HDMI is cool, it has audio too, so I can select the TVs as audio output.

But how to aggregate them?

On Mac I think you can just make a Multi-Output Device. But looks like there is no simple way on windows. I tried selecting two outputs in Voiceemeter but it didn't work.

It would be smart if windows could output their sound to whichever display it's mainly on, unfortunately I guess this sort of thing hasn't been designed for in these complex systems.

Imagine if it did a fade effect when the window is partially on multiple displays, that would be really smart!