Part 2: The Delay Effect - The Rhythmical Mix

Previous part of the delay-theme concerned how delay can be used as a delicious atmospheric effect. Today it's going to be about how the use of delay can make miracles for the complexity level of the instrumental capacity, and how it can make your music more interesting to listen to.
Imagine that the frame of time before the signal will be reproduced on a given sound source is connected to the tempo, the original sound is aligned in. Then imagine determing at what rate or rhythm the signal will be repeated in.

There are also others who have imagined this, one can safely say!

The Audio Technicians Troubles:
Once, echo-deployment technique's was one hell of a nightmare. We have not always had these manipulating machines, but fortunately we got those a few decades ago, but once we had to record the requested echo in a natural, echoing, reverbing environment. It was hard enough, and therefore, the demand for machines that could record the signal and echo it a certain range of times with certain intervals between each play was high, in order to emulate the effect of an echo in natural surroundings, even with or without the resonance of the room.

The mega-popular "Roland Space Echo" was introduced as one of the first and very best machines for fulfilling this demand. Back then, it came in a very compact package, for the time being.

But it was only with John Martyn (Just listen to Glistening Glyndebourne) that we began to experiment with echoing as a rhythmic effect. Martyn developed his own playing technique, based on dotted 1 / 8's rhythmic delay. This delay was then taken to mainstream heights through U2's The Edge, through whom Martyn's technique has become Edge's trademark. Later, many others have been using the effect in the same manners, including some of my own personal biggest idols, and in my own music, this effect is used strongly.

From there it went in different directions. The delay formed the basis for other new things such as reverb, chorus, flanger and phaser, and sound engineers tested a lot of different techniques with phase accuracy and doubling. While it also gave us as musicians the opportunity to develop new playing techniques, such as we saw Martyn and The Edge do it, it became founder of tons of wavemanipulating techniques.

But it is not only to the dotted 1 / 8's that the delay has had its foothold as a rhythmic effect. The popular reel tape recorders that could record a signal, and reproduce it hundreds of times, made it possible for everyone to loop their own performance and build on it. This was especially smart with the aforementioned Roland Space Echo, and it was a hit for solo guitarists to play long solos where they reconciled with their own performance, and moved into the wildest of soundscapes.

But let's move on ...:
I will now show you how this works in practice. I'll give you examples of which delays work for what, and how they should be adjusted.
For absolute guideness, I have made a few audio samples of a signal in the dry and delayed mode.

Lets first hear the dry signal without effects attached:

Dry signal

Let's play back the same sound. but with a 1 / 4 rhythmic delay at (400 ms at 150 bpm):

1/4 signal

And finally a signal with a dotted 1 / 8 's rhythm (300 ms at 150 bpm):

1/8 Dot signal

Let me throw a comment to the last specified audio clip, which is set accordingly as Martyn first introduced it in the old days.

When you want to give your instrument or basically any signal an echo, and you want it to be rhythmically based, you must first find out how many milliseconds time between sound and reproducement the given rhythm needs, in connection to the songs tempo. Either you calculate it yourself, or find something to do it for you. Last mentioned is the only solution that works in practice and are perfectly correct every time.

Here I can recommend you everybody's favorite -Delay Time Calculator 1.0.
And why? Well, first and foremost it is freeware, it requires no installation, take up less than half a megabyte on your harddrive, and most importantly, it has a simple, accessible interface. You type in the tempo, the desired rhythm, whether it should be punctured/triplets (or not), and lastly, you press Calculate. Then you'll have your milliseconds in front of you.

For everyone's sake, I've taken a picture of the program and directed a small example. Here's how we would calculate if we had a song in 150 bpm, and wanted dotted 1 / 8 'e.

What we choose to delay our audio is one big jungle. Should the delay be retro and analogue? Should it be digital and modern? Should it be spacy and run multiple times? Must it be even more spacy, and also modulate the signal with filtering and so on? All this is up to you and your preferences.

Therefore, I will write up some examples of ways you can use the delay and what you can produce with it.

Firstly there is the essential choice:

Must the delay be placed before or after the recorded signal?

Well, you see, have you ever sat with Guitar Rig, Amplitube, ReValver, Pod Farm, or other of similair ampsimulators, and tried to incorporate a driven guitar, but recorded your tracks without the drive - instead with a completely dry and clean signal? Your point has surely been to then put on the driven sound afterwards. If yes, then you have learned, that the lack of "feel" is very audible. Because you recorded your drive guitar completely clean and raw, you've needed the feel of playing with the correct sound, and therefore your performance lacks the sense of rightness.

