Mendelssohn Lieder ohne Worte

Yes, very nice, but ... the sound is unfortunately quite severely overdriven at higher levels.
Here, I've isolated two of the worst moments. Neither of you hear that?

But yes, without that distortion, this performance makes for very enjoyable listening as well as a great demonstration of the appeal of the Bluthner.

_

Well, Pianoteq provides a limiter (hard compressor) that can be turned on and off, at the right of the volume meter directly above the volume slider. The volume can be increased and the limiter will still monitor and reduce the volume of the transient high-volume peaks as necessary to prevent clipping (to prevent reaching the maximum digital level or number) and thereby prevent distortion as much as possible, but although the average volume will be increased, some of the dynamic range will be lost (even the softer parts will be louder).

One of the problems with trying to have a maximum sound level, is that when other programs such as YouTube or SoundCloud transcode the audio (or audio/video) to other formats or resolutions, those programs will often create distortion if there is no "headroom" in the audio file (or audio portion of the video file). Headroom involves having the volume peaks lower than the maximum possible, such as -2 to -4 dBFS (2-4 decibels quieter than full-scale maximum decibels possible in the digital (numerical) format, which varies according to the bitrate of the audio encoding, such as 16-bit, 24-bit or more). Lowering the volume at least a little allows the programs that change the format of the audio to perform their algorithm, which can add or subtract from the volume in a dynamic way from moment to moment, the ability to do so without reaching the maximum level possible and in doing so, create clipping and distortion of the audio file.

You can import a WAV file created by Pianoteq into another program (such as Audacity (for Windows, Linux and I believe OS X) or Ardour, or another audio editing program, and use the "normalize" function to maximize the volume while still leaving headroom, by specifying -2 to -4 dB as the maximum allowable volume for the normalize function to use.