In a previous post we discussed some of the intangible aspects of mixing like arrangements and finding the point of interest, but this post is about a more subtle, yet vastly underrated aspect. Most engineers just developing the art of mixing usually concentrate on instrument and frequency balancing (EQing), but perhaps the most important item after that is the use of effects.
Effects can make or break your mix since their addition can make it sound larger than life or washed out and distant, depending upon their use. And the best effects are sometimes the ones that you don’t even know are there, but you’d notice immediately if they were muted, which is why this post is titled “The Subtle Side Of Mixing.” So let’s take a look inside the world of effects and how they’re applied to your mix.
When And Why To Use Effects
Most of us usually add effects almost by feel, never thinking too hard about why we’re doing it, at least in the beginning. For most of us, this leads to a lot of experimenting until things sound right, which can be a big time suck. That’s why it really helps to have an idea about the reason why effects are added and have a strategy for approaching the process as well.
There are four reasons why a mixer would add effects to a track:
- To Create An Aural Space. One of the reasons why we record elements in stereo is to capture the natural ambiance of an instrument, or the “aural space”. Since we can’t record everything this way due to track or storage limitations, a limited recording space, or because we’re close-miking, we usually just create this aural space artificially.
- To Add Excitement. Sometimes a delay or modulation effect added to an instrument or vocal will be just the thing to make it the hook or a major focal point of the song.
- To Make A Track Sound Bigger, Wider And/Or Deeper. This is the usual reason that a successful mixer will add effects to a track, but one of the hardest for a neophyte mixer to grasp because it can be so subtle. You can’t really tell the effects are there until they go away, but they never do, so how do you know?
- To Move A Track Back In The Mix (give the impression it’s farther away). One of the easiest ways to keep instruments from fighting each other for attention is to layer them by moving them back in the mix.
So now that we know the reasons why we we add effects, let’s look at a strategy on how to do it. Here are a set of rules that can help you choose the best effects for each track more efficiently.
Rule 1 - As A General Rule Of Thumb, Try To Picture the Performer in An Acoustic Space and Then Realistically Recreate That Space Around Them.
This method usually saves some time over simply experimenting with different effects presets until something excites you (although that method can work too). Also, the created acoustic space needn’t be a natural one. In fact, as long as it fits the music, the more creative the better.
Rule 2 - Smaller Reverbs or Short Delays Make Things Sound Bigger.
Reverbs with decays under a second (and usually much shorter than that) and delays under 100 milliseconds (again usually a lot shorter than that) tend to make the track sound bigger rather than push it back in the mix, especially if the reverb or delay is stereo.
Rule 3 - Long Delays, Reverb Predelays, or Reverb Decay Push a Sound Farther Away If the Level of the Effect Is Loud Enough.
As stated before, delays and predelays (see below) longer than 100 ms (although 250 is where it really kicks in) are distinctly heard and begin to push the sound away from the listener. The trick between something sounding big or just distant is the level of the effect. When the decay or delay is short and the level loud, the track sounds big. When the decay or delay is long and loud, the track just sounds far away.
Rule 4 - If Delays Are Timed to the Tempo of the Track, They Add Depth Without Being Noticeable.
Most engineers set the delay time to the tempo of the track (see below on how to do this). This makes the delay pulse with the music and adds a reverb type of environment to the sound. It also makes the delay seem to disappear as a discrete repeat but still adds a smoothing quality to the element.
Delays are measured tempo-wise using musical notes in relation to the tempo of the track. In other words, if the song has a tempo of 120 beats per minute (bpm), then the length of time it takes a quarter note to play would be 1/2 second (60 seconds ÷ 120 bpm = .5 seconds). Therefore a quarter note delay should be .5 seconds or 500 milliseconds (.5 X 1000 ms per second) which is how almost all delay devices are calibrated.
But 500 ms might be too long and just sound confusing in the mix. Divide that in half for an 1/8th note delay (500 ms ÷ 2 = 250 ms). Divide in half again for a 1/16th note delay (250 ms ÷ 2 = 125 ms). Divide again for a 1/32nd note delay (125 ÷ 2 = 62.5 ms or rounded down to 62 to keep it even). That still might not be short enough for you so divide again for 1/64th note (62 ÷ 2 = 31). Again this might not be short enough, so divide again for a 1/128th note (31 ms ÷ 2 = 15.625 rounded up to 16 ms). And yet this still might not be short enough so divide again for a 1/256th note if there is such a thing (16 ms ÷ 2 = 8 ms).
