In last month's T&P, we examined reverb and some of the ways that you can manipulate it for maximum effect (pun intended). This month, we'll dive into the bouncing world of delays and how you can use them to enhance your mix. As we did with reverb, let's briefly revisit the parameters available in a typical delay, whether it's hardware or software.
Delay time: the amount of time between each repeat, usually expressed in milliseconds (mS).
Feedback: generally controls the number and intensity of repeats (some delays feature a "Number of Repeats" parameter for more precise control). Feedback can sometimes be set to a negative value, which puts the delay out of phase with the original sound. This may not be audible with longer delay times, but will be critical with delay times between 0 and 15 mS used to create flange or chorus effects.
Filters: at the very least, you usually get a low-pass filter. The LPF does two things: it emulates ye olde days when echo came from tape machines, and it helps keep the delay from masking the original signal. I typically chop off everything above 4 or 5 kHz and everything below around 300 Hz (using a high-pass filter), which sonically separates the delay from the original sound.
Depth: the first "modulation" parameter. The word, "modulate," means "to change." In this case, we are talking about changing the delay time while the delay is happening. Why? Because it sounds cool and provides an out-of-tune wobble to the delay. Depth is the severity of modulation.
Rate: the second "modulation" parameter. Rate is how fast the delay time is changing. Rate and depth can be used to create pitch-shift, chorus or flange effects. (We're going to examine modulation effects in more depth next month).
Multiplier or x2: On older hardware units, you may find that the delay time can be varied up to a point (say 500 mS). To achieve longer times, you push the x2 button, which doubles the delay time but halves the sample rate. (The required memory is the same). You lose fidelity, but that's not necessarily a bad thing.
Tap tempo: a front panel button allowing you to tap along with the music to get a delay that repeats in time with the song. (In the old days, we had to actually calculate delay times. Yikes). You can usually choose the note value (quarter, eighth, sixteenth, etc.) associated with the tap.
Mix: the amount of dry signal versus the amount of delay.
Some delays provide a choice of algorithms. A "normal" or traditional delay feeds the output back to the input for increased repeats. When adding delay to a lead vocal, watch out for excessive repeats, because they clutter your mix – much like long reverb times.
I typically use four delay times for lead vocals, roughly based on song tempos: 165, 205, 265 and 305 mS. The shorter times are for faster-tempo songs and have no feedback; 265 and 305 have about 4 percent feedback providing two repeats, which is plenty. The 265 mS delay is my go-to delay for Eric Bloom's vocal, and the 305 is dedicated to Buck Dharma's lead vocal. If the delay provides the ability, I'll offset one channel by about 10 mS to add a bit of left-to-right imaging. Older hardware delays may not have any filtering, so when you return the output to a channel on the desk, you can use the channel EQ to shape it.
A distinctly different style of delay is the "Multi-Tap," where you are given several delay outputs, each of which may be set to a different time. (Don't confuse this with Tap Tempo.) Depending upon the plugin or hardware you may have access to pan, level, feedback and filters for each delay, enabling you to create complex delay effects (and if you're creative about it, you can create an early reflection effect). One of my favorites is Yamaha's Multi Tap algorithm with 60 mS panned hard left and 40 mS panned hard right, both set for maximum output and no feedback. Add this to a vocal or guitar, and it becomes huge.
"Ping-Pong" delay refers to a delay that bounces back and forth between the left and right channels, typically with independent adjustable delay times on each side. This is probably useful more as a special effect than something you'd run on a lead vocal all night. Ditto for high feedback/high number of repeats and very long delays times. For example, on the song, "Godzilla," we use a delay of 665 mS on one side, 400 mS on the other, with 68 percent feedback and (if the program allows) feedback of left channel to right and vice-versa. You have to be careful with very high feedback percentages, because some delays will "take off" on their own and feed back without an input signal – very cool if you are doing a sci-fi film, but not so much in a music mix.
At least one manufacturer – t.c. electronic – offers a variation of Tap Tempo where you tap a rhythm and the device calculates delays based on your tapping. The first time I heard this, it was mind blowing. You input a rhythmic pattern with up to ten taps, and the D•Two from t.c. electronic plays it back. Awesome.
An effective technique that can be applied to any style of delay is ducking. The idea with ducking is that the level of the delay is automatically brought down while the original signal is playing. For example, let's suppose you have a long (300 to 400 mS) delay with several repeats on a lead guitar. It'd be nice to hear a rolling delay as each guitar phrase ended, but adding that much delay to the entire solo is likely to interfere with the original guitar. You could manually ride the level of the delay up and down as each guitar phrase ends, but that's a pain. (Plus, you already have enough to do). Ducking does this for you.
Some delays actually include this function, but it's easy to create if you delay doesn't offer it. Simply patch (or plug in) a compressor after the delay, and use the guitar (or vocal, etc.) to trigger compression via the compressor's key input. (You can route a bus or an aux send from the guitar to the key input). Set the ratio around 4:1 and the threshold so that every time the guitar starts, the delay is compressed by 4 to 6 dB. Set the release long enough so that, as the guitar phrase ends, the compressor "lets go" of the delay and allows the delay level to come up. When the guitar starts playing a phrase the delay drops, and as each guitar phrase ends, it comes back up, creating the illusion of a huge delay without getting in the way of the "dry" guitar. If you are doing this via plugins, make sure that the compressor is inserted in a slot below that of the delay. (Signal flow follows insert slots from top to bottom). Ducking is equally effective on lead vocal.
Next month we'll take an in-depth look at modulation effects.
