I'm asking what is the preferred method of adding effects to MIDI's and Synths. Is it it better to add them BEFORE you bounce them to audio or AFTER they are an audio file. I have done it both ways, and have had success either way, just wanting to know what is most common method.

If you are using a Cakewalk DAW there is no requirement to bounce tracks to audio so the question is a little unusual.

Well, thank you. Never mind then.

OK... never bounce anything to audio that you don't plan to ever want to change.
 
So, no I would NOT include the FX in the bounce. Always bounce a clean file. If there are FX in the midi/synth track, either delete them or shut them off for the bounce unless, as I said, you don't plan to do any audio FX later.
 
You can then insert the FX in the audio track's FX bin to your heart's desire. If something doesn't work, doesn't sound right, or you just want to experiment, you simply shutoff that FX or delete it from the bin. If you have bounced it with an FX such as reverb or whatever..... nope.... you are stuck with it forever. Unless you still have the original midi track and synth, but then you have to go through the work to re-bounce it..... simple just to bounce clean and use audio FX later. 
 
I employ this method and it has saved my butt more than once. Bounce clean.
 
I use the same principle with recording audio tracks.... record them clean. My exception to this is my guitar. I add reverb to taste and it gets printed "as recorded" and rarely gets changed or edited.

Thanks, I was thinking that was the way to go, like you said, if you bounce it with the effects you are stuck with it. Kind of a silly question, but, I was just asking. Thanks again.

There isn't a right or wrong way, really.  Personally, I usually bounce down tracks - old habits die hard and that is a left over from the low cpu days.  however, my computer today (and most, I warrant) should do fine using real-time effects w/o bouncing.  So it is a matter of preference, not rule.
 
Again, if you bounce first before affecting, use freeze, copy and paste the audio then unfreeze and archive the midi so you can go back to square one if you change your mind.
 
@

For my sample-based soft synths such as Kontakt, Superior Drummer and Omnisphere, I prefer to freeze them but exclude the fx bin from the freeze. This conserves RAM by unloading the synth and its samples from memory, avoids long load times on startup, and reduces CPU usage during playback while still leaving options open for subsequently tweaking EQ and effects.
 
Non-sample-based synths often aren't as critical and are more likely to warrant further tweaking as I go, so I usually don't freeze/bounce them until late in the mix process. The exception would be particularly CPU-intensive synths/patches and tracks with CPU-hungry effects such as reverbs on them. 
 
Another trick I often employ is to split a MIDI track into two or more tracks if there are CPU-intensive effects on just part of the track. For example, a drum track might employ a long reverb on just one or two hits. I'll separate out those hits, insert a second instance of the synth and freeze that section along with the reverb, leaving the main body of the track available for editing.

Thanks gentlemen, for the input. I have been bouncing everything clean and it is working well. I copy my files that I use live, so I will always have a "start over" point. I have found editing the clip automation's back to flat (no pan, reverb, chorus, ect.) track are bouncing real clean and require very little effects. Since my struggles with EMu, and going to the Quad-Capture, I'm having a ball with this, and am learning a lot without waiting for the next thing to go wrong. Thanks again!