When the epic, studio-created drum sounds of the '80s had to be translated to a stage setting, sample triggering from drum kits came to the fore — and our man Gavin Harrison began his personal quest for the ultimate setup...
During the early 1980s, drum sounds on records began to change. Snare drums mutated from a quick bash into something more like a nuclear explosion, accompanied by bass drums recorded in the Taj Mahal and tom-tom fills resembling rapid cannon fire. These new 'super-sounds' were the result of clever processing — one memorable example is the drum sound on Phil Collins' 'In The Air Tonight', created using a combination of heavily compressed room ambience and triggered noise gates.
The sampling revolution transformed drum sounds even further. At first, samples were used to enhance 'real' sounds, and it became common to add a processed, ambient-sounding clap sample to a snare drum to 'big up' the backbeat. But before long, producers began to abandon acoustic drum sounds altogether, and within the space of a few years, sampled drums and percussion ruled the earth. Inevitably, gigging bands wanted to take advantage of this brave new sound world, but not all were prepared to replace their drummer with a drum machine (even though you only had to punch the rhythm into the latter once). At this stage, many musicians (myself included) began to look for an answer to the question 'how do you trigger samples from a drum kit on stage?'
My personal quest for the Holy Grail of immaculate sample triggering began in 1986, when I went on tour with Iggy Pop to promote his Blah Blah Blah album. David Bowie (Dr Jekyll to Iggy's Mr Hyde) was the producer, and, typical of its time, the album featured sampled drum sounds — big, explosive snares and heavily-treated, massive bass drums. A real kit could not produce those kinds of noises on stage, and to make matters worse, the drum sounds on the album were significantly different from track to track. My mission was to play a real drum kit on stage with Iggy, while somehow also triggering those fabulous sounds from the record.
The master tape was procured, and off I marched to a London studio for a sampling session, proudly clutching my cutting-edge, Hertfordshire-built Greengate DS3 sampler. (This idiosyncratic gadget, actually a card which fitted inside an Apple 2 computer, was cutting-edge technology back then, but seems laughable now — its total RAM memory was 1.5 seconds of glorious eight-bit mono sampling.) I duly sampled Bowie's monster drum sounds from the original multitrack tape into the Greengate and saved them on five-inch floppy disks, half afraid that the Thin White Duke would burst though the door and demand to know what the hell I was playing at.
To trigger the samples from my drum kit, I used miniature piezo contact mics (I can't recall the make, but I believe they were the same ones Simmons used inside their octagonal drum pads). I attached these gizmos to my snare and bass drums and plugged their flimsy cables into the audio inputs of another miracle of modern science, the Roland Octapad. Each time the snare or kick was played, the contact mics emitted an audio 'spike' which the Octapad duly transformed into a MIDI note-on command (MIDI, of course, being another new wonder of the day). The MIDI notes were relayed into the Greengate, which pondered what sample should be played, and about six weeks later, the selected noise would come flying out of its audio output. Hurrah!
Whether triggered samples are involved or not, no drum kit can sound and feel right if it is mixed incorrectly. Sound engineers tend to fixate on kick drums, and I've often stood out front at a concert in which the bass drum is very loud in the PA, but the snare drum (which in real life is much louder) is being left out of the mix. This can happen because the sound engineer feels that the snare drum is acoustically loud enough to be heard without amplification. However, the sonic separation of kick and snare, depending on the size of the venue and the distance of the drum kit from the PA speakers, can cause terrible imbalances of timing and effectively ruin the precious groove.
The reason for this is that sound takes roughly three milliseconds to travel a metre, so if a drum kit is set ten metres back from the front of the stage, its acoustic sound will arrive in the auditorium 30 milliseconds later than the sounds coming through the PA. Even 30ms is enough to really upset the integrity of a groove; if the bass drum is heard through the PA and the snare isn't, the two will lose their all-important timing relationship and the groove will suffer. The same phenomenon affects all loud on-stage sound sources, such as guitar, bass and keyboards. A smart front-of-house engineer could compensate by delaying the PA speakers by the appropriate number of milliseconds, to ensure that the stage and PA sounds arrive at the listeners' ears at the same time.
The contact mics had their own self-adhesive backing, but during 90 minutes of non-stop Iggy Pop thrashing they were liable to jump off the drums and land on the floor. Attaching them to the drum shells didn't work, as that failed to generate enough level for a reliable trigger. The only solution was to fix the little blighters directly onto the drum skins with silicone glue — definitely not ideal from an audiophile point of view, but with the samples belting out of the PA at high volume, who cared about the tone of the real drums?
Unwanted triggering was a big problem — sometimes just the sound of me clicking my drum sticks together during a song count-in would be enough to activate the snare trigger, causing an embarrassingly loud snare sample to come blasting through the PA. Vibration from my feet on the drum riser, loud bass notes (depending on the proximity of the bass cab), or Iggy himself jumping on the drumset from a great height smeared in peanut butter while howling 'I Wanna Be Your Dog', were all events which could set a trigger off. Loading the samples between songs was a nightmare; the crowd would be going mad, Iggy would be screaming 'come on you muthaf***ers, kick it off!' at the band, and I'd be down on my hands and knees staring at a blank green screen, trying to remember which 'key word' to type into the Greengate so it could load the right sample.
