For me, I start with a 200Hz hit which I then layer with a transient which can vary what frequency it hits at, I'd say roughly between 500Hz - 1kHz maybe? I then round off my snare with either a noise from a synth, a clap or an acoustic snare layer which can be lifted from a drum break or from a drum sampler like Addictive Drums or Superior Drummer.

Cool, I dig it. I usually load up a Drum Rack in Live and find complimentary samples from all forms of sources: live instrumentation, sample packs, NI's Battery.

Apart from sample selection, I've been finding more and more that compression on the drum bus is of paramount importance in not only making the drums glue together, but also enhancing the groove. Taking those original drum layers and using parallel compression on them, then bouncing them out is something that beefs up the percussive elements almost every time.

I like to go one step further and be fancy by adding foley sounds to the claps and hats. Everyone by now knows that deadmau5 used his own naked ass slaps as claps on his early stuff - that's one form of foley - but using an apple crunch or someone biting down on chips can also really enhance the attack of your downbeat (I hear this quite often in Amon Tobin and Koan Sound material).

I prefer to work with straight audio samples being dropped onto the grid in ableton. I'll find samples with parts that I like, for example one hat sample has a transient sound I like and another has the sustain/decay that I want and I'll stitch the samples together and fuck around with the crossfades in ableton until I'm happy with the sound. Freeze and flatten (bounce to audio) and use that new sample. I find that layering samples can sometimes just become a complete mess and this method will generally sound cleaner. That's not to say that I avoid layering, I just find this method works better the majority of the time.

I end up smashing a bunch of samples together (up to 5) with minimal EQ. Each one plays a role - some are for variance (use them only on specific hits), some are for punch, some for rolls, and often I'll put them all together on important hits. I layer them with claps as well and my snare sounds have been pretty interesting thus far.

Foley, acoustic, and analogue sounds (like from drum machines) are great for layering because they'll be interesting but simple enough to not take up the whole spectrum fully.

For the 'standard' snares I make I usually compress a 909 to get the 200-hertz body and a bit of top end, then I layer with some acoustic samples and make a transient that I mix in. Imo the transient is what makes or breaks a punchy snare, if the transient is weak, the snare's not gonna be as punchy.