The same applies to delay. Try to imagine a performer playing a rhythmic echoed guitarpattern, without being able to hear the rhythm. When delay are added afterwards, you will be surprised by how uneven and untight the performance ended up.

So, be sure to delay your signal before recording it. But - place the delay last in the chain. After the amp, not before. If you place it before the amp, the delay will be echo the raw sound, not the driven one. There are many ways to use the delay. Here, I present a couple of them.

The Analog Tape Echo:
This has had its strong point as a looping machine, bringing crazy delaytimemanipulating effects to life through the ages, and is particularly good because the signal is sent back through the tape sounding very warm and flat without any annoying top frequencies, and therefore, this works well as an anonymous feature, running "under" the drysignal.

The drawback to this is the rhythmical feel invalidated when dry and wet signal are very far apart. Since we are not all running around with reel tape recorders today, Boss had a smart solution: Boss RE-20 Space Echo. Modulated upon the Roland Space Echo, this one is a digital floorboard-sized pedal that can be used to record very long passages, and loop them, or to produce (and by that the name) spacy echoscapes.

The Modern Delaypedal
In many ways, one can safely say that "Digital" is a term, the fewest people are looking for when looking for outboard gear, pedals, instruments and so on. But sometimes the digital world has advantages. Let me introduce one of the world's most popular pedals, Line 6 DL4. Why choose a digital delay?

Well you see - with digital technology you get your signal recorded in 24-bit, real stereo, which allows you to manipulate your signal in the most grandiose ways. You can get a guitarist to sound like he has overdubbed his takes, and you can create a much more wide stereospectrum on a single soundsource. I will describe how to use the DL4 or other stereodelay's for just that in the last section of this series.

With a stereodelay, your signal will be reproduced 100% as the sound came, apart from the built-in filters you can modulate in different directions. This could be bass and treble, for an example.

A disadvantage is that you can not dynamically manipulate your signal in the wildest ways, like some alternative units can. Where the Boss has 2 expression pedals and gives tons of modulation and looping choices, you must settle with "only" the high-fidelity rhythmic style reproduced signal, produced by most of the modern delaypedals.

Equipment Recommendation:
Last time I recommended one of the most simple delay plug-ins. Today, I will recommend one of the most heavy ones.
The animal is called PSP 608 MultiDelay. They brand themselves upon the statement that they have the largest amount of options on a delaypedal on the market, and I must admit that I believe them. The 608 is a delightful mix of tape echo and the modern echo, I have been praising so much. You can do crazy varieties from your drysignal, as it supports the entire 8 different delay "channels" on your track. It allows you to manipulate the sound in almost infinite ways. The sound is hot, and it has lovely built-in filters, and lots of opportunities for pans, Stereo Widening, reverb and much more. In addition, it has a fantastic bank of good presets that can really pull you in many directions. But I think it is a drawback that it is very hard to adjust properly, especially for rhythmic patterns. None of the included presets I have tampered with rhythmically functions well. This shall be used more as an atmospheric effect.

STRONG TIP:
This is one of my most frequently used techniques, and it involves the 608.

If you are going to copy a track for more fill and widening, and want to avoid phase problems, it will never be the same as a double-take to copy and move the track 15-20 ms. But actually, I ball out of the same street. Move the copied track 15-20 ms, open the PSP 608 MultiDelay, and open the preset "Voice Doubles 1". It really creates a stereo image, and you will suddenly hear a more natural reproduction than you do when you simply move the track a few ms.

This tip is especially recommended for filling out the vocal or solo guitar. It should also be mentioned that the Voice Doubler 1 acts as a strong Stereo Widener, and therefore, it just creates more fullness in the sense that the drysignal lies somewhere solidly in the middle, while the wet track "space out" and create a wider, fuller sound, underneath dry signal. Great feature!

Follow the next and final post, when it is about studio tips (which I already gave a little attention already, whoops).

Part 1: The Delay-Effect - The Empty Mix

One of the most basic and frequently used effects in sound art in recent decades is without a doubt the delay effect. Even reverbalgorithms consists of lots of small delays, which all together adds up to a whole. Delay is to be understood as "Echoes". What delay does in practice is to reproduce a sound with a given space of time between the original sound and the reproduction. It is then optional to estimate a given number of times the echo will repeat, and a particular rhythm or "time", the echo will be running at. It can either be used as a rhythmic effect, an artistic effect, or a "fill-out" effect to widen the perspective of the mix. It has the power to manipulate sound waves and the ear's perception of sound. This philosophy is also to be seen in chorus, flanger and phaserpedals do. Here too, it is modulated echo algorithms that create the wild sounds we can get out of a sound source.

