Docs
Rodrigo has complained that most modern music is all alike (and dull, with few exceptions). The complaint is not new. There's a bucketload of reasons for this...



1. Compact Disc Audio, the Redbook standard of digital sound recording. It's too low-resolution, and places many constraints on the music. Without too many details, the problem is an overly coarse sampling grid. Sampling quantises signals (sound in this case) into discrete series of dots. The coarser the coordinate grid, the less dots in a waveform, and there is a couple important issues with the axis resolution. In digital signal sampling, amplitude is represented by current rise/fall coded with bit assignment to each frame, e. g. 16-bit, 24-bit, 15-bit, 14-bit, 32-bit, etc. The issue is, unlike open reel tape which does not exhibit progressive signal detail loss until hitting hiss/noise floor, amplitude detail in sampling is lost with loss of bit depth. 1 bit of resolution is assigned to each ~6 dB of dynamic range. 0 dB is the highest loudness a digital format can go, "digital ceiling", from there the loudness of a signal is counted in negative dB. So for CDs, the peak detail is 16-bit at 0 dB, which equals 65536 possible amplitude division levels divided by two (negative/positive current flow; of these, only the positive is actually played by headphones/speakers). The issue is, at ~-6 dB and lower there're 15 bits, which is half that (32768/2, so really only 16384 levels for the audible positive signal). At ~-12 dB and lower there're 16384/2 levels, which already starts sounding rather cold/hollow. The result is that most modern CD-mastered music is kept with average dynamics around 12 dB - it's simply maximised loudness (possibly monotonous, but necessary to preserve some warmth). Search for "loudness wars" - there're some good clips on Youtube.com about it. There're other issues from underquantising, such as loss of amplitude detail and rectification, but usually it all just tends to sound cold and hollow unless there're enough bits allocated to amplitude. In reality, CDs are inferior to open reel and in some ways cassette tapes (you don't hear the hiss, but the waveform is coarser). 96 KHz/24-bit already starts approaching vinyl quality, but common CD 44 KHz/16-bit is simply hollow.
Sampling frequency is also an issue that's not understood correctly. Whereas a Nyquist limit of 44100 Hz in theory is enough to sample less than 22050 Hz, in reality what you'll get at right under 22 KHz is technical noise. This is because there're only 2 coordinates/cycle (practically 1 for the audible positive part). Waveforms are distorted. Practically, what matters is the number of coordinates available for a given frequency bandwidth. To more or less accurately present a varying waveform, a minimum of 8 coordinates (4 coordinates/half-cycle) is needed. What this means is that for 20 Hz of hi-fi range the sampling rate has to be over 160 KHz.
The fallacy of CD audio is that it only about accurately covers 44100/8=5512.5 Hz, which is just about the midrange. Treble and space definition are distorted, the higher frequency range where the actual vitality and energy of a musical performance is, is distorted, the overtones are damaged. CDs kill off presence and vital energy, which is why vinyl lovers call them flat and lifeless.





2. Wrong tuning. Modern Western tuning is overly harsh at A4=440 Hz.
Without going too much into detail about musical scales, for a Western equal-temper scale that can have corresponding notes on many octaves, the proper tuning is at most around 434 Hz, properly 430.53 Hz as this makes the C/do notes fall on multiples of 8 (e. g. 64 Hz, 128 Hz, 256 Hz, 512 Hz...). This actually used to be the tuning a couple centuries ago, and in practice tuning depended on the fork used (ex.: Handel used several forks, one tuned to 418 Hz, another to 422.5 Hz, another at C=512 Hz). The lowest tuning for an acoustic instrument is 380 Hz for an old English organ (pipe instruments have fixed pitches). Historically tuning forks used to go higher and higher approaching the 20th century, until the Anglo-Saxon (US/UK) standard of 440 Hz was adopted in 1918. The background reason for higher tuning (even higher than 440 Hz, e. g. 451 Hz at La Scala, 455 Hz at Crystal Palace in London) is simply loudness. Acoustic instruments play louder when tuned higher. So for the gallery listeners (cheapest tickets, and farthest from the stage) to still hear the orchestra and singers in detail, La Scala tuned up to 451 Hz (nevermind the strain this produced on singers).

The issue with high tuning is, it's overly stressful. It's harsh and is associated with hopelessness and despair and undue overdriven effort. It may empty the feelings and promote mindlessness (an asset to military orchestras that tuned to 452 Hz, and Wagner, whose aim was to capture and lead the listener as if on a leash). While it may work for some music (Wagner, as an example, was party of the high tuning above 450 Hz), it does not work for properly warm, loving, relaxing, lush music. Harmonical music. Lower tuning is usually slower, easier to digest/discern, and more welcoming and naturally joyful.





3. As a conclusion, CDs exacerbate the overly high tuning and combined the low-res digital formats and high tuning promote hard-hitting, highly companded (compressed/expanded) records. The dilemma a CD places is: either the songs are kept in the higher -12 dB dynamic range, or they are washed out and sound cold and hollow. The issue with high tuning to 440 Hz is, first, it masks some of CDs' inherent lifelessness by sheer impact of overstressed musical instruments, second, it evokes despair. It works for cold, neurotic, monotonous and dull music which has become more or less commonplace. CDs, on the other hand, favour highly saturated, highly expanded dynamics such as overly "punchy" drums and more piercing vocals (you can't master soft and subtle for CDs as easily as for higher-res digital or tape/vinyl, the life gets lost unless most music is in the upper 12-18 dBs). Basically, it's a trap. Singers are forced to overdrive themselves like electrical guitars, synthesisers, which are often low-res 44/16 samplers, are driven into cold, hollowness and screechiness (have you ever tried to work with a 44/24 "realistic" sampler of a string orchestra?), and the overall result is all records being alike: dull and/or using the same tricks for standing out and preserving at least some life on CDs.

As a conclusion, records have to be processed and released in a digital sampling resolution of at least 96 KHz/24-bit, and with a master tuning of 430.53 Hz or any scale with an octave middle of less than 435-436 Hz.



http://www.mcgee-flutes.com/eng_pitch.html -- here's a curious article on different tuning pitches, which also tells how the current (crazy) tuning pitch came to be.

You have no rights to post comments