superman - I assume it is volume levels that you want to automate? To start with, automation in Sonar is done after you have recorded your material.
The simplest way (for me at least) is to record your track first. Once the track is recorded, play it back and adjust the general volume level of the track to roughly where you want it to be using the "vol" box from the left frame of Track View.
After you have the "general" volume to your liking, right click on the recorded waveform and select "envelopes" from the popup menu, and the "create" and "volume" from the submenus.
Once you do this you will get a straight horizontal line drawn through your wave (the volume "envelope"). This line represents the "default" volume setting for the track. (If you hold your mouse over the line, a popup window will tell you what it is set at.)
Now start your playback. When you reach a point where you want to implement a volume change, stop the playback. Put your mouse pointer on the volume envelope line exactly where you want the volume chanqe to start. Right click the mouse and choose "add node" from the popup menu. It will place a square box on the volume envelope. Now move slightly to the left and add another node.
Begin you playback again and stop it at the point where you want the volume to change again (or return back to the original setting). Place a node at this point, and one just slightly to the right.
You should now have 4 nodes on the volume envelope line. Put your mouse on the volume envelope inbetween the two middle nodes. Left click and drag the line upwards (to increase volume) or downwards (to decrease volume). As you do this, a window will tell you the db setting of the change.
Start your playback again from before the first node and listen to the change. Adjust it again according to your taste.
While this might sound complicated, it is actually quite easy. It took me longer to explain it then it would to actually do it.
The beauty of this is:
1. It is all non-destructive. You can make all the adjustments you want without effecting the orginal volume of the wave.
2. You can visually see where all the changes are taking place.
3. You can go back an readjust any change (or delete it entirely)
4. You can make very rapid changes (e.g., increase/decrease the volume of a single word in a vocal).
5. Your timing can be impeccible, since you are visually placing the node, rather than trying to move a fader at exactly the right time.
Good luck, and check back if you have any problems.