Digital Music Formats: Tragic, Compressed, and Encumbered

I gave you a GA. I couldn't have said it any better than you. I too am part of the small market segment. I might add; people like what sounds good to their ear even if they never heard the music live, hence no means of comparison. If one has never heard a live symphony, he cannot critique a recording; you must have something to compare it to.

I liked what you posted, though I personally am not a believer in either vinyl or valve amps, mainly for the reasons you mentioned.

We had a recent blog of great similarity that sort of "foundered" when a few more here understood better the problems in the past with RIAA setups and how many companies followed their own ideas......On Wiki is a great article about that.

And as you correctly point out, few of us have enough money to afford good enough speakers......

All in all, the Vinyl/Tube systems are each different to each other, the valves age up in different ways, temperature plays a role as does power supplies of older vintage......then you have the record companies ideas as well, its like playing the lottery.

I prefer my audio CDs on my trasistorised playback equipment with a (reasonably) good pair of speakers and a, I set up the frequency response to suit me, my mood, my music and how I feel on the day......digital is for me the best compromise there is.....all the rest is just sailing on an unknown sea, with no stars, no sun, no radar and no compass.....and no idea.....

The difference between vinyl and digital recordings is analog vs digital. The best vinyl supposedly would get up to the upper frequency limits of human hearing. Not only that, but the vinyl was a continuous form of recording, not a chopped up approximation of the sound waves.

Digital has problems having to do with sample rates. Back in the 70's when I was selling the first digital oscilloscopes for Tektronix, the rule of thumb was that the sampling frequency had to be 10X the highest frequency in the signal in order to accurately record the signal. That is still a valid requirement.

So, if we assume that 20,000 Hertz (I still like cycles per second) is the highest frequency in a music source (band, etc.) then the minimum sampling rate should be 200,000 Hertz. As far as I know, there are no commercial digital recordings made at that rate. In fact, not even 1/2 that rate so it is no wonder that vinyl is technically better in reproducing music that the current digital formats. The digital formats do not have the ability to accurately capture the music.

As for the other issues of tubes vs solid state and zip cord vs oxygen free super cables, I only point out that most of those arguments are made in the marketing departments of the suppliers, not the engineering departments.

That sample rate you cite is not quite correct. The nyquist point only requires 2X the desired maximum frequency.

While I am not a good source to explain the full theory (my friend did his masters thesis on the subject), the sample rate is not the real issue, but the dynamic range is more the issue.

There really is no problem reproducing music digitally given enough analog resolution in the voltage domain and a minimum of 44 kHz in the time domain.

The issue is more to do with how the music is engineered at the studio and the playback equipment's fidelity (clock frequency stability, DAC fidelity, etc.).

Nyquist states that 2X is only to detect the frequency, not the proper shape of the wave. For that, you must have 10x the highest frequency.

If the proper shape of the wave is not a sine wave, then there are higher frequency components present.

Another thought: The ear breaks sounds down into their frequency components in the cochlea before it enters the auditory nerve and the brain. If you generate two tones, say 1000 hz and 2000 hz and you change the phase of one with respect to the other, the waveform of the sum will change, but the ear will not detect any difference. The ear is only sensitive to the amplitude of the frequency components and not to their relative phase.

Rixter makes valid points, but also consider a note is more than one wavelength long, particularly at high frequencies. So you can easily reconstruct the waveform based on that information when you have multiple wavelengths of time to sample over.

There are mathematical proofs that show that a 44 kHz sample rate is adequate to produce reproduction of an audio signal. They are much too complex to go over here and because of that most people would rather believe in magic or rumor than take advanced courses in math and physics just to understand the principle.

So, while it is provable, it requires a lot of work to get to the level of understanding to understand that proof. For most of us we simply have to accept what is presented as true or go back to school and extend our knowledge or reject it and be a skeptic without due cause.