Now such small increments like 8 and 16 ms might not seem like much, but they’re used all the time to make a sound bigger and wider. It’s something that you might not exactly hear, but you can perceive it since it acts as the critical “first reflection”, which is the loudest and most important echo of a sound in any environment. Even a short delay like this will fit much more smoothly into the track if it’s timed.
Another way to determine the delay time is to use the following formula:
60,000 (the number of milliseconds in a minute) ÷ Song Tempo in bpm =
Quarter Note Delay In Milliseconds.
Example 60,000 ÷ 128bpm = 468.75 milliseconds (rounded down to 468 to keep it an even number).
All the other values can be determined from this by either:
- dividing by 2 for lower denominations (i.e. 468 ÷ 2 = 234 ms for 1/8th note delay, 234ms ÷ 2 = 117 1/16th note delay, 58.5 1/32nd note delay, 29ms 1/64th note delay)
- multiplying any of the above by 1.5 for dotted values (i.e. 234 ms x 1.5 = 351ms for dotted 1/8th note)
- multiplying any of the above by .667 for triplet values. (i.e. 234ms x .667 = 156ms 1/8th note triplet)
Dotted and triplet values are very effective delay settings and many times take precedence over straight note delays since they have an interesting feel, providing movement to the part in a subtle way. Plus they fall in between the “can be heard” and “can’t be heard” crack. In other words, they’re noticeable without sticking out like an untimed delay.
It’s also an interesting effect to sometimes use a stereo delay with a straight delay of a ¼, 1/8th, or 1/16th note on one side and a dotted note or triplet on the other. If the delays are under 100 ms or so, it simulates the sound of a room. During the early 80’s and 90’s, a delay of around 25 ms on one side and around 50 ms on the other was used to enhance the sound of a clean electric guitar, for instance.
Rule 5 - If Delays Are Not Timed to the Tempo of the Track, They Stick Out.
Sometimes you want to distinctly hear a delay and the best way to do that is to make sure that the delay is NOT exactly timed to the track. Start by first putting the delay in time with the track, then slowly alter the timing until the desired effect is achieved.
Rule 6 - Reverbs Sound Smoother When Timed To The Tempo Of The Track.
Reverbs are timed to the track by triggering them off of a snare hit and adjusting the decay parameter so that the decay just dies by the next snare hit. The idea is to make the decay “breathe” with the track. The best way to achieve this is to make everything as big as possible at the shortest setting first, then gradually get longer until it’s in time with the track.
The predelay of a reverb (the space between where the note of the source track dies off and the reverb begins) can change the sound of the reverb considerably and is usually timed to the tempo of the track. Back in the days of real plates and chambers, predelay was achieved by using a tape slap. This was the natural echo that occurred when playing back off the repro head of a tape machine while recording onto it. Since there was a gap between the record and playback head, it gave a noticeable delay and engineers of the time used this to their advantage.
Because the early tape machines didn’t have a way to vary their speed, it wasn’t possible to time the delay to the tempo of the track. The best that could be done was to select either a 7 1/2 inches per second (which came out to about 250 ms) or 15 ips tape slap (about 125 ms). Many legendary engineers that grew up in that period still use a predelay of around 125 ms or so (timed to the track, of course) on just about everything.
Predelay can be very effective on vocals, providing a separation between the vocal source and the onset of the reverb effect that maintains the focus of the vocal while not washing it out. The same reverb without the predelay can make the vocal sound undefined or even a little distant. Predelay allows you to add more reverb without making it sound like the vocal is “swimming.”
While following the above rules will help you to determine when and where to add an effect, there’s still more that you can do to help everything blend together, especially in an era of almost unlimited available effects. One of those things is something that’s overlooked by a lot of mixers, which is to EQ the send or the return of the effect itself.
From the early days of reverb chambers and plates, it’s always been common to EQ the reverb returns, although the reasons for doing this have changed over the years. Back when plates and chambers were all that was available, usually some high frequency EQ at 10 or 15 kHz was added because the plates and chambers tended to be dark sounding and the reverb would get lost in the mix without the extra high frequency energy.
Nowadays EQ is added to reverb in order to help create some sonic layering. It’s pretty common for an engineer to have a problem with the sound of a delay or reverb during mixing. Usually the engineer will spend a lot of time trying different presets in an attempt to help everything blend better, but most of the presets are just different internal EQ settings of the same preset anyway, and just a little EQ added or subtracted externally to the effect could’ve made everything fit nicely in the first place.
