If you are using "Application" audio recording then the audio is extracted from the application right before it is sent to the audio subsystem. Provided you haven't changed the gain setting or enabled peek normalization then the audio is the same as what is extracted (and then formatted to the output format). No gain is added to it.
If you are using 'Stereo Mix" audio recording then this is what the audio subsystem is playing. This is affected by the volume you are currently playing sound at.
From CRS @ 24 Jul 2017 08:07