fully clipping 60db and look totally unusable and when I pull the volume down a bit it's usable again.
 

Not for me. A lot of FX hard limit as you start to get to the over-driven point. I think You'll notice that if you hard overdrive and then try to restore it you will lose considerable dynamic range.
 
I guess, by this I mean, the distortion will not identically mimic hardware clipping, but that doesn't mean it isn't distortion that you don't want.
 
Frankly, clipping between different interface varies significantly. I have an old Tascam that clips mildly. And m-audio card that clips terribly and a RME UCX that just hard limits at the top. With the RME blind I can't tell when a clip slips by when it happens.

I agree that high input levels  are most often not a problem for FX either, and I emphasize that the key phrase in my previous post (with italics added) is "if you want to be sure that FX are all working as intended it's best practice...".
 
Now I'm thinking I might need to try to brew up a scenario that shows exactly how it can make a significant difference.