Oversampling in Music Production: What It Is and When to Use It

What Is Oversampling?

You've probably seen the button.

2x.

4x.

8x.

Sometimes 16x.

Oversampling.

Turn it on and the plugin uses more CPU.

But what is it actually doing?

The simple answer is that the plugin temporarily processes the audio at a higher sample rate.

If your project is running at 48 kHz, 2x oversampling processes the signal at 96 kHz internally.

4x takes it to 192 kHz.

Then the signal is brought back down to your project sample rate.

Why?

Mainly to reduce aliasing.

How Oversampling Works

Oversampling happens in three stages.

Upsample.

Process.

Downsample.

The signal is temporarily moved to a higher sample rate.

The plugin does its work.

Then the signal is filtered and brought back down to the original sample rate.

Upsample.

Process.

Downsample cleanly.

That's oversampling.

Why Use Oversampling in Music Production?

The main reason is aliasing.

Aliasing happens when digital processing creates frequencies above the Nyquist limit.

The system can't represent them properly.

So they fold back into the audible range.

But they don't come back as clean harmonics.

They appear as false frequencies.

Distortion.

Saturation.

Clipping.

These can all create new harmonics.

Push them hard enough and some of those harmonics can cross the Nyquist limit.

Oversampling raises that limit while the plugin is processing.

The unwanted high-frequency content can then be filtered out before the signal is brought back down.

Less aliasing.

Cleaner processing.

When Should You Oversample?

Oversampling matters most when a plugin is creating new harmonics.

Distortion.

Saturation.

Clipping.

Some compressors.

Some analogue-modelled processors.

These are the places where aliasing is more likely to appear.

A gain plugin doesn't need oversampling.

A pan control doesn't need oversampling.

The question is simple.

Is the processing creating frequencies that weren't there before?

If it is, oversampling might help.

If it isn't, there's often very little reason to use it.

The Tradeoff

Oversampling uses more CPU.

2x processes more data than running at the project sample rate.

4x uses more again.

8x more again.

That doesn't mean 8x automatically sounds better than 2x.

Sometimes the difference is obvious.

Sometimes it's tiny.

Sometimes I can't hear a difference at all.

So I don't automatically turn oversampling to the highest setting.

I listen.

If 2x fixes the problem, that's enough.

If 4x sounds cleaner, I'll use 4x.

If I can't hear a difference, I leave it alone.

Does Oversampling Always Sound Better?

No.

Sometimes it sounds cleaner.

Sometimes the top end feels smoother.

Sometimes distortion sounds less harsh.

And sometimes I can't hear any difference at all.

Different plugins handle oversampling differently.

And sometimes the aliasing is part of the sound.

So I don't turn it on because a bigger number looks better.

I turn it on.

Listen.

Turn it off.

Listen again.

If it sounds better, I use it.

If it doesn't, I don't.