Nyquist theorem Shannon's theorem -> noise sampling + information encoding. Overall if I understand correctly GP says that audio is recorded without any smart techniques besides the basics. Implies that this is bad because memory resources are wasted. Does not discuss how smart techniques might waste computational resources.
Back in the day, home computers did not have enough resources to play music like modern computers do. They instead relied on custom hardware to make noise. An example is the MOS SID 6581, which was on the original C64.
Once you have enough resources this is unnecessary, because you can just make a giant sample of the entire song you want to play, and that's going to sound at least as good as analog recording (and possibly better, depending on what technology you are using to make the analog record).
The Tandy Color Computer did not have such a chip, but it did have a 6-bit DAC (Digital/Analog Converter) that could be used to both sample and play back sound. Yes, it was CPU intensive, but you could get some pretty impressive results (I recall sampling songs, and even playing music with 4-part harmony).
Watching The 8-Bit Guy's youtube channel I'm convinced the Tandy was probably the best DOS computer ever. A shame I never owned one; I went from an XT clone with monochrome graphics (the Hercules card had a charm of its own!) to an AT with a VGA videocard. I understand that the EGA emulation the VGA was capable of was similar but not exactly the same as Tandy color graphics.
It may seem weird, but the Tandy 1000 copied the graphics modes from the IBM PCjr, not the EGA standard. That included a 320x200 16 color mode, as well as 640x200 four color mode that the original CGA couldn't do (because it lacked memory). The IBM PCjr also included a sound synthesis chip (don't recall which one off the top of my head) that may also have been included in the Tandy 1000 (don't know, had the PCjr, not the Tandy).
Not at all! They are just saying that nowadays, computers and sound output devices are so powerful, that they can effectively create arbitrary sounds in the entire human hearing range. This used to not be the case: Sound was synthesized using simpler methods, which could not create arbitrary sounds, but had its very own character. Colloquially called "8 bit music" is part of that.
sorry what is this?