Still, when the triggering system worked (which was about 80 percent of the time), everybody was pleased and I was hailed as a conquering hero. Our front-of-house engineer was particularly happy with the enhanced sound of the bass drum (I used a big, crunchy kick from the album, and also a tight, dry Linn LM1 bass drum sample, a useful weapon in overcoming the boomy bass response of some large halls). The audience joined in the celebration too, showing their appreciation by throwing things at me: during the course of the tour I was hit by various items including rolled-up tee-shirts, a woman's blonde wig, a metal ashtray, and, strangest of all, a bible, clearly one satisfied customer's way of indicating his faith in the Godfather Of Punk's music.
Amidst all this unconfined joy, one person was less than impressed... me. The samples always sounded late (when I could hear them — the stage monitoring wasn't exactly reliable), and if the acoustic drum source and the triggered sound were mixed into the PA at equal volume, everyone would hear a pretty obvious flam. What I really wanted was an instant, delay-free triggering system where the sound of the real drum and the triggered sample happened at precisely the same time and became one new super-sound. It seems simple enough, but this divine synchronicity proved to be highly elusive.
When I first played electronic drums in a rehearsal room, I realised that it's a very strange feeling to hit a pad and hear a sample come out of a monitor some distance away. We drummers are used to hearing acoustic drums 'explode' in our faces, and after many years of that experience it's hard to play any other way. For that reason, it's crucial to have a good live monitoring system. I recommend 'in-ear' monitors or headphones as the best way of hearing your samples directly.
Drum-pad technology has improved greatly over the last twenty years. The original Simmons SDSV electronic drum pads were made out of an incredibly hard, shiny plastic substance also used in police riot shields.
Hitting the pad hard with a drum stick sent a nasty shockwave up the arm, causing carpal-tunnel syndrome, loose teeth and various minor personality disorders. The pads' surface also made a loud 'clack' when struck, which would cause problems when the noise spilled onto drum mics. It also pointed up the latency problem, as the 'clack' was liable to flam with the attack of the sample! The Roland Octapad's plastic pad zones had more 'give', generated less impact noise, and were much nicer to play
Nowadays, units such as the Pintech drum pad use mesh drum heads. These are incredibly quiet, and a joy to play from a feel point of view. As you can't really hear the sound the pad makes when it's struck with a drumstick, any obvious flamming with the triggered sample is eliminated, which helps create the illusion that triggering is instant. Clavia also make drum pads with a choice of mesh heads or real drum skins, the latter padded out with foam to kill all resonance. As an alternative to drum-shaped pads, the KAT MIDI controller is a good professional device for percussionists.
If you want to get away from the traditional drum kit, there are now several electronic drum kits out there with agreeable playing surfaces. The Roland V-drum kit (which also uses mesh heads) is very nice to play and reasonably responsive, but the big problem with such units is that you're stuck with the factory sounds, and however great and versatile these seem at first, you'll always want to add something new at some point. (Yamaha's new DTXtreme IIs has on-board sampling). But personally, though I'm impressed by the technology and came close to buying a V-drum kit, I feel most at home playing acoustic drums; their nuances can't really be captured by a set of samples, no matter how comprehensive.
Fast forward ten years to 1997. A leaner, wiser, more experienced drummer takes the stage with Iggy Pop, but he wasn't me — I was touring with soul diva Lisa Stansfield, and still trying to get really tight sample triggering from my acoustic drums. By this time I was using a Trigger Perfect trigger on the bass drum, running its audio output into a Roland Octapad II and the Octapad's MIDI Out into an Akai S3000XL sampler. The Akai contained a sampled bass drum triggered by my kick, and a variety of TR808 samples that I played directly from the Octapad. The results were still (to my ears at least) pretty unsatisfactory — everything sounded late, and the bass drum sample and the real kick still produced their usual horrible flam when mixed together.
In an attempt to get rid of the flams, I tried other triggering interfaces, such as the Akai ME35T. According to the sales pitch, this audio-to-MIDI box was capable of instant triggering, but I discovered that it had a 13ms delay from input to output. When I mentioned this to an Akai salesman, I was told that I was splitting hairs, and that no one could hear that kind of delay. That made me even more obsessed. Determined to hunt down the fastest system in the west, I bought an Alesis DM5 drum module, which has a set of internal sounds and built-in audio trigger inputs, but this also suffered from late timing. I began to suspect that the problem was caused by the very thing that was sent to this planet to save our souls... MIDI. Every time MIDI was involved, I got horribly late triggering and, more worryingly, inconsistent timing. Finally, I decided to use Logic Audio to test the timing response of the Alesis DM5, by recording a live bass drum playing a series of single hits into Logic; feeding the output of the recorded bass drum track straight into the DM5's audio trigger input; playing back the recorded bass drum track while recording the DM5's analogue output onto a new Logic audio track; and then finally repeating the previous step, but this time recording the DM5's MIDI output onto a Logic MIDI track.
