Recently Blender received an audio upgrade to integrate 3D sound effects for game building. This came through the Pepper development branch and was engineered by NeXyon (Jörg Müller).
Here is a link to the GSoC proposal page:

http://wiki.blender.org/index.php/User:NeXyon/GSoC2011/Proposals

And here is the development blog to see what happened:

http://audaspace.wordpress.com/2011/08/25/gsoc-2011-wrap-up/

Fortunately some of this development spilled over into Blender’s VSE too. there are some really useful tools now, like

waveforms on strips

a working f-curve fade effect

You can also change the pitch of an audio strip, which is good for changing audio length to match video with different frame rate. Or you can f-curve the pitch change over time (weird but cool).

And somewhere in there is shelving for high and low frequencies.

Now it’s no Digital Audio workstation (DAW) but at least the VSE has the basics for video assembly built in now. And while Blendervse suggests a more robust workflow (rightly so for quality), of rendering vision from Blender and muxing (joining video and audio) externally. I often need quick and dirty renders of synced effects or videos. So muxing a file in Blender is a necessity.

The one issue I have had with this was a simple corruption of a project that I was working on. At some point the Scene settings for Audio were changed to a null (illegal) value, which killed ffmeg at render time and crashed Blender.

So I suggest that you make sure that Properties > Scene settings > Audio is set to Stereo or Mono before rendering (you can use multitrack too if you use the right output codec).

Properties > Scene > Audio settings

To take advantage of the new Panning effect, that you can apply to a strip (via strip properties), you must ensure that the audio is mono first.

To do that you have to go to Outliner then change Display mode button to Datablocks. Scroll down till you find Sounds. Expand the sound you want and set it to Mono.

datablocks > audio > mono” width=”496″ height=”624″ />