Lorewegian 📚 on Nostr: the typical format that you see for processing digital audio is called PCM - pulse ...
the typical format that you see for processing digital audio is called PCM - pulse code modulation.
it's not a very helpful or descriptive name though.
what it refers to is storing audio waveforms as a series of instantaneous sound pressure measurements called samples, typically at rates of 48000 Hz (samples per second) or above.
the samples can be taken at varying levels of precision. 16-bit PCM might use a value of +32767 for the maximum positive sound pressure, 0 for no pressure and -32767 for the maximum negative pressure.
a tone would consist of a series of such samples that trace out a repeating sound waveform if you plot them on a graph.
actually measuring or reproducing a sample to degree of precision electrically is difficult though. using a classic converter design, you start running into difficulties above 8 bits of precision - insufficient for high-fidelity audio reproduction.
sometime in the late 80s, a new method of sampling was devised. around that time, manufacturers began to switch to what is sometimes called 1-bit converters. the technically correct name is sigma-delta converter.
instead of trying to take absolute sample values, what they do is track the relative change since the last sample, but in a very simple way: 1 means a step up, 0 means a step down. and they do this at 128-256 times the advertised PCM sample rate.
to produce a PCM sample from this bit stream, they add these samples up digitally to produce an average.
SACD - the failed successor to the CD - ditched PCM altogether and uses this bitstream format on disc instead.
PCM hasn't been the "native" format of digital sound for at least 30 years now. but it's what everyone is used to, and it's easier to process than the raw sigma-delta bitstream, so it has stuck around.
it's not a very helpful or descriptive name though.
what it refers to is storing audio waveforms as a series of instantaneous sound pressure measurements called samples, typically at rates of 48000 Hz (samples per second) or above.
the samples can be taken at varying levels of precision. 16-bit PCM might use a value of +32767 for the maximum positive sound pressure, 0 for no pressure and -32767 for the maximum negative pressure.
a tone would consist of a series of such samples that trace out a repeating sound waveform if you plot them on a graph.
actually measuring or reproducing a sample to degree of precision electrically is difficult though. using a classic converter design, you start running into difficulties above 8 bits of precision - insufficient for high-fidelity audio reproduction.
sometime in the late 80s, a new method of sampling was devised. around that time, manufacturers began to switch to what is sometimes called 1-bit converters. the technically correct name is sigma-delta converter.
instead of trying to take absolute sample values, what they do is track the relative change since the last sample, but in a very simple way: 1 means a step up, 0 means a step down. and they do this at 128-256 times the advertised PCM sample rate.
to produce a PCM sample from this bit stream, they add these samples up digitally to produce an average.
SACD - the failed successor to the CD - ditched PCM altogether and uses this bitstream format on disc instead.
PCM hasn't been the "native" format of digital sound for at least 30 years now. but it's what everyone is used to, and it's easier to process than the raw sigma-delta bitstream, so it has stuck around.