Sound, measured thousands of times a second.
A microphone gives a smooth voltage. Digital audio writes down its value at fixed moments (the sample rate), with a fixed number of bits each (the bit depth), once per channel. Three choices set everything else. Change them here and the whole page follows.
A tone below half the sample rate is captured exactly: only one smooth wave fits the samples. Above it, the same samples also fit a lower tone, and that lower tone is what plays back. Real recorders filter those frequencies out first. This samples a pure tone with no filter so you can hear why that filter exists.
After a 44-byte header, a WAV is just the samples, one per channel per moment, interleaved. These are the first moments of the tone above at your settings, exactly as the file below stores them: little-endian, signed, except 8-bit which is unsigned with silence at 128.
Uncompressed size is simply rate × depth × channels × time. Compressed formats sit below for comparison; their size depends only on the bitrate.
Sample rate is how many times per second audio is measured. CD uses 44.1kHz (44,100 samples/second). Higher rates capture more detail.
Bit depth determines the dynamic range of audio. 16-bit (CD) gives 96dB range. 24-bit (Hi-Res) gives 144dB range.
For recording and editing, yes. For final listening, the difference is essentially inaudible on consumer audio equipment.
FLAC for lossless archiving. AAC or OGG for compressed listening. WAV for professional production.
Lossless audio (FLAC, ALAC, WAV) stores all the original data perfectly. Lossy audio (MP3, AAC) discards some data to achieve smaller file sizes.