Recording the analogue outputs enabled me to look at the DM5's audio response time, while recording the MIDI notes it generated in response to incoming triggers gave me an idea of its MIDI processing speed. Having recorded these tests, I compared the timing of the DM5 recordings to those of the original bass drum hits, and found significant delays in both the audio and MIDI events. I also found that the delays were slightly different each time, even though the recording conditions were identical.
I also tried the same test on my Akai sampler with similar results, eventually concluding that although MIDI wasn't entirely to blame for triggering delays, it was a major contributory factor. Using a triggering device to fire off custom sounds in an external sampler, I could expect unpredictable MIDI processing delays from both units — clearly a recipe for timing disaster. In the end, I found the only fix was to trim the front off the bass drum sample to reduce its attack and minimise the flam effect. Far from ideal, but it got me through the tour!
Whatever the triggering system, accurate sample triggering relies on the manual setting of an audio threshold for each trigger source. The threshold value determines the minimum level at which the incoming audio generates a trigger impulse — all signals below this level are ignored. The level therefore has to be set low enough to pick up quieter hits (not that there were many of those in Iggy's set, for example), but high enough to rule out cross-triggering from adjacent triggers or background noises. With multiple drums triggering samples, crosstalk becomes an even bigger issue, as a loud hit on one tom-tom can easily cause its neighbour to trigger. Onstage background noise is often extremely loud, so getting the threshold settings right in a live situation can be very tricky.
Another problem is that trigger boxes sometimes over-react and generate two or three impulses in response to one loud hit; some units cleverly overcome this by supplying a 'hold' function which can be set to prevent multiple hits occurring within an excessively short period. I've learnt that setting up live drum triggers is a lot like making love to a beautiful woman — it requires a lot of patience, determination and fiddling, and you seem to have to tweak the damn things almost on a daily basis.
By the time a second tour with Lisa Stansfield came around in 2001, I had bought a Clavia ddrum4, a professional drum module from the company who make the Nord synthesizers. The ddrum4 has a large menu of high-quality mono sounds, 8MB of Flash RAM for user samples and its own set of built-in audio trigger inputs, enabling internal samples to be triggered directly, without MIDI being involved (having identified MIDI as a source of timing slop, I was happy to see it eliminated). But even with direct audio triggering, the ddrum still outputs its samples slightly late — 4.6ms, to be precise. The good news is that this delay is at least consistent, so once you find a sample that blends well with the acoustic drum source, it will sound the same each time — a major breakthrough! Although 4.6ms sounds insignificant, it's enough to cause samples with really sharp attacks to flam with the acoustic source, but by choosing samples with a less pronounced attack, you can get satisfactory results. Thanks to the ddrum, I was able to use a triggered bass drum sample all the way through the tour with no problems.
Clavia have shown their commitment to sample triggering by manufacturing their own, very good drum triggers — these clamp onto the rim of the drum and have built-in jack or XLR sockets so that you can attach your own cables.
Being able to use your own samples in the ddrum4 is a great selling point, but loading them in is pretty painful; the unit doesn't have front-end sampling, so you have to use an external sampler to transfer your samples into the ddrum4 one at a time. This process relies on that slow, tedious transfer method known as MIDI sample dump, which, along with hexadecimal code and haemorrhoids, is one of the greatest pains in the rectal zone known to mankind. Using MIDI sample dump, it takes a long time to transfer even a very short bass drum sample — having got to the point where I thought I could dispense with an external sampler, it was ironic to discover that I still needed my Akai S3000XL to perform the transfers!
Once my drum samples were resident in the ddrum4, things started to look good. They got even better in 2003 when I discovered a really fantastic bass drum pedal made by US company Axis, who have invented their own ingenious system of mechanical triggering. Axis trigger devices are quick off the mark — so much so that they can actually be set up to trigger a sample before the bass-drum beater hits the skin!
Here's how it works: that funny-looking miniature hammer's thin metal arm (right) is attached to the central crossbar. As the pedal is pressed, the crossbar rotates towards the head and the beater (which is just out of view) is thrown forward to hit the drum. At the same time, the tiny hammer moves down and strikes the surface of a black steel disc, visible on the left of the picture. The trigger sensor lies just underneath the surface of this disc. Using an Allen key, you can adjust the angle of the hammer's arm so that it hits the trigger disc fractionally before the pedal strikes the bass drum. The triggering mechanism has a female jack socket on its underside, from which a cable can be run to one of the ddrum4's trigger inputs.
After some experimentation with the hammer angle, I was able to effectively 'advance' the trigger point and eliminate the ddrum4's 4.6ms delay, thereby achieving instantaneous sample triggering in real time. Eureka! The Axis mechanism is very precise, so accurate triggering does depend on you having a nice clean bass drum technique. Sadly, no-one has yet manufactured viable 'advance triggers' for snare drum and tom-toms — one solution would be for a drum stick to break a beam of light as it descended, but the potential for cross-triggering would be enormous. However, over in nutty California, a friend reports seeing a small group led by US synth designer Don Buchla using light-beam technology to trigger percussion samples, the joke being that samples of cymbal, snare drum and cowbell were sounded when the player's stick hovered a foot in the air above the instrument in question!