FIG. 1: Using two mono reverbs can be a good alternative to using a stereo reverb. In this example from Mark of the Unicorn's (MOTU's) Digital Performer 4.12, the source track is bused to two mono instances of MOTU's eVerb plug-in, which are panned left and right.
Anyone who has ever mixed a piece of music knows how important ambience is. Adding space to an instrument or to a vocal track can bring it to life and give it a sense of depth. For many years, delay and reverb were used heavily and often without much subtlety. From the tape slap and echo chambers used in '50s music to the over-the-top effects of '80s pop, conspicuous ambience was frequently applied.
Musical tastes have changed since the '90s, and it's become fashionable in pop music to mix the vocals drier. Clients and producers specifically request less reverb, but they still want to hear space around a vocal, a dramatic coloration for a solo instrument, or a heightened sense of separation between foreground and background. Applying ambience has become trickier than it was in the old days — you can't just crank up the reverb send anymore. Instead, you have to be subtle and use a range of techniques.
This article discusses ways to tweak reverbs and delays to create other ambience options that go beyond the “same old thing.” Chances are that your current setup has the potential to dial in unusual and original effects.
Reverb and delay are the two basic building blocks of ambience. Reverb is designed to simulate the sonic characteristics of a real physical space: room type, size, and surface. After selecting an appropriate setting, you can then determine its virtual size by manipulating the decay time.
One way to achieve a subtle yet effective ambience is to drastically reduce the decay time of a reverb patch. More intense presets such as churches, large plates, or concert halls — which typically have decay times of three to four seconds — work best for achieving subtle ambience. Decrease the decay time to 0.5 seconds, and increase the predelay to 100 ms or more to create a sound that's spacious without being wet. That technique allows you to retain the sonic character of the space in a shorter, more concentrated package, although you may have to increase your send levels to make those effects work in your mix.
Many reverbs also include EQ controls. Highpass and lowpass filters are the norm, but some processors have even more elaborate EQ capabilities. Using a lowpass filter to reduce the high-frequency content of the ambience can quiet a noisy track. Cutting highs in the reverb on an acoustic guitar track, for example, can subtly deemphasize occasional finger squeaks. Similarly, a not-so-bright reverb may be a better choice for a snare track that has lots of hi-hat leakage. At the other end of the sonic spectrum, adding highs to a short, intense reverb can increase the impact of the effect.
More elaborate reverb processors allow you to go even deeper and manipulate the balance between the room reflections (called early reflections) and the sustained reverb sound. You can experiment with creating ambiences that are all reflections, all reverb, or a blend of your choice. Experimenting with ambiences also allows you to see how your processor is able to simulate so many different spaces with so few controls.
When working in a DAW environment, it's always tempting to use multiple instances of your most expensive reverb plug-ins — the ones that have the most features and the longest list of presets. Unfortunately, those plug-ins can significantly tax your CPU. Even if you have a reasonably powerful system, you may still need to find effects alternatives as you run out of processor power. One way to lower CPU usage is to set up the reverb as a bus effect and send multiple tracks to the same plug-in, rather than using individual inserts that require an instance of the plug-in for each insert. You can also bounce tracks to disk or freeze them once you've applied effects.
FIG.2: In MOTU's Digital Performer 4.12 (pictured above), a single reverb plug-in is being fed by three different-length delays, which are functioning as predelays.
Another way to save CPU resources is to use two instances of a power-efficient mono reverb plug-in to create a unique stereo effect (see Fig. 1). That approach offers interesting sonic possibilities and also works with hardware processors. If your hardware effects unit has multiple effects engines, then you can build a powerful stereo patch with an individual engine dedicated to each side of a stereo return.
Whether in software or hardware, start with each side of a stereo return set to the same patch, and then modify one side. One way to accomplish that is to add a hefty amount of predelay, which results in the reverb sounding as though it's moving across the stereo spectrum. Keep in mind that the ambience doesn't have to be symmetrical. If you choose a predelay time that has an exact rhythmic relationship to the tempo of the song, then you can create an ambience that throbs with the beat.
Delay times that are synchronized with a song's tempo can help generate bounce and flow in a mix. To make that happen, you'll have to accurately compute the delay times that fit your song's tempo. Many delay plug-ins automatically calculate that for you. If, however, you are using a hardware-based processor or your song was recorded into a DAW without a click, then you may need to calculate the delay times. Shareware or freeware delay calculators are available on the Web (you'll need to know the song's tempo to use them), or you can use simple math (see sidebar “Finding Delay Tempos”).
If you mix live music, then you may have developed a knack for turning the delay knob at the right moment to send a phrase or word into a sonic spin. When mixing in the studio, however, there are more accurate ways to set a delay to kick in at a precise moment. In a DAW environment, one method is to generate a dedicated delay-effect track.
To make a dedicated delay-effect track, create a new track with the same output assignments as those on the original track. Copy and paste the section (a word, note, or phrase) that you want delayed into the new track. Set the track's aux send to a reasonable level, set the aux send to prefader, and then turn the track's volume all the way down. Put a delay on the aux bus or aux track (depending on your software), and set the delay to 100 percent wet with a tempo setting that works with the song. When the song arrives at that section on the delay track, the delayed signal from the targeted section will play back precisely at the appointed time.