In this series of posts, I will describe all my ways to use the delay as inspiration to both me and you.

I will divide it into 3 sections - divided into 3 posts. A rhythmic section, a "fill-out" part, and finally a section on study tips, where delay can be used as a power that can create miracles in your mix. In this post, I will be talking about the delay as "fill-out" effect, which purpose is as previously stated to widen the perspective of the mix.

The empty mix:
Do you feel that you have a solid basesound on vocals, but that it feels like it's standing outside the mix? Or maybe you are rocking out a lead-guitar solo, but it just doesn't stand enough apart, no matter how you look at it? I've certainly thought this a few times or 10 in my life.
In this case, it's a very good idea to consider using delay. In very grandiose music - rock as pop, orchestral ballads, dance, metal and anything else punchy, delay is very often used to fill the vacuum left out.
Take a listen to the vocal track on one of your most powerful songs. Is there not a delay, reproducing the vocals, somewhere deep in the mix? Or what about your favorite guitar solo on any Iron Maiden album? It's to be found anywhere in my record collection at least.

And so to the heart of it. If you, like me, have been using delay on singers' vocal tracks, guitar solos or the like, but they seemed fill too much in the mix, you probably turned them down to where they are so far back in the mix one barely hear them. They appears to be filling out too much of the mix, because the reproduced audio signal is just as dry as the original one, just in a different volume. This is not the way to do it!

In all qualified Workstation software, you can work on aux buses, and now you need to involve just that.
This little guide will help you out with opening an aux bus up, to then followingly guide you through creating a clear, yet distant delay as you hear it in your favorite music.

In the attached pictures you can see a series of framed "boxes". I have framed them to highlight what is important to focus on. Since I did not have internet in the studio at the time I wrote this article, I produced all material for these guides elsewhere, and because of the lack of audio software at this location, I just installed a Cubase 4 LE, and found a couple of free plugins on the internet. It is not plugins I use ever at my serious work, and this just functions as examples of how a chain could be.

Open the project where you would like to fill the empty void with a delicious delay, which WON'T be dominating or annoying the mix.
As I mentioned earlier, there is labeled an "aux bus." While your whole mix is running through a master bus, which you can turn up or down on the master fader on your transport bar, there are also a lot of internal busses in your project; or at least you can create them. In this case, we must put up an effectbus.

I run with Cubase, and has also made this guide so that it fits Cubase. But I'll try to explain this as objectively as possible. If you are a regular Cubase user, find where to add new tracks to the project and select "FX Channel". This could arguably be called "Create Aux Bus", "Effects Bus", "Aux Channel" or something like that in all other DAW's, but here it is entitled FX Channel. From here, select your preferred delayplugin. I will be listing plugin recommendations in the end, in order to give you an idea of what to choose. As you can observe in the picture, I chose the plugin "KXDELAY2--p", which is a freeware delay plugin. There'll now be created an effect bus for the plugin KXDelay 2.

The smart thing is not just that you can assign a plugin several plugins on top of it, but also that it will save you a great amount of computer resources. Instead of having to put plugins on a track "insert" each time, and thus constantly increase the pressure on the CPU and RAM, you can now send your tracks through a plug-in bus, and by that have plenty of extra power to play with. There are also other benefits you'll learn about in practice.

Once you've done it, find the audio tracks you would like you to assign to the delay. As you can see on picture 1, you'll find "Send" between the various options you have on the audio track. It is through the "Send" you send the track's signal through a second channel. Choose to send the signal through the newly created channel "KXDELAY2--p", or what it will now be named among you. How much of the delayed signal that will go through is up to you. Here it is about finding the perfect balance. It should preferably not fill up too much. I usually run it up to about -7 dB as a maximum.

Now you are halfway. You have sent your audio signal through an effects bus, which you've also just created. In reality, you have already assigned delay to your signal. Try for yourself to play back your tracks! But that would also be just the same as just adding delay as a insert on the channel - and then all this hassle was for no matter!



