In the last few years, more advanced upmixing (from stereo to 9.1 or something similar) technologies have surfaced. In the past, there was Dolby Pro Logic for example, but AFAIK that was fairly limited. It did a matrix decode and a few tricks like delaying the rear channels.
The current crop of upmixing techniques like Dolby Surround (not to be confused with the Dolby Surround from the 1980's) and DTS Neural:X seem to be performing differently. Here is an old news article describing/selling it: https://www.soundandvision.com/content/meet-new-dolby-surround
Apparently, Dolby Surround splits the source materials into frequency bands, which are separately steered. Does anybody have any more information on how this works in detail?