No matter how good your microphone, there will always be a time and a place where you’ll need to “sweeten up” some recorded audio.
It might be rogue aircraft flying overhead, an air conditioning hum, someone on set leaving a door open that suddenly slams or something worse such as a “boomy” room. The worst I came across (to date anyway) was way back when the recording of band was in a room that had mirrored walls.
Fixing audio glitches such as these generally comes down to 5 steps:
- Remove Noise
- Cut Off High Frequencies
- Compress The Dynamic Range
So what do these terms mean? Here is a quick crash course.
Using audio sweetening software such as the free Audition, all sounds are boosted equally until they reach the loudest possible available before distorting. Generally it is accepted that sound in a file should be in the -8 to -14 dB ranges with the peaks getting near +-0 db, but also bear in mind that some sounds are meant to be naturally quiet and others loud.
Using the wave form is the easiest way to pick up anomalies in audio such as those mentioned above, and depending on your audio program, there will be various tools to assist in quickly cleaning these out. In Adobe Audition for example there is an “Auto Heal” tool, “Capture Noise Print” tool, plus “Reduction” and “Reduce By” sliders to all assist in the removal of unwanted noise.
Cut Off High Frequencies
As the human ear (generally) can only hear sounds up to 20kHz it makes no sense to have higher frequency audio in an audio file chomping up space. High sounds are also generally it bit harsh on the ear drums so it is good practice to use a cut off filter to remove sound above the 16kHz level. Some suggest this can be pushed to 17.5kHz if the sound(s) contain high pitch frequencies and thus retain fidelity of the overall sound. A cut off filter will also remove any annoying hiss by the way.
Popular belief is that this is the same as volume levels but it is not as cut and dried as that. It actually relates to the dynamic range of audio which is the difference in decibels between the highest and lowest sounds and compressing reduces this dynamic range.
The dynamic range of a sound should be wide enough to allow variations in the sound, but not too narrow that at low volume sounds are inaudible.
Two terms associated with compression are attack and release. Attack is the amount of time it takes for the compression to ease in once the threshold has been breached and release is how long it takes for the compression to ease out once the audio drops below the threshold.
Equalising a sound lets you, without any fundamental change, accentuate a sound. Many audio packages have a bunch of presets letting you play with various equalisation settings and therefore you can hear what will happen and adjust accordingly. As a rule of thumb, often pushing the midrange and higher levels by between 1dB to 3dB will give a clearer sound.
If you use the free Audacity program for your audio sweetening, this 7 minute tutorial will introduce you to the basics of audio sweetening.
While we have you, on our “Donate” page we have four options: $2 $5, $10 and $20 donation buttons via PayPal (which also allows payment via VISA and Mastercard etc. These are in US$ by the way).
If you decide we are worth a small ONE-OFF donation via this method, we would greatly appreciate it and will allow us to continue at the same professional level we have strived for for almost 14 years now!
More information as to WHY we have added this option is on the Donate page.