Oversampling is a technique of digital signal processing that helps audio software plug-ins to minimize the level of artifacts often caused by aliasing distortion.
The oversampling process involves sampling audio material greater than its Nyquist rate. For example, if you want to capture (or record) high-quality audio with a bandwidth of 20 kHz, the sampling rate must exceed 40 kHz. That's why the standard sampling rate for a music CD is 44.1 kHz.
Likewise, when the sampling rate is 44.1 kHz, the Nyquist rate will be 22.05 kHz. A steep anti-aliasing filter removes information above 22.05 kHz.
So it goes without saying that recording at 44.1 kHz sample rate isn't necessarily to hear sound up to that frequency range but to avoid the aliasing errors that would rather occur during playback.
However, after the plug-in has processed the signal, it needs to be downsampled back to its original 44.1 kHz or whatever the sampling rate of your music project is.
The oversampling process involves sampling audio material greater than its Nyquist rate. For example, if you want to capture (or record) high-quality audio with a bandwidth of 20 kHz, the sampling rate must exceed 40 kHz. That's why the standard sampling rate for a music CD is 44.1 kHz.
Likewise, when the sampling rate is 44.1 kHz, the Nyquist rate will be 22.05 kHz. A steep anti-aliasing filter removes information above 22.05 kHz.
So it goes without saying that recording at 44.1 kHz sample rate isn't necessarily to hear sound up to that frequency range but to avoid the aliasing errors that would rather occur during playback.
Oversampling in audio software plug-ins
The amount of oversampling depends on the sampling rate of your current project. Therefore, when you set your virtual instrument or a sound effect plug-in to operate at 4x oversampling, digital signal processing will occur at 176.4 kHz, assuming your music session has a sampling rate of 44.1 kHz.However, after the plug-in has processed the signal, it needs to be downsampled back to its original 44.1 kHz or whatever the sampling rate of your music project is.
Some cons of oversampling
Even though the option of enabling oversampling on an audio software plug-in may increase processing accuracy for non-linear plug-ins, some trade-offs are to be encountered, such as:- Increase in audio latency (with some plug-ins, that is)
- Increase in CPU usage (you can't oversample every single plug-in)
- Bad downsampling or anti-aliasing filters can mess up your sound
Examples of some audio plug-ins that support oversampling are:
- Fab-Filter Pro-C (compressor)
- Fab-Filter Pro-L (brickwall mastering limiter or maximizer)
- Cytomic: The Glue (compressor)
- Image-Line Sytrus (FM synthesizer)
- Xfer Records Serum (wavetable synthesizer)
- IK Multimedia Stealth Limiter (brickwall mastering limiter or maximizer)
- etc. (The list is endless; go figure it out!)
