Assignment: Algorithmic and Convolution reverb

Hi fellow students!

My name is Hans Hendriksen, im from Chicago and this is my 5th assignment for the

course Introduction to music production.

This week my topic will be convolution and algorithmic reverbs. I will explain how they work and function, as well as show the differences between the two.

     

   recording IR's for convolution reverb..

 

         

...and programming an algorithmic one.

Reverb is a very commonly used effect. It's not just useful to give audio a sense of "space"
but it can also "color" the sound, give it character.
There are several ways people create reverb, and i will focus on the two above mentioned.

What is reverb?

To understand how these reverb methods work, you first have to understand what reverb is.

Reverb comes from the word reverbaration.

Reverbaration is a collection of sounds reflecting from a surface or enclosed space. Very simply: your living room, as well as your bathroom, car and oven all have specific reverbarating sounds.

If the surface it bounces off of is smooth, it will have a different effect than for example a brick wall.

You are probably sitting in front of a monitor or laptop reading my blog now. Imagine that if you say something, the sound will move to the monitor and bounce straight back to you. 

This is called direct signal or direct sound. The soundwave reaches your laptop, bounces back and enters your ear. 

Of course sound goes in all directions. So when it hits the wall behind your laptop, it bounces back as well but it will probably not reach you again in a straight line, but indirectly, bouncing off several surfaces.

These indirect sounds are called early reflections. The early reflections will take longer to get to your ear than the direct signal, so the eventual result will be slightly delayed. This mix between reflections and direct sound will give the listener the sense of reverb. It gives you an idea how big an area is, whether it has stone tiles or wood paneling on the walls.

Very often you can tell the type of space you are "in" by listening to the effect.

Convolution Reverbs

In a convolution reverb the reverb engine ( usually in the form of a plugin ) places samples "over" 

the existing audio signal. These "samples" are called Impulse Response. ( IR)

If you dissect that word, Impulse would be the sample, and response what a sound does ( how long it takes) when it is reflected on a surface.

So an impulse response is how a given sound "reverbs" in a room, space etc.

Convolution reverb uses that method. Sounds are being recorded in a space, and sampled and used in a reverb unit or plugin. The plugin will set the parameters of the IR . In a way the IR file could be compared to the Meta information of audio files. It contains certain information and instructions what to do with that information once it reaches a plugin.

If you use that impulse response ( sample) over your own audio signal, you place the characteristics of a given "space" over your own signal, thus creating a specific reverb effect.

Often convolution reverbs work with a lot of presets like "hall", "room", "church", etc.

You can, usually for creative purposes, edit these presets. 

Here, in my opinion, is where the weak point of a convolution reverb becomes obvious. It's flexibility. 

You can usually edit common parameters in convolution reverbs, but unless you want to experiment with unconventional sounds it usually doesn't work. Why?

Because the Impulse Response ( Sample) loses its integrity once you change certain parameters. And with that the effect quickly becomes confusing, rumbly, even distorting. The best way to use these type of reverbs is to use the presets as they were intended and using that characteristic in your mix.

Another potential problem is the cpu power these plugins use. In a way they are "samplers" and it takes a lot of resources to use them in realtime as insert or even as an auxilary send into your tracks. I use them very sparingly and usually as a send over the mix. 

The advantage is the quality of the reverb. Software companies go through great lengths to carefully record 

these IR's and the results, when used right, are often stunning. 

Algorithmic Reverbs

The Algorythmic reverb is widely used in most DAW's, and usually are provided as a standard effect included in the software. They almost function as a synth in that they modulate a signal with mathematical algorythms that can be manipulated.

An algorythmic reverb uses several delay feedback signals to create a reverb effect. It uses a method where it simulates early and late reflections using algorythms. This gives an abundance of parameters to play with and nearly any representation of reverbaration can be made using an algorithmic reverb.

Usually the early reflections are represented as a group of parameters and late reflections or simply "reverb" in another.
Common parameters are: * Input level
                                       * Room Size
                                       * Predelay/delay
                                       *Dry/wet mix
                                       *Depth

Examples of different digital reverb interfaces with common parameters

The big advantage with algorythmic reverbs is their flexibility. Since it simulates relections of sounds that are practically always unique, it has to be able to create basically any type of reverberation.
Compared to convolution reverb these plugins are usually very cpu friendly.
Disadvantage is that the reverb quality isn't as realistic and good as convolution reverbs.

I hope this lesson was useful and clear to you, thank you for reading it.
Good luck with your assignments and for later on good luck with your exams!

Hans Hendriksen