I use FL 8 XXL Producer edition, Sonar 8 Producer edition, and Soundforge 8.0b, though I also own Cubase SX3, both Live Lite 4.0 and Live 7.0, and Reason 4.0. I like to keep my options open and be compatible with just about anything that crosses my desk.
Keep in mind, I produce Hip-Hop and R'n'B mostly.
I produce predominantly in FL and Sonar, using FL for sequencing and FX on the main production, ie. beat/backing track. I use Sonar to mix and master the final track, recording and mixing vocals inside Sonar and mastering the final mix via Ozone, Sony/Oxford, Modern, Sonalksis and other various programs, dependent on what is needed to balance the track and put it in the right place for where and how it will be presented.
On the backing track side, I generally start with the drums, laying in a simple beat to start and then bringing out the Nanopad to track the actual drum parts a la MPC. I like to layer 2-3 good kicks, using EQ on each drum to meld it into one, cohesive sound, then strapping compression across a Buss with all three channels routed to it to really sync them up. I'll then lay in clap/snare/main percussion sound, possibly layering those in the same manner. Alot of times I will choose not to layer a snare/clap because I like the sparse sound of a single sample. Next is the rest of the percussive elements, congas, bongos, toms, hats, shakers, tambs, etc. All channels will be mixed for balance, eq'd for color, then all drums are sent to a final drum Buss and a drum compressor is strapped across them to push them all in line with one another and give further cohesion of the disparate samples.
Now I move onto the main part of the beat, whether it be bass for a more southern or west coast beat, melody and harmony for a lyrical style track, or samples for that east coast grimey vibe, or just any of those three for the original joints when I am not on contract or haven't been given an overarching theme by an artist.
For bass, I'll choose a solid sound to start with from one of my favorite VST's, then tweak away at the preset until it is giving me the feeling I want. I rarely use 3xOSC because of it's aliasing problems, though on an old school joint I might actually use it with a bitcrusher or dithering to give it that classic, noisy feel. I've got a series of FX that I really love to use on bass, but I won't give out the exact fx chain I use.
For melody and harmony parts, I start much like the bass, with a VST I like to use on that section. Right now I'm using the FabFilter Twin 2 ALOT, as well as Minimoog V and Alchemy. For pads and drones, I have always been an AbSynth 4 fan, and for strong, easy to work with, Lead parts, I go with any of the Rob Papen stuff, from Blue to Predator. For that matter, Predator has some excellent ability to produce good bass parts.
As for samples, I deal mostly with 70's and 80's funk, R'n'B, and the rare Blues joint. I have also worked with some of the early 50's Blues and Jazz pieces. I don't sample loops. I am a microsampler, choosing to sample the little bits and pieces like a single horn stab, a piano chord or single key, a guitar strum, etc. I rarely reconstruct the original sound of the sample, though I might intro a track with a pitched version of a sample. Generally I use the microsamples to put together a new interpretation of the older work. I don't use breaks, though I do sample out the individual hits in a break for later drum use.
To finish a track up, I take all of the variations of the melody, harmony, bassline, drums, etc. and put a simple pattern together, then add "glisten" to the track by throwing in sound fx, stabs, pads, etc. to changeup the pacing through certain sections. I will then throw a 15:1 ratio brickwall limiter across the master, followed by a Sonnox Dynamics with only the Warmth at a comfortable level, and listen through the track probably close to 100 or more times over the course of a day and take notes on where the pace is too dry, what alternate arrangements I can go with, how much/how little reverb is needed to smooth it out, where chorus is needed, what instruments need what fx added, where the EQ is muddied, etc.
After re-mixing the track from the notes, I'll kill the limiter and warmth on the main, load a Stereo Tools up to check phasing and spectrum, then slap a Modern Multi-Limit behind the Stereo Tools followed bby a second Stereo Tools. I use the Multi-Limit, with 5 bands, to even out the frequencies a bit if there are any freqs which are a bit on the weak side, never pushing a band beyond 1.5dB up or 2.5dB down. If a band needs to be pushed beyond those limits, I track down the offending instrument(s) and reevaluate their EQ structure. I use the second Stereo Tools to visually reference that the Multi-Limiter has evened the frequencies and when I am finished, I drop both Stereo Tools and export the finished track as both a FullMix and the multi-mix with only the individual Busses as stems in case I need to re-mix later.
Then the artist will hear the Final Mix and start writing. After they've written and are ready to get into the studio to record, the fun begins. The individual busses are ported into Sonar and put on their own channels. The FinalMix is put into the first mix channel as a reference mostly, and the recording begins.
I generally have artists record the verses before moving to the chorus, intro and outro, but it's up to them if they have some idea that they wanna try. After all, they are paying. We loop the section they are recording and begin recording takes. I don't like to comp much, so minor mistakes are ok, but I really hate throwing together bits and pieces from 5 or more takes to get a full verse, so most of my artists just retake and retake until we get a great run-through. After a verse is done we throw down seconds, then adlibs. I let most artists do their own thing and put their own stamp on it, but I will rein in some of the stupid shit. There will be a sign up in the studio I'm building right now that says "You ain't Jeezy and you ain't Lil Jon, mufukka!" I don't do the "Yeeeeeeahhhhhhhh!" shit.
After the verse mains, seconds and adlibs are done, it's chorus, intro and outro time. Yet again, seconds, adlibs, etc. are all recorded in loops and takes. After the whole thing is done, I start comping, cutting, moving, arranging, layering, and throwin' the FX for the artist's vocals down. I have some presets for particular artists which are good starting points to give them that signature sound that they have on other tracks. I keep a signature EQ, Compression level, and Reverb/Delay Send channel setting for most of the artists who come through, each of which is merely a starting point and gets some tweaking per track, but really gives cohesion to recordings across their albums, mixtapes, etc.
After the FX chains are layered in, then comes the mixing of the vocal tracks, instrument tracks, drums, etc. I will generally mix drums solidly with no automation. Instruments get minimal automation as most of their volume and panning automation should have been done in the beat mixdown. And finally I mix in the vocals and give them the most time "riding the faders", as they are the most important aspect of the music. At this point, I start looking at my EQ bands for conflicts between vocals and instruments and begin programming a second layer of EQ on instrument busses to notch out a little bit of the conflicts, generally in the mids and highs. I will automate the EQ a bit so that during vocal dropouts the instruments won't sound as thin. After the EQ and volume automation is taken care of, it's time to mixdown.
Now it's time for Stereo Tools-Multi-limit-Stereo Tools again. Remember, I used the instrument busses so there was no limiting on the beat carried over from the mix. Throw a brickwall on it, give it a tiny bit of gain, and export it as mp3 for the artist's reference and .wav for archival. That's about the extent of it and I almost used up all of the characters NG gives you, so...
For TL;DR, I do alot of magic, measure twice, and cut once.
