FabFilter

Hi Fred,

It's 32-bit float.

Cheers,

Thank you, Frederik.

-Fred

Frederik,

Given that the output is 32-bit float when the Dither setting is Off:

Is the output being *dithered* from the internal precision (64-bit?) to 32-bit float?

Or is the output being *truncated* from the internal precision to 32-bit float?

Thanks,
Fred

Hi Fred,

We just round from 64-bit to 32-bit float. Honestly there is no way this would make an audible difference in the end result -- it is even debatable whether 16-bit integer dithering makes a difference on real-world signals. (Of course you can put together an artificial test to hear the difference but that's something else entirely.)

Cheers,