From the description, I assumed that "night mode" would be useful for high background noise environments, such as in a moving car.
I have here a track from an orchestral soundtrack that has a quiet into with at average (RMS) -43 dB, followed by a louder section at -17 dB. A 26 dB difference. That's obviously too much dynamic range to be listenable in a car. One has to turn up the soft part, and turn down the loud part.
After processing with both "Volume leveling" and "adaptive volume | night mode" enabled the loud part was at -14 dB and the soft part at -38 dB. That's still a 24 dB difference. Not a significant change.
Worse, though, after processing by "night mode", there was considerable clipping in the loud part. This even though I had enabled "clip protection" in the DSP Studio Adaptive Volume dialog.
"Peak level normalize" isn't right either: "Boosts low volume content, while preserving dynamic range." No, I don't want to preserve dynamic range.