Therefore, as the second picture suggests, find the FX Channel where you've put up the delay. Form your delay, so it fits you well. When you are satisfied with the delay, then it's time to be creative!
You have shaped your delay, now is the time to shape the sound of it. Here is what I usually do.
In order to make the delay not dominate the mix too much, I usually cut most of the bottom- and topfrequencies. Maybe not all of them, but that would not be a completely stupid idea. The freeware EQ I found had neither Low- or Hi-Shelf, Low- or Hi-cut, but only just traditional peak bands. I pulled both the bottom and top completely to a minimum. It must be so that you only get the mid-frequencies of your signal. That will make it audible, (mids make things stand strong), but distant (removing top also means making things feel more "distant"), and without mud (bottom frequencies tend to mud up the mix).
Then, to create the "space" and fill in your delay, all you need is a simple (and yet complex) thing - reverb. Add a reverb on to your delaychannel and adjust the wet / dry until there is a fine balance - I tend to think 50% is a good option.

You can play with this for an eternity, that is really the point of learning and finding your own way, but this is enough for your delay to stand clear in the mix, without dominating anything. It feels distant, and is serving as "fill-out" on your mix. Try this example on your lead guitar, or your lead or backing vocals!

----------------------------------------------------------------------------------
In each of these posts, I will include a plugin recommendation.
Here is my favorite Stereo Delay that you can use for just that.
Normally I am not a fan of Steinberg plugins. I always use plugins from other sources than Steinberg, yet I see myself never going away from a single plug from Steinberg. It is recommended for all its good transparent sound and absolute simplicity.

Steinberg Double Delay:
A Stereodelay plugin that was introduced with Cubase SX3. The simplest of the simple but yet with a completely transparent sound - and you can run hundreds of them without playing out the CPU. Normally one might wish for colored sounding delay, but the goal here is to color the sound with our own effects for full flexibility and individuality, so transparence is very important. I use this plugin as the first plugin in the chain of almost all delaychannels I have ever run.

Next time there will be recommendations to the slightly larger, serious and (therefore) expensive delayplugins.

Dynamic Processing and the bass

It has recently been troubling my mind that the bass was always muddy on my mixes. In this case, I do not think of the bass as the total amount of bass of all the instruments, but simply the bassfill of the single instrument - the bass guitar.

It was quite enigmatic to ensure a clean, smooth volume that stood clear and bright in the mix, without problems with muddyness at the bottom. Before I was able to let the bass guitar play back on a regular volume, I compressed it with a deep threshold (-30 dB, was not quite rarely seen), and with a ratio between 3 and 4. In addition, it should also be the instrument that put the largest amount of bass in the mix. As a matter of fact, it should be the only instrument playing around in the deepest frequencies, together with the kickdrum.

I quickly realized that it was easier said than done. As described earlier, it was one hell of unmastered transients, and you could clearly feel the bass hanging over all elements in the music - rather than under.

I switched between a wide range of compressors in my quest to find one that did find the sweet spot between keeping tight tone, but also dynamics and punch, but no clicks and crackle. I had fine results with API or SSL compressors. Waves Rcomp or C1 Comp also did a good job, but as soon as I started to fill in the bass and treble, the problems were back again.

Suddenly it dawned on me. If I had fine results with the compressors, but the problems first appeared when I added or decreased decibels to some frequencies - why did I not compress some bands more than others?

I thought quickly upon the idea of a multi-band compressor. If you do not know what this one does, here goes first a little explanation of what a normal compressor does. A compressor limits the dynamics, by squeezing the highest and lowest peaks together, so that the sound level becomes more even. This can be done by degrees, and you can use it as a little addition, or you can make it so tight, that even the weakest impact on the sound level is on volume with the hardest.

A multi-band compressor, however, is pretty much the same. The only difference is that more compressors than one are active, spread out over different spectrums of frequencies. A typical multi-band compressor of the higher class usually has 4 bands, all of which may be set individually and completely different from each other.

Like I wrote in the last article about Limiting, many of the same tendencies exist on this market too - there are many bad and few good software-based plugins. Since I typically work in software, and has no special outboard to speak of, I have been looking for a great multi-band compressor.

And finally I found something that absolutely cured all my problems!
This time it's a plug from the Waves series. C4 is it's name. It compresses like Waves own Rennaisance Compressor, which I also have had fine results with on Bass, Toms and other instruments in the low frequency spectrum. RComp is not a punchy compressor, and goes to great lengths with smoothing out the dynamics, without turning the instrument "Clicky" (like you would see from a lot of cheap compressors - it's not a bad thing at all, clicky compressors are used in large scale on snare drums, kick drums, and everything else that need a transient boost). And the built-in EQ is the same technology as in their Equalizer series "Q", which simply put is my favorite Equalizer, along with Flux Epure.