By applying long delays to whole phrases, it's possible to create rhythmic call-and-response effects. To achieve cascading rhythmic repeats, increase the amount of the regeneration or feedback. Higher feedback levels increase the number of repeats. Shorter delay times increase the frequency of the repeats.
Mixes featuring rhythmically delayed phrases are in fashion in contemporary pop music. The delayed portion, however, is often not a sonic replica of the original line. It's often treated with an extreme effect, such as telephone EQ, deep distortion, or a rotating speaker sound. An extreme effect can help focus attention on the delayed line.
Despite the popularity of rhythmically timed delays, they don't benefit every song. Sometimes it's better to throw away the bpm chart and dial in the delay that feels right.
PUTTING THE PRE IN DELAY
One very useful way to tweak a reverb patch is to postpone its onset with predelay. A reverb with predelay adds simulated depth to a close-miked recording because it imitates the delayed reflections of a larger space. A definitive separation between the dry sound and the simulated reflections of a reverb helps to maintain clarity even with an extreme effect. Many engineers set predelay values that are related to the tempo. (Usually you have to use shorter note values than quarter-note or even eighth-note delays; otherwise, your predelay will be too long.)
Unfortunately, the predelays built into many reverb plug-ins and processors sound muddy and indistinct. One way to get a cleaner-sounding (and potentially longer) predelay is to use a dedicated delay line before the reverb. Turn the reverb's predelay off, and dial it in from the dedicated delay processor or plug-in. If it's patched the right way, then it's possible to make one reverb return do double or triple duty with several different predelay values (see Fig. 2).
I created a short, predelayed hall reverb and panned it to the rear channels recently for a surround mix of a track featuring a spunky, percussive horn arrangement. Every horn stab from the front resulted in a short reverb kick from the back of the surround space. That effect works in a stereo environment as well and can give a horn section (or other percussive element) a memorable ambient space.
A classic technique for creating depth without an obvious reverb effect is based on a popular Eventide Harmonizer patch called Micropitchshift. The idea is to create a slightly detuned, slightly delayed effect on either side of the primary signal. It's best if the delay values are small and are different for the left and right sides.
Delay times of 12 to 14 ms and detuning as slight as plus or minus 5 or 6 cents can be effective. Traditionally, one side is pitch-shifted up and the other side is pitch-shifted down. I recommend configuring that as a bus effect rather than an insert because a little bit goes a long way. If you're too heavy-handed with that effect, then it will send your song straight back to the '70s.
WORKING THE MIC
Some ambience qualities are determined before the signal even reaches the microphone. The sonic character of your recording space impacts everything you commit to tape or disk. Take advantage of a great-sounding room to create ambience that is based purely on mic technique.
FIG. 3: Using differing mic distances creates a natural sense of space between the lead and background vocals.
If you record a lead vocalist three-to-four inches from the mic and the backing vocals six-to-eight inches from the mic, then the lead vocal will always sound more present or up front than the backing vocals. Even in mono, the difference in depth is clear (see Fig. 3). The farther the vocalists are from the mic, the more the mic will pick up the ambient or room sound.
THE FULL MONTY
This article has mainly focused on techniques for creating small, intense spaces that are subtle. But when the time comes for that big-ballad lush reverb, lots of predelay combined with a rich vocal plate is the way to go. Predelay values in the neighborhood of 110 to 120 ms produce a clear distinction between the vocal and the space around it. That is the time to bring on your most luxurious sounding reverb.
A real plate reverb or a live chamber can make this ambience truly spectacular. If you're working in a DAW environment, this might be the time to crank up that CPU-hungry, convolution plug-in.
When working with ambience processors, the bottom line is that there are no rules. You do want to keep the ambience subtle most of the time, but don't be afraid to try some arrhythmic delays and asymmetrical effects. An unusual or a seemingly inappropriate effect can sometimes rocket your mix to exciting and unexpected heights. A willingness to experiment and to take a counterintuitive approach can lead to new discoveries and push your music into uncharted territories.
Julian McBrowneis an engineer, producer, and self-styled digital-audio guru who lives in southern Vermont.
FINDING DELAY TEMPO
If you're working in a DAW and your song is in sync with its metronome, the software can calculate the delay time for you. But what if your song wasn't recorded in a software-based environment or recorded to a click? Below are tips on how to figure out delay times in ms so that you can set your delays and predelays.
To calculate the tempo in beats per minute (bpm), count the number of quarter notes that go by in ten seconds, then multiply the result by 6:
6 × (number of quarter notes in 10 seconds) = bpm
Next, divide 60,000 (the number of ms in a minute) by the bpm of your song. The result will be the length of a single quarter note expressed in ms:
= (quarter-note value in ms) bpm
You can multiply or divide that number to find the ms value of different note lengths:
(quarter-note value in ms)
= (eighth-note value in ms) 2
2 × (eighth-note value in ms) = (half-note value in ms)
Assuming that you're in 4/4 time, multiplying the quarter-note value by four results in the whole-note value. If you're in a more relaxed dub-style mix, you might want to know the quarter-note-triplet value, which can be calculated by dividing 40,000 by the bpm:
= (quarter-note-triplet value in ms) bpm
Once you've found some useful values, jot them down in your mix notes and use them for all of your effects settings.