This post was originally published on www.dbr-training.eu
In this series of post, I want to share my way of using Adobe Audition to produce audio assets of high quality for my eLearning courses made with Adobe Captivate. In this second post, I will focus on the second step of this process : the Equalization step.
What is equalization?
According to Wikipedia, Equalization is the process of adjusting the balance between frequency components within an electronic signal.
When applied to an audio signal, the equalization process lets us adjust the energy of a particular frequency or of a range of frequencies.
To achieve this adjustment, sound engineers use tools called equalizers, and, because things are never as easy as we want them to be, there are different kinds of equalizers available. The ones that we will use in this post are the high pass filter, the low pass filter and the parametric equalizer.
The High Pass Filter
A high pass filter lets the higher frequencies pass through the filter and, consequently cuts the lower frequencies. (Some people call this filter a Low Cut filter).
All we need to do is to set the cutoff frequency. All the frequencies that are below the cutoff frequency are reduced, while the frequencies that are above the cutoff frequency are left untouched.
The Low Pass Filter
A low pass filter lets the lower frequencies pass through the filter and, consequently cuts the higher frequencies. (Some people call this filter a High Cut filter).
All we need to do is to set the cutoff frequency. All the frequencies that are above the cutoff frequency are reduced, while the frequencies that are below the cutoff frequency are left untouched.
The parametric equalizer
This particular kind of equalizer let us control three parameters for each of the filter we want to use.
This first parameter let us decide the amount of boost or the amount of cut that we want to apply.
- Leave the gain at 0 if you don’t want any processing to occur.
- Give the gain a positive value (for example +6 DB) if you want to apply a boost to a particular frequency or range of frequencies
- Give the gain a negative value (for example -6 DB) if you want to reduce the energy of a particular frequency or frequency range.
This second parameter let us choose the frequency at which we want to apply the boost or the cut described above. If you want to filter a range of frequencies instead of just one frequency, this setting represents the frequency that sits at the center of the range.
The Q Factor
This third parameter is less obvious. It let us choose the width of the frequency range affected by our boost or our cut.
- If the value of the Q factor is high, a very narrow frequency range is affected by the filter
- If the Q value is low, a wider frequency range is affected by the filter
Using an equalizer
Correctly using an equalizer is not an easy task. It requires a lot of attentive listening and a bit of practice. To help you out in your endeavor, here are a few guidelines about how to use an equalizer.
Reduce the proximity effect.
The proximity effect is a distortion of the audio signal that is due to the use of directional microphones. A directional microphone is also called a cardioid microphone, because its polar pattern looks like the shape of a heart. Omnidirectional mikes (mikes that record the sound comming from all directions) are therefore not affected by the proximity effect.
The proximity effect is characterized by a significant increase of the energy of the lower frequencies of the spectrum. The following audio recording has been made by a cardioid microphone (an AKG C3000). It has been normalized with the technique explained in the first post of this series, but no equalization has been applied. Pay close attention to the lower frequencies of this sound clip. Notice how ‘Fat’ the sound is because of the proximity effect!
To combat the proximity effect, we will apply a High Pass filter to the audio clip.
I have applied a high pass filter with a cutoff frequency set to 100 Hz and a gain of -48 DB / octave to the following audio file. Compare it with the previous audio file. Pay close attention to the lower frequencies of the signal.
OK, the change is not (yet) obvious and the sound is still a bit fat, so let’s make it clearer.
Make the sound clearer
For a normal human ear, the audible audio frequencies span from 20 Hz to 20.000 Hz. The higher the frequency, the higher the pitch of the sound. The following table comes from Wikipedia and describes each range of frequency by the sensation it induces.
|16 to 32||1st<||The human threshold of feeling, and the lowest pedal notes of a pipe organ.|
|32 to 512||2nd to 5th||Rhythm frequencies, where the lower and upper bass notes lie|
|1024 to 2048||6th to 7th||Defines human speech intelligibility, gives a horn-like or tinny quality to sound|
|4096 to 8192||8th to 9th||Gives presence to speech, where labial and fricative sounds lie|
|16384 to 32768||10th||Brilliance, the sounds of bells and the ringing of cymbals.|
The frequencies that made the voice understandable are centered more or less around 1000 Hz (or 1 kHz).
Below that frequency, are the medium bass and the basses. If they are present in excess, those are the frequencies that made the sound ‘fat’. The following audio file is the same as above, but with a parametric filter that has the following values
- A cut of -20 DB
- A frequency of 165 Hz (Medium bass). For a female voice, this frequency will probably be higher.
- A reasonably low Q factor of 7,4, so the range of frequencies around 165 Hz is affected by the cut.
Other adjustments are to your liking! They all depend on the microphone you used, on the voice you recorded (male or female) and on your personal taste. To help you out, these are a few tips that I discovered along the way.
- Ask yourself “What is there too much in this sound” and not “What is missing in this sound.” That way, you will apply more cuts than boosts and the resulting sound will contain frequencies that were already present in the original sound and not frequencies that are artificially added to the sound.
- Keep in mind that in eLearning, most of your audience will listen to your audio files through cheap headphones. These headphones have a tendency to artificially boost the lower frequencies to cope with the requirement of the noisy and loud modern music.
- Also, keep in mind that you want your voice to be intelligible. Your primary focus should be on the voice intelligibility, not on the respect of the original tone of the voice.
- Test your equalized audio clip with an array of different headphones (your studio headphones, your iPod, etc.). You‘ll be amazed of the differences!
In the following audio clip, I’ve added a small boost of 6 DB around 5500 Hz to enhance the sound brightness a little bit. I have also added a Low Pass filter (also called a High Cut filter) with a cutoff frequency of 16000 hz (16 kHz). Frequencies higher than 16 000 are barely audible anyway!
Equalization is an essential part of the processing of your eLearning audio assets. Equalization will help you shape the sound in order to make it clearer, intelligible and optimized for the cheap computer headphones found in eLearning labs and computer stores.
Correctly equalizing an audio clip is not an easy task. Each clip probably requires its own unique equalization settings that depends on the microphone used during the recording and the person whose voice was recorded among other factors.
Adobe Audition has a very nice parametric equalizer available. It also features a very convenient preset feature that enables you to save your Equalization settings should you reapply them on another audio clip.
I hope this post will help you shape the sound of your own audio clips.
Untill then Happy Captivating