Voice Conversion for Lombard Speaking Style with Implicit and Explicit Acoustic Feature Conditioning