I placed the C4 on the bass bus, and loaded a preset that was optimized for bass, which also functioned as a Deesser. It had a tight release on the topfrequencies.
The result was a surprisingly smooth bass with no mud at all - and then I could go back in my plugin chain and raise the high mid / high frequencies for more clarity, without the "click" - it did both function as a great Deesser and as RComp limiter-like compression.

Recommended for bass instruments, pianos and synths, or generally any instrument acting in the spectrum between 30 and 4000 hz.

Hi World - Introducing Limiting

With these words, I am introducing the first written words on my brand new blog, through which I hope I will help some desperate souls, or at least just myself. The goal here is to make myself smarter on the techniques I use myself in my daily work, and to get them down on paper, so I can go back and read up on my own tips.

Today's post is about my favorite limiter plugin; Voxengo Elephant. For those of you who do not know what a limiter is, here goes a simple explanation of its function.

A limiter limits the signal level of a given sound source to an attached value of decibels. Since digital media is "distorting" any peaks above 0dB, virtually all technicians limits, or in popular terms, "clips" the signal at 0dB.

What this means in practice is that with a limiter on your track, you will never see the track distort; it will never exceed 0dB.

This technique is mostly used to get the volume up on a track, or typically to raise the volume of a whole mix. This is very similar to a compressor, which compress the lowest and highest peaks, to make the audio levels even, in order for us to be able to raise the volume further without the digital distortion. In terms of limiting, the more you'll gain the volume, the harder the job gets for the limiter to keep peaks below 0dB. Particularly cheap or free plugins on the market are especially troubled with this, and therefore limiters are pricy, and well worth it. But also the very biggest and most expensive limiters from Waves, PSP, etc, have problems when the gain turn over 3dB. It may not distort the signal, but it will kill all transients and dynamics ; your track will not either sound or feel right, for the most types of music. It will sound like a digitized overcompressed mud box, nobody will praise.

Therefore, all the major software and hardware manufacturers of processing gear are battling and competing right now, and they have done so since the 90s, to make the limiter that can press the bit of extra decibels into a track or mix, that the other ones can't - without destroying the dynamics. This is a question of minimal loudness differences, which nonetheless makes a world of difference on a competitive level between volume in different publications.

For quite a while, Waves sat on the tip of the Loudness throne with their "L2 UltraMaxiMizer" . It is an fairly transparent, yet coloring limiter which comes in both software and hardware form. Later they updated it to an "L3" edition, though not all agree that it wins over the L2 in all scenarios. Let me say that my own experience with the L2/L3's are mixed. If you come around 4dB gains, it will color the sound in an undesirable manner. It gives exactly the digital "oomph", few of us are looking for.

That's why you must resort to other means, if you want to achieve the highest gain of decibels without sacrificing the dynamics and transients. IZotope came shortly after, with their "Ozone" mastering limiter / eq / multibandcompressor. It goes to 4dB or over fairly well, and feels much more transparent than Waves otherwise very strong L2/L3 limiter.

But is it enough? The struggle has continued, and recent additions is from Voxengo, who just released their Elephant limiter. Elephant is an exceptionally cheap limiter, and is by far the cheapest of the professional limiterplugins. It comes at a price tag below the budgets of a homestudio. It is very CPU-friendly, and in terms of sound? - crazy good.

Here is what makes Elephant a brilliant limiter. There are none that are as transparent as Elephant. That's what you need to know. For those who want further explanation - Elephant retains punch and frequency relationship up to about 6 dB gain, and with the right settings, it can keep a track's original punch and clarity without any nasty feeling of digitization. Elephant presses volume higher than any limiter on the market without artifacts.

And we know little yet! Besides being the ultimate best on the market, it too is ultimately the most flexible limiter I've ever seen. Voxengo has given us so many features available that we'll have to work for months with Elephant before we understand all its features. When we do, we can tweak it to fit completely in with our track, and the loudness level we seek. It supports up to 8x Oversampling, 100% customizable Knee and release settings, built-in dithering, DC offset filter, and much (much, much) more.

Even the harshest of the loudness critics, the ones who always have spoken out against the loudness war and any kind of hard limiting, have had to recognize that the Elephant really functions as a nice transparent way of increasing output without digital artifacts. And it costs literally nothing.


Can be used at both individual tracks and the mastering the bus, last in the chain obviously. The built-in dithering is highly recommended.