by Kbyrnes (2018-12-11 21:58:56)
Edited on 2018-12-11 22:00:22
In reply to: I disagree with your first sentence posted by mjv8198
...The superiority of an analog mode, vinyl, to a digital mode, Spotify, cited upwards in this thread is not based on analog being inherently better than digital, because that's not the case.
I suppose "better or worse" could be supported, but it would have to be on some quantitative basis: which system can better replicate the sound you hear live?
One thing that the writer, who has presented a nice basic explanation of recorded sound, does not get into is the difference between lossless and lossy compression, which is right at the heart of digital sound quality. See link for one brief discussion.
EDIT: After posting this I read Duke Camaro's post, which is exactly on point.
are so many variables.
My comments are based on discussions with my father you was a sound engineer early in his career which are confirmed in the article I attached . Digital chops analog and averages. However he felt the best format was tape but there are conditions there as well. A 1/4 < 2" tape etc. Recording speed is important as well higher speed recording and playback > lower speed.
In the end there are so many factors such as equipment quality both recording and playback, even playback environment. There are concert halls for a reason.
I have enjoyed the conversation
.
...I went back and re-read the math PhD's article. I wasn't a math major, so beware of the following! The digital approach is indeed a sampling, so kind of an average, but the phenomenon is more like integration, where the number of divisions is made smaller and smaller to approximate the smooth curve of whatever function you're dealing with.
There are graphic comparisons of analog to digital that could be quite misleading, like this one:
One sine-wave period corresponds to the stated frequency; if we're talking about A 440 (the note A on a piano above middle C), then one period of the sinusoidal wave form is considered to represent 440 hertz. The highest note on a conventional acoustic piano, C8, has a frequency of about 4200 Hz (in equal temperament).
So going back to that image above, if one period is for A 440, the period last for 1/440 of a second. The graphic image shows about a dozen divisions, so about 440 x 12 = 5,280 components per second. But sampling rates are commonly over 50,000 Hz, meaning 50,000 components per second. So imagine those 12 boxes in the image each sliced into about 10 narrower pieces.
Current digital technology samples the sound data at rates far above the ability of the human ear to hear (about 20,000 Hz). In live music, the sounds combine pure tones into a complex tangle with harmonics reaching well above what our ears can hear. It would be interesting to take recordings done on wide-track tape recorded and played back at high speed to digital recordings at, say 96,000 Hz and ask people with experienced ears to compare. Not to mention--digital recording still often uses analog components; e.g., "digital" microphones with diaphragms, etc.
Sound reproduction is a fascinating field. A few years ago I valued the Pritzker Pavilion in Millennium Park, Chicago, and was given a mini-lecture on and demonstration of their outdoor sound system, which serves people seated up front and then sitting on a lawn that extends several hundred feet back and away from the stage. They fashioned a system with delayed sound reinforcement using LARES technology, so that the sound from the speakers, wherever they are placed, arrives at the auditor close to the same time as the actual sound from the stage. See link for a little story about that.