There is no benefit whatsoever in running separate Midi tracks for each kit piece.

When you insert SD3 (or any other multi output vsti), on the Insert Soft Synth Options page, select Midi Source and All Synth Audio Outputs: Mono

You can then direct each of SD3's kit piece channels in the Mixer page to any of the audio outputs exposed.

Delete any remaining tracks when you've finished. Convert any mono tracks to stereo if you want.

Right, freezing/bouncing.

If you choose to freeze the synth, Sonar will do one pass for each and every audio track and place the rendered audio into the clips pane of each track. The synth will be removed from active duty, but it's only a cvlick away to unfreeze it, make any Midi tweaks and re-freeze.

Right click the freeze button and go into Freeze options. I always make sure "Track Fx" is UNticked because at this stage, I haven't finished mixing.

If you choose to do a bounce to track, Sonar will make ONE pass and render down to brand new tracks. (if you have stereo tracks, you'll need to do these separately in a 2nd bounce) 

This will still leave SD3 live & active, so the first thing you'll want to do is to archive the tracks, or mute them, but if you want to conserve CPU/RAM, you'll have to disconnect it from the Synth Rack. The Browser needs to be undocked in order to reveal the Connect button.

If your bounce is slamming the way you're seeing, I'd check what you're using under Source Category in the Bounce to Tracks dialog. You'll never go wrong if you select Main Outputs.

Some people suggest using Entire Mix, but this can cause problems if you have tracks or busses routed to any additional output pairs on your interface. This caught me out when I first started because I run a dedicated Headphone buss and route this to output pairs 3/4. If you use Entire Mix in this situation, Sonar will sum all active interface busses together and your rendered mix will be 6dB higher for each & every active output pair.

Hope this helps

Thanks for answering but I disagree about the separate MIDI tracks.  I split the MIDI out intentionally as it makes it easier for me to edit.  I can also reassign the MIDI if I want to use an external box or different VI for a particular drum.
 
I will check the Source in the dialog.
 
Note that this is not just an SD3 problem:  it took me multiple attempts to get an Ivory II part bnounced at the right level.

One thing you can do with Session Drummer 3 is, archive all the Audio Outputs that you don't want to freeze. This is very handy if you use the SD3 Templates that come with X1 as they set up all 12 audio outputs. I use those templates then move the outputs I don't use to their own folder and archive the folder. I do that in case I want to use one of those outputs later down the road. It's easier than setting up the output manually.

Archive also works nice for when you have to unfreeze a track, tweak it, then re-freeze it. Say you have kick, snare, 3 Toms, 3 Cymbals, all frozen and you need to fix something with one of the Toms. You can unfreeze just that Tom, tweak it, archive all the other drums/cymbals except the Tom you tweaked, then re-freeze just the Tom.

As for the volume differences, the only thing I can think of on that is, you are freezing using effects you have on the Master Bus or some other bus you may have the track routed to, and you are processing the frozen track through that bus a second time during playback.

Good luck.

Bristol, I fully agree with everything you posted except there are good reasons for having multiple MIDI drum tracks, One but not the only one is to have the option of using more than one drum synth.

Another is the easy ability to adjust each track as needed on playback for volume balance.

One can do very well indeed with a single track however outside of the addition of more tracks as a drawback multiple tracks offer more options to the user.

Putting them into a folder is very helpful though.

Konradh, I have not had a problem with different levels after a freeze or a bounce.

If that were the case freeze would not be so useful. Also I use fast bounce and freeze. I have not run into a problem doing so. I would check why you do if you do. BTW there is no audio difference between a fast bounce and a real time bounce.  CW has made this very clear. Same would apply to freeze.

The only reason to use real time is if a plugin can't handle a fast freeze or bounce.

konradh


 Freezing or bouncing is required before exporting as a WAV or mp3 (as far as I know).

Not necessary but a good idea if you are resource bound. Also a good idea before archiving project just in case the synths/fx are unavailable at some point in the future when the archived project is reloaded.

I usually have the midi in one track these days and multi track bounce (in a single bounce) all the drums separately. I described how to do this step by step in one of my posts. But, if you Prefer separating the midi notes (maybe to Quantize the Kick and snare and hats differently or to use midi velocity to adjust only certain things like toms, sometimes I split the snare midi track into 2 identical tracks then filter velocity 40 and above on 1 and filter velocity below 40 on the other. Then you can quantize Loud notes stronger than the soft ones so as to let buzz rolls through) TO DO THIS: highlight the midi clip and go to Process, Run CAL, then Split Note to Tracks. It will separate each midi Note to a separate midi Track.

Also, Freezing is Not required pre-epxporting or bouncing. The synth will bounce during export. I usually freeze, but have accidentally left them on and they still make it. :)

Since you really don't need to freeze anything before exporting a mix, I guess it comes down to resources.  I don't have the fastest PC around and I rarely freeze anything - not unless the CPU simply can't keep up.
 
But if you do still want or need to freeze, it sounds like you need to work out the difference in levels before anything else.  And for that, I would agree with Bub that there are probably FX plugins coming into play somewhere within your routing.   I'm assuming that if you only freeze one of the MIDI/audio pairs, you get the same difference in levels?  If so, that might help make it easier to track down the culprit.
 
As far as not being able to drag+copy a frozen clip... no idea.   Can you copy it later in the timeline on the same track?
 

Edit: Never mind misread the original post.. thought he wrote SD2 (sorry)  

Rail

Lots of good info here.  Particulary 1-Good to know freeze/bounce is not required and 2-that Fast Bounce is not lower quality.  Thanks.