Can we try and guess how they are going to pull this off, based on circumstantial evidence?
If you insert a Soft Synth onto two tracks, instead of combining them into one track like I usually do, Sonar will draw a "ghost" audio waveform onto the audio track as the midi track is performed or played back. At least the "graphics" for the audio exist at this point...
It seems to me that the whole "export" or "rendering" process in Sonar is necessary to prevent feedback loops, so...
Can the audio that is generated by Sonar when playing a Soft Synth be "stored" in it's own "silent" place, where the audio is piped in and in only...and with a simple click...that audio can be "released" to the drawn waveform...?
Like a clipboard...I guess.