Hi there! My name is Francis G. Loch and I am from Glasgow, Scotland. This lesson is for week 5 of Introduction to Music Production at Coursera.org. In this lesson I will be comparing and contrasting algorithmic and convolution reverbs.
Please note that the images in this lesson can be enlarged by clicking on them.
What is Reverb?
Before looking at what algorithmic and convolution reverbs are, let’s take a moment to look at what is meant by reverb.
Reverb can be described as the sound or sonic signature of a space. In a room a sound will travel, reflect off walls and other surfaces, and will arrive at a listener’s ears as a collection of delayed echoes. The size of the room, the surface materials, and a number of other variables will all affect the reverb characteristics of that room. A person speaking in a small room will sound quite different than if they were speaking in a large concert hall for example.
Reverb is an important effect in music production. During a recording, a lot of effort might have gone into removing the sound of the room such that the recording loses its sense of space. By using reverb, the sense of space can be added back in and can also make the recording sound as if it were recorded in a larger space than it really was.
Algorithmic reverbs are created by using a formula. Below is Kjaerhus Audio’s Classic Reverb plugin which is an example of an algorithmic reverb.
The interface for this plugin is split into three sections – reverberation, filters and output. Here is an explanation of each of the parameters for each section.
Defines how big the room is. The bigger the room, the more reverb that is produced.
Simulates the reflectivity of the room’s surfaces. The higher it is set, the less the sounds will be reflected. This can also be called diffusion on other plugins.
Adjusts the delay time between the original (dry) sound and the start of the processed (wet) sound.
HI DAMP. (High Frequency Dampening)
Similar to the damping parameter, except this will only dampen the high frequencies.
LO CUT (Low Frequency Cut)
This is an EQ filter that cuts off the lower frequencies (adjustable between 20 and 1000 Hz).
EARLY REF. (Early Reflections)
Sets the gain for early reflections. Early reflections are the echoes created when a sound first reflects off multiple surfaces.
Controls how much of the original (dry) sound is mixed with the processed (wet) sound.
Controls the amount of gain at the output.
Convolution reverbs are created by using a recording of a real space. This sampled recording is called an impulse response, or IR for short. Below is Native Instruments’ Reflektor plugin which is an example of a convolution reverb.
The controls on the left third of the interface deal with the selection of presets and setting up the impulse response category and sample to be used for the reverb.
Some of the other parameters have similar functionality to what was found on the algorithmic reverb, so only the new ones will be explained:
Affects how long the end of the sound is heard. It can be decreased reducing the length of the sound, or it can be increased amplifying the end of the sound.
Adjusts the start position of the IR sample.
Mutes the send of the original (dry) sound to the reverb processor.
Reverses the IR sample.
R POS (Reverse Position)
Sets which part of the IR sample is reversed.
LO-ENV (Low Frequency Envelope)
Dampens the low frequencies. The ‘F’ knob below the slider affects the frequency at which the dampening occurs.
HI-ENV (High Frequency Envelope)
Dampens the high frequencies. The ‘F’ knob below the slider affects the frequency at which the dampening occurs.
Adjusts the cut or gain of the peak EQ filter. The ‘F’ knob adjusts the frequency and the ‘Q’ knob adjusts the width of the filter.
Synchronises the timing of the decay and predelay parameters to the tempo of the host DAW.
Adjusts the width of the stereo image.
Simulates the depth of the room, bringing it closer or further away.
Controls the panning of the wet signal.
Algorithmic vs Convolution
As already stated, algorithmic reverbs are created using formulas much like a synthesiser creates sounds using formulas. As convolution reverbs are created using recordings of a space, they are more like a sampler.
Because algorithmic reverbs are built using formulas they have far more potential for manipulation over convolution reverbs. Convolution reverbs on the other hand will often sound more realistic since they are a recording of a real space.
Here is a recording of a guitar being played without any reverb added:
Here is the same recording, but this time with the algorithmic reverb being toggled on and off throughout:
And this time with the convolution reverb being toggled on and off:
Prior to doing this lesson I had always just used the supplied presets in the reverb plugins I have. I had never really taken the time to learn and explore what all the parameters do.
I also found myself forced to look through the various reverb plugins that I have to see if I had one that done convolution reverbs, so it was a surprise for me to actually find out that I did have a couple already at my disposal which I will be making more use of in future!