Here are a few tips for EQing effects that will help everything blend without spending a lot of time experimenting. The type of reverb (digital, real plate, etc. - delays too) doesn’t matter as much as how these are applied, and that depends on your ears and the song.
|Equalization Tips For Reverbs And Delays|
Here are a couple more interesting EQ tricks that work really well:
1) When adding reverb to drums, don’t be afraid to roll off all the low end and even most of the mid-range (like up to 4 or 5k, even 8k). You’ll be surprised how much more natural the reverb on the kit sounds and how much of the mud on the low end will be eliminated.
2) Another good trick is to EQ the send before it goes to the reverb. Most digital reverbs, either plug-in or hardware, have EQ controls on them, but they’re usually set up post-reverb. If it you treat the signal before it gets to the reverb, you get a whole different kind of sound. In fact, this is the way it was done at Abbey Road for many years and you can hear that sound on the reverbs of most Beatles records (for those of you who purchased the any of the reissued albums). The trick to that sound was a filter that rolled everything off below 600Hz and above 10kHz, which are still pretty good starting points for any effects alterations that you might want to try.
Just for the record, one of the chambers at Abbey Road also had a 6 dB boost at 3500 Hz with a wide slope, and that’s something you might want to try as well.
Layering of Effects
Sonic layering means that each instrument or element sits in its own ambient environment, and each environment is usually artificially created by effects. The idea here is that these sonic atmospheres don’t clash with one another, just like in the case of frequency ranges.
The following chart features some suggestions to try so that the sonic environments don’t clash.
|Layering Tips For Reverbs And Delays|
Compression on the Mix Buss
The reason why we introduce compression into the discussion of effects is that most great mixers don’t use the buss compressor so much for level control as much as for the sound that it imparts, since they only add a decibel or two to the final mix. In this case, the type of compressor that you use across the buss is important because it’s the sound of the unit that you’re after, and many units actually add an intangible sonic quality. Current favorites are the Fairchild 670 (at a hefty $25,000+ for the vintage hardware model, and they’re mono – see Figure 1 for a look at the Universal Audio version), the Neve 33609 (Figure 2) or the SSL (Figure 3).
Of course, the other reason why most mixers insert a compressor across the mix buss is to get a sound that’s more finished and mastered. Originally this came about when artists began asking why their mixes sounded so different in the studio from what they heard on the radio or when their record (it was still vinyl in those days) came back from the pressing plant. Indeed both the radio and record did sound different because an additional round or two of compression was added in both mastering and broadcast.
In order to simulate what this would sound like, mixing engineers began to add a touch of compression across the mix buss. The problem was, everybody liked it, so now the majority of records now have at least a few decibels of compression added to the stereo mix despite the fact that it will probably be recompressed again at mastering and yet again if ever played on the radio or television.
The SSL Mix Buss Compressor
The sound of a great many (some say the majority) of records in the 80’s and 90’s comes from the sound of the built-in mix buss compressor on an SSL console (Figure 3). This is an aggressive compressor with a very distinct sonic signature. Some have even gone so far as to call the compressor IN button (meaning it’s currently inserted in the signal path) the “Good” button because it makes everything sound better.
If you happen to get a chance to work on an SSL (any vintage, they all have a mix buss compressor) or any of the software plug-ins, here’s the time-honored setting to use as a starting place.
Typical SSL Buss Compressor Settings
- ATTACK – all the way slow
- RELEASE – all the way fast
- RATIO – 4:1
- THRESHOLD – to taste
A Little About Modulation Effects
You’ve probably noticed by now that we’ve not covered any modulation effects and the reason why is that, after years of overuse, effects like chorusing have really fallen out of favor in terms of mixing trends. There was a time there during the 80’s when you’d listen to a mix and it seemed like every instrument had a little on it and many mixes just became a jumbled mess as a result. I’m not saying that you shouldn’t use a modulated effect because there may be just the perfect place for it, but in general, your bread and butter effects will always be reverbs and delays.
One more thing about chorusing: Many keyboards still come with a stereo simulator that’s nothing more than a cheap chorused output that does more harm than good since it doesn’t collapse to mono well and can mush up you mix. In a case like that, you’re a lot better off if you can find a mono-only patch and use that, adding your own effects in the mix to simulate stereo if you must.
So there you have it. If you think about these concepts regarding effects during your next mix, I promise that you’ll find what you need faster, and you’ll be able to layer and blend them better that ever before.