Is it possible to play the MIDI file through your softsynth in Sonar and record the output into an audio recorder like Audacity? If so, then you can import that audio file back into Sonar.

Very, very many years ago somebody gave some kind of an explanation. I don't remember if it came from CW or a forumite, and I don't know if it's true. Back then it sounded logical to me.
 
Settings that  enable you to record VSTs in real time cause easily very bad, ear/loudspeaker damaging feed back loops.
 

have a solid check of your interface as many have a built in return path for recording mains out. This is what I use in Motu 828mkIII
If not, if you have a spare spdif or toslink pair on your interface you are all set.... I cable the spdif in/out on the Komplete Audio 6 just for this.
 
Strangely, on the built in audio in both PCs and the laptop, none of which I use, all three have a return/loopback path for recording.

Similar to @fb.seeker ...
 
I was wondering whether your could do someting with the donationware at VB Audio:
(Virtual mixer/cable(s)/loopback, dual-bus output).
http://vb-audio.pagesperso-orange.fr
 
Either just Voicemeeter and/or in combination with Virtual Cable and ASIO Bridge.
 
I'm finding this software great and still exploring the possibilities. My initial goal was to record track and direct mic commentary simultaneously for video tutorials (which works well).

Thanks for all the great suggestions.  In reference to the concept of exporting just certain audio tracks:
 
•  That would be good if there were no latency problem; or, if there were and I could nudget them into the right place.  Not ideal, but possible.
•  Also, if I could record, I could put all the audio tracks for drums )e.g., kick, snare, hat, toms, OH) into record mode and record them all at once.  With export, I would have to export them one by one and bring them back in.
 
Lots of good thoughts here, so thanks again.
 
Let's hope Toontrack and Cakewalk fix the EZ Drummer 2 issue that started this.

I'm not understanding what the issue is.  I understand that one doesn't record a soft synth in real time, but there's no need to.  You just bounce, keeping the audio in the project without exporting.  Why worry about not being able to record the slow way --??
 
EDIT:  I understand there's a techno hangup with SD2 that seems to be preventing any freezing or bouncing, - my question is to the general question of why soft synths aren't recorded in real time.
 
RB

konradh
With export, I would have to export them one by one and bring them back in.
 

If you bounce to tracks with Source = Tracks, you will get one track/file for every selected audio track. and if you put all the audio tracks adjacent to each other and select the first one as the Destination, each track will be bounced in place in one operation, effectively the same as freezing.

rbowser my question is to the general question of why soft synths aren't recorded in real time.

in the ongoing quest to model analog kit, some synths & fx plugins introduce subtle randomness (or "humanisation") so you never get quite the same playback each time - how's that?

lawpin the ongoing quest to model analog kit, some synths & fx plugins introduce subtle randomness (or "humanisation") so you never get quite the same playback each time - how's that?

Interesting post, lawp - And you understand, I wasn't asking the question.  I was pointing out my reply was to the thread's original question.  I'm not concerned about recording synths in real time.  I remember years ago, when I got the first version of Cakewalk that supported VSTs, I was flummoxed for a day or two while making the transition from using hardware synths to using soft synths.  I expected to record them in real time, the same way I was used to, then saw that we bounce.  Haven't worried about it for years!
 
Randy

lawp

rbowser my question is to the general question of why soft synths aren't recorded in real time.

in the ongoing quest to model analog kit, some synths & fx plugins introduce subtle randomness (or "humanisation") so you never get quite the same playback each time - how's that?

In that case, there would be no advantage to recording rather than bouncing.  Either way, the randomness will be equally unpredictable.