CDs Vs LPs

sleepwalker65

The video that @cleeds referred to is heavily slanted toward trusting that the conversion back to analog fills in the missing pieces with a perfectly synthesized replacement.

Actually, the video is not slanted at all. It scientifically and visually demonstrates that your claim of the CD’s "fundamentally flawed ... staircase effect and artifacts" is mistaken, no matter how intuitively reasonable your claim might appear to be.

As as a person who can’t put on the blinders and “un-see” what I know ...

What you think you "know" is mistaken. Don’t get me wrong - I have my issues with the compact disc. But claims of the "staircase effect" have been shown to be false. The reason this is important is that if you seek to get the most from CD - and many of us do - it’s important to know what it gets right, and what it gets wrong. And claims of the "staircase effect" are just nonsense, as shown in the video.

@cleeds the fact is that while the data is still represented as a staircase, (lollipop diagram if you prefer), the nature of sampling means that some information is left out during digitizing and then on playback, it is artificially synthesized. That is the indisputable flaw in the process. No matter what quantization resolution and what sampling rate, you cannot escape this fundamental. 

As is usually the case, I am in 100% agreement with the post by Larryi. As an illustration of his point about dynamic compression of classical recordings, I’ll mention that I have had occasion to examine the waveforms of a couple of classical orchestral recordings that I believe have been engineered with no compression. Namely the Telarc recording of Stravinsky’s Firebird Suite, Robert Shaw conducting the Atlanta Symphony, and the Sheffield Labs recording of excerpts from Prokofiev’s Romeo & Juliet, Erich Leinsdorf conducting the Los Angeles Philharmonic. I have done this on a computer, using a professional audio editing program.

In each case I observed a dynamic range (the difference in volume between the loudest notes and the softest notes) of approximately 55 db, which I consider to be an amazing "tour de force." Correspondingly, when I listen to those recordings at my preferred volume settings, with average SPLs somewhere in the 70s at my 12 foot listening distance, the softest notes are in the vicinity of 50 db or so, with the loudest notes reaching close to 105 db at my listening position. Also correspondingly, a dynamic range of 55 db means that the amplifier must supply approximately 316,000 times as much power to reproduce the loudest notes as to reproduce the softest notes. (And 316,000 is not a typo).

I believe it is safe to say that many and probably the majority of audio systems, including the majority of systems used by serious audiophiles such as those who are members here, would not be capable of handling such recordings at volume levels that are high enough to satisfy many listeners. With the most common limiting factor perhaps being the maximum volume capability of the speakers. And perhaps in some cases, and at some times, the ambient noise level in the room.


On another subject:

Sleepwalker65 10-4-2018

@dynaquest4 and @cleeds you are glossing over the fundamental issue with going digital: it does not capture the infinite range of undulations. Rather, the process quantizes the input program material at the sampling frequency, and then stores it as a sequence of discrete samples.


What you are mistakenly thinking of, is the reverse process, taking that stored sequence of samples and generating a facsimile of the original analog program material, doing bit-sum averaging to compensate for dropouts.


Perhaps this is the missing information you needed to see why digital is inherently flawed.


Sleepwalker65 10-5-2018

@cleeds the fact is that while the data is still represented as a staircase, (lollipop diagram if you prefer), the nature of sampling means that some information is left out during digitizing and then on playback, it is artificially synthesized. That is the indisputable flaw in the process. No matter what quantization resolution and what sampling rate, you cannot escape this fundamental.

First, I’m not sure how familiar you may be with the Nyquist-Shannon sampling theorem. There is no "bit-sum averaging" involved in the process. And if by that you meant "interpolation" that is not involved either. And the reference to playback being "artificially synthesized" is a misconception. What is involved is low pass filtering, in the recording process at the input to the A/D converter, to prevent "aliasing," and in the reproduction process to reconstruct the analog signal. And the 16 bit quantization that is used in CDs can, if well implemented, provide a dynamic range of approximately 96 db (actually slightly more than that), which can be further enhanced by "dithering." And theoretically/potentially, at least in the case of an infinitely long series of samples, the 44.1 kHz sample rate can support a bandwidth of 22.05 kHz, without any "averaging" or "interpolation" or "artificial synthesizing" or information being "left out."

If you believe that the LP medium is inherently superior to the CD medium I suggest that instead of focusing on (and misapplying) theory you consider the potential side effects of the two filters that I mentioned, and probably more importantly on the engineering of the recordings, and perhaps most importantly on the quality of the circuit designs that are used in the specific equipment that is used for both recording and playback.

Regards,
-- Al

P.S: Regarding the mention of 96 db and 55 db in my previous post, to be sure it’s clear I should add that those numbers refer to two different things. The mention of 55 db refers to the dynamic range of the music (meaning the difference in volume between the loudest notes and the softest notes), on a couple of recordings having extraordinarily wide dynamic range. The mention of 96 db refers (approximately) to the dynamic range of the medium, which must be far greater than the dynamic range of the music, so that the information that is present within each note can be captured.

Regards,
-- Al

sleepwalker65"What you are mistakenly thinking of, is the reverse process, taking that stored sequence of samples and generating a facsimile of the original analog program material, doing bit-sum averaging to compensate for dropouts.'

You are confused, disoriented or misinformed provided that what is under discussion hear is as I believe it to be which is the Compact Disk Audio Standard as defined by the "Red Book" protocol as promulgated by Sony/Phillips there is no "bit sum averaging." There should be no "dropouts" unless of course there is a substantial failure, defect or fault in the playback system because the CD audio standard relies on Cross-Interleaved Reed-Solomon Coding by using 24 8 bit words and encoding them in a RS code with parity check symbols.