- Pan layers L/R
- Play multiple octaves
- Chords
- Detune saw waves, place a LFO on it (~1%)
- Flangers, phasers, sausage fattener
- decrease Mod Envelope pitch below 0.0, maybe like -.5 (play around with it)

FWIW I have found that the "theory" of making good supersaws isn't so much as how it is made *technically* as much as it is the chords/note combinations that you play together. Some chords and combinations of notes just sound better with supersaws than others.

any stereo widening effect helps in my opinion.

you can copy the channel, pan one left all the way and the other right all the way, then give some a few m/s delay and you should hear it widen up. Alternatively, you could download ADT by vacuumsound, found here:

http://www.kvraudio.com/product/adt_by_vacuumsound

Don't know if it helps but one trick I always do in my supersaws is put a turned down square wave in them. I feel like with most saws it's the bass that makes the biggest difference.

I'm curious to know what everyone does to make their supersaws hit their audience in the face. What techniques do you use?

I have used 3 layers. Each layer does something different. I've used one layer for the mid to low portion that is detuned but with not much stereo spread. Then I use 2 higher frequency layers. One has some detune but not a lot, and moderate stereo spread. I sometimes try to use this layer for the attack/plucky part of the sound as needed. The top layer is detuned quite a bit and has maximum stereo spread. The amount of detune depends on your taste. You could also add another layer in the low mids that has a more plucky sound if you desire more attack. As far a reverb?  I add the most reverb to the layers with the most detune and stereo spread. Kind of a cumulative effect. Reverb is not as critical on the lower frequencies. The high frequencies usually have quite a bit of release as well so I use reverb cautiously.

Rob

I told my lil' supersaw to drink a glass of milk everyday so he would grow of big.

Worked for me.

- White noise (Can't believe no one mentioned it yet, such a big difference!)
- Several layers (each layer fills eachother up at different frequencties) (can put effects such as phasers, saturators on some) and change what they play. For example the top pads play high notes only.
- Low cut around 100hz with a soft slope. That means overlap with the bass pad. So many cut around 250hz.. don't do it.. unless you want a high hz pad, then cuts lots.
- Send to one bus and compress (+ whatever effects you like)
- Panning (L + R)
- Choosing the right sounds.
- Lots of reverb (send channel to bus channel)
- Delay (send channel to bus channel) 
- A->B between mono and stereo (phase issues)
- Proper EQ (m/s)
- Pluck playing with pads

+ some already mentioned above, just play around.

That's more or less how I do it. Effects varies on the channels depending on the sound or what I want to archieve.