Sampled drums lacking realism? Synthesize stereo room mics to add authentic ambience with Sonar's built‑in effects.
One of the most sought‑after sounds in rock is a drum kit recorded in a room with great ambience: think Led Zeppelin's 'When the Levee Breaks', where John Bonham's kit was located in a large open stairway. Two Beyerdynamic M160 microphones were placed at the top of the stairs, and the spacious quality was further emphasised by compressing the sound.
I've been working with room mics lately, but in a very different context to Led Zep, with classical music by solo performers. Working in a studio with a superb recording room always reminds me that good room ambience is one of the qualities most lacking in a lot of music created in project studios, simply because there isn't a room, or at least a good one. Instruments are often recorded direct, and for those using sample libraries, although some have drums recorded with ambience, it's not always the ambience you want.
However, by analysing the characteristics that make up 'room sound', we can synthesize a stereo set of room mics to beef up a track. You can use this alongside standard reverb to give a fuller sound, but in many cases you might find that the 'artificial room' is all you need. So let's take a dry drum track and do some magic. If you want to follow along, insert Session Drummer 2, load the Hard Rock Tight kit, and start playing a pattern.
One of the main characteristics of an acoustical space is the quantity and diversity of early reflections, which are challenging to simulate with electronic reverb. Some get around this by placing a speaker in a bathroom, basement, or other space with hard walls, sending some audio to the speaker, and picking it up with a microphone to augment any existing reverb sound. We'll assume you don't have that option, and need to create those early reflections artificially.
The strategy is to first set up a room sound using reverb, then augment the early reflections electronically. Several send buses will be required, as we'll need to trim levels precisely, and add processing independent of the drums. To simplify matters, we can use a mixed stereo output from the drums and apply the 'room mic' to the entire kit.
1. After inserting the drums, create a bus and call it something like MainRoom.
2. Create a send on the drum kit's audio track and assign it to the MainRoom bus. Set the send to Pre‑Fader, and turn down the drum track fader, so you can concentrate on the room sound only.
3. Insert your basic room‑sound reverb of choice in the MainRoom send bus FX bin. I use the Perfect Space convolution reverb and load the 'Open Sounding Blues Club' impulse (VSTplugins / Perfect Space / Real Spaces / Blues Club / Open Sounding Blues Club – XY Rear).
4. Create two more buses, one called Reflections1 and one called Reflections2.
5. Create two sends on the MainRoom bus. Assign one to the Reflections1 bus, and the other to the Reflections2 bus. I usually choose Pre Fader for both of these.
Now we need to insert a delay-line effect into each of the two Reflections buses. Sonar has the perfect plug‑in for this application: the DSP‑FX Delay, which has four independent delay lines, each with Delay, Feedback, Level and Pan controls. Adjusting these parameters can create a variety of early reflection effects. To achieve the most open sound, I try to avoid delay times that will build up into resonances, so the Reflections1 bus has times of 3ms, 5ms, 7ms and 11ms, while Reflections2 has delays of 13ms, 21ms, 23ms and 25ms.
Adding feedback to each delay increases the sound's complexity. You can get away with as much as 40 percent, but with more than that you'll probably start to hear an unrealistic, 'boingy' tonality. Start with all delays on around 20 percent feedback, then try increasing from there.
As we're trying to synthesize stereo room mics, panning is essential to creating a sense of space. I tend to pan the two longest delay lines extreme left and right, and the two shorter delays about mid‑right and mid‑left. Initial levels for each delay line are up at maximum, but even slight changes in the panning, delay or delay‑line levels can make a major difference to the overall sound.
We now have four different components to the drum sound and they need to be mixed in the right proportion. Here's one way:
1. Start with all four faders down. Bring up the dry drums first, in context with the other tracks, and set the level just a little bit lower than optimum.
2. Solo the four drum tracks.
3. Pan Reflections1 about 80 percent or so left, and Reflections2 an equal amount to the right. Bring them both up, and adjust the levels for a good stereo balance. Trust your ears on this, because setting them to the same level may tilt the image to one side or the other of the stereo field, in which case one bus may need to come up or down a little bit compared to the other.
4. Once you have the balance set correctly, assign the two Reflections bus faders to a group so you can control them together when doing your final tweaks.
5. Bring up the MainRoom bus fader to add body to the sound from the Perfect Space reverb. You probably won't need much. Go back and re‑tweak the four drum components to optimise the sound.
For each Reflections bus, add a compressor after the Delay (the Sonitus Compressor works well). There are specific requirements to make this sound right: you don't want compression to 'smear' the sound, but accent it.
One key is setting the attack at around 4‑5ms to let through the 'crack' of the initial drum transient. Even more importantly, turn off TCR (or, with non‑Sonitus compressors, whichever parameter gives automatic control over release time), and set a short release time of around 20‑50ms, or even less. This allows the compressor to recover quickly after the signal goes back below the threshold, retaining a crisp, percussive character while still applying significant amounts of compression to the overall sound. The end result brings up the ambience without losing any percussive qualities.
Incidentally, if you want to expand your palette of drum‑friendly compressors, download SSL's free (yes, free) LMC1 plug‑in (VST Windows, AU Mac for Power PC G4 or better, but not Intel) from www.solid‑state‑logic.com/music/LMC‑1/index.asp. (This is a version of the 'Listen Mic Compressor' that was built into the SSL SL 4000E and first used for drums on Peter Gabriel's 'Intruder'.)
Another useful tweak is pulling back the MainRoom bus high‑frequency response to prevent the highs from competing with the dry drum track (one of the characteristics of ambient sound is that high frequencies are less pronounced compared to the direct sound). This needn't be drastic; I typically use one stage of EQ set to low‑pass, Q around 1.0, full cut (‑18dB), and a frequency of between 10kHz and 15kHz. Be careful, though, as this also affects the high frequencies in the Reflections buses.
After making these tweaks, revisit the drum component levels. You may find that the MainRoom component isn't necessary when the Reflections buses have compression, or that you can bring down the dry level and use only the Main Room and Reflections sounds. In any case, your drums will now sound bigger, fuller, and more natural, so it's worth the effort.
Visit sonaraudio.mp3 to hear audio examples illustrating this technique. Four bars of drums are presented first dry, then with early reflections added (the heart of the process), then with early reflections and some room reverb.