Acoustic drums are arguably the most difficult instrument to record and, while relatively easy to sample, are very difficult to fake. For home recording musicians like me, getting realistic drums in the mix is probably the biggest hurdle to laying down a convincing track. The problem is two-fold: first, the audience has to believe a human is actually playing the drum set; and second, the samples must sound authentic, especially the snare.
In my case, tracking live drums is out of the question – the number of things that have to come together is pretty daunting: a properly tuned drum set with the right sounding snare, a drummer who has learned the song I want to track drums for, a sound engineer, a nice sounding room and a DAW with enough inputs for the four or more mics it would take to lay down the proper tracks. Plus, if I decide I want to change something (extra chorus, different fill on the intro), there’s nothing I can do about it. Because of this, I’ve always used free soundfonts or the bundled soft synths that came with Cakewalk Sonar. I programmed them manually or used Groove Quantization to get a more human feel, but the sound of the samples was always lacking.
Deciding to spend some money on decent drum soft synths, I got recommendations from some friends and began comparing the sounds. I looked at Wizoosounds Acoustic Drumkits, BFD, EZDrummer, Drumkit from Hell 2 (DFH2) and Drumkit from Hell Superior (DFHS). I downloaded all the demos I could find for each kit, and after listening to them all I determined that DFHS had the biggest sound, the most authentic-sounding samples and snare timbres that I liked the best. A friend of mine confirmed that like BFD, the 35 GB or so of DFHS’s samples were not processed, so that was another mark in its favor – I wanted to start with raw samples and process them as needed for each song. The only thing I didn’t like was the price – $300 – but BFD was the same price, and the demos just didn’t sound as good, so I got out the credit card and made the order.
Installing DHFS was definitely the hardest part of using the kit. DFHS has two interfaces for host programs, Rewire and VSTi. Rewire wasn’t supported with Sonar, so I attempted to set it up with VST instead. The Sonar template provided on the DVDs didn’t work, but the example project did, so I created a new template from the sample, adding common tracks and busses that I usually work with. I had to import all the audio from my pre-DFHS songs to fresh projects, but the task was made easier because DFHS provides some Cakewalk CAL files to automate the conversion of MIDI drum tracks from GM mapping to DFHS mapping.
Working with the actual interface was fairly easy. There are three sections – Construct, Main and Bounce. The Construct page is where you select the drum samples to load for the song. DFHS has 3 VST instruments: Drummer, Cocktail and Percussionist. Drummer is the main set and comes with five basic kits. However it’s easy to create your own kits and you can select from six kick drums, 17 snares, 17 toms, eight hats, seven ride cymbals and more crash cymbals than I cared to count. You can also “prune” the samples loaded with your kit to cut down on memory usage. (Loading a full kit can easily take 2 GB or more of memory.)
The other big feature of the Construct view allows you to preview as much or as little of the final quality of the DFHS samples as you like. DFHS has 12 virtual mics that represent the recording setup in place when the samples were recorded by real drummers. While working in the Construct and Main pages, the sound you hear in your project typically won’t be the true quality of the final samples, because when DFHS bounces the drums to WAV files (which you can do on the Bounce page), it adds in “mic bleed” from all of the other instruments. In the Construct view you can turn this mic bleed on (for a better-quality preview) or off (to save some CPU cycles).
The Main page gets into the details of controlling each instrument (called a super-pad), including mic level and ambience, fade times, MIDI note assignments, hard and soft velocity cutoffs and even pitch. Each super-pad has sub-pads as well, and each of those can be individually controlled in the same way for the truly obsessive. Clicking on each pad will produce a velocity-sensitive preview of the sample, and each pad can be soloed or muted from this interface as well. The Main page also controls the humanization features, which randomize the samples played back by DFHS.
The Bounce page handles options for bouncing DFHS VST tracks to audio. The process itself is simple enough: press record, play the song in your host software, select some output options and hit “Bounce”. But like the other two pages the Bounce page offers some extras for the control freak in all of us. Select all the splitting options and DFHS will output a close-mic and bleed WAV files for the snare (top and bottom), hat and toms (except rack tom 3) and a separate stereo WAV file for each ride and crash cymbal. There’s also a stereo ambiance track created with every bounce. And if you really want to impress your friends, you can set your preferences to split stereo as well.
With all these features I thought DFHS would take up a lot of resources to run, and in fairness, there are some option combinations that’ll take down just about anything (turn on “Semi Seq” in the Main page and all the mic bleeds in the Construct page for some serious CPU burn). But with sane configurations DFHS added a maximum of 20% CPU overhead on my P4 2.8GHz (hyper-threading on) with 1 GB of RAM. Not a trivial amount, but I can still mix some pretty big projects without bouncing the drums first.
The bottom line is that DFHS is relatively easy to use and highly configurable, and the output is stellar. If you can swing the $300, don’t look anywhere else.