I think we all agree that if you sample a signal, that all the information up to the Nyquist frequency (Fs/2) is captured, at least to the resolution that it is sampled at. Displaying the waveform on an oscilloscope takes more samples because it is a lot less compact way of representing the signal.

Quantizing noise is introduced by digital sampling, depending on the resolution, and some noise is contributed by the lossy Mp3 compression. I'm sure someone must have come up with a noise figure.

Take a piece of graph paper and a pencil, and try to describe a 11K sound wave with 4 points of reference. (44K sampling)

It doesn't look pretty, does it? Now try it again with 8 points of reference. (approximating a 96K sampling) Much better, isn't it?

Now go for 17 points of reference. That's what 192K sampling does for you. Tell me you won't hear the difference between 44K and 96K, and I'll say you're deaf.

What a great paper and written in a manner that anyone can easily understand! (Well I did!!)

Thank you.

It was really good.

To say anything else would be unfriendly.

Well done....

I agree Andy. Thanks AH.

The graph reminded me of how a guitarist defined it the other day on the radio.

"Digital sound perception is like standing in a shower and being hit by cubes instead of droplets".

Once it hits you it runs off and has done its job, that's my interpretation. Even the most basic sounds can carry emotion and that's what its all about. Red House blues could be played over a single speaker of the worst kind and would still give me the creeps, shudders and a wet eye.

Do read up on the history of RIAA, here is a link:-

http://en.wikipedia.org/wiki/RIAA_equalization

A comment at this website is:-

The RIAA equalization curve was intended to operate as a de facto global industry standard for the recording of vinyl records since 1954. However, it is almost impossible to say when the change actually took place.^[1]

Also don't forget that playing a vinyl record actually damages it. The hiss and scratches came as dirt was attracted onto the surface, then we played only "wet" for many years, but that brought its own problems as well.....Which is why years ago I bought the record, recorded it on a high quality tape machine and only played it back from the tape......I still have vinyl that has been only played back once!!!

I have made tests with many people, many times, where I played both back simultaneously and I switched back and forth between the sources, nobody ever got it right more than average - only luck.....and there were only 2 sources to keep apart!!!!

The so called audiophiles that have left my house enraged in the years of the 70's and early 80's.....were sadly many! Enraged with their own ears basically!!! My equipment was high quality and they could not hear the difference between tape and the original vinyl....even today, I am sure done properly, the same test would show that most people only imagine they can tell.....I don't have a tape deck anymore sadly....and it wasn't digital then either, all analog.....but there was still the same controversy that many thought they could tell tape from vinyl, but actually couldn't!!

I can tell the difference, especially on very quiet passages of a brand new classical disk.....but play it a few more times and I cannot tell anymore myself.

I do have two turntables still though to allow me to digitize and clean up old vinyl recordings.....

With most people its a way of life to only love Digital or analog and spit on the other, but it is not a cast iron fact that they can actually tell the difference!!!!

I am of the opinion that a small number of people can tell the difference, but far far fewer than many think......

A good valve amp, properly set up with big electrostatic speakers is still a great listening experience for me personally, though I don't have such equipment anymore.....too expensive nowadays....

There's music available on DAD that is sampled at 192K. I have one by Alan Parsons.

Some might argue that this could be a cash grab by the artist and recording industry. Lots of people would buy a "remastered" album, I think, especially if you could play it on current technology.

Others might argue that a change in format would be a huge money maker for technology manufacturers, especially if they obsolete the old format. Wasn't BluRay threatening to wipe out the current DVD format?

And this might be a non-starter for others. Audio DVD never really took off, did it?

No, Neil is NOT way too late. He's been "working" on this for a long time. I don't know if he understands the science behind it and that science may not be important to him. I think he truly wants a better experience for listeners that won't force them into an expensive niche market to enjoy that experience.

You wrote, "Audio DVD never really took off, did it?"

No, by the time it started (both the Super Audio CD (SACD) and DVD) they were killed by MP3 and its derivatives and here we are today.

Most people do not critically listen to music. It is background while they work, play, use the computer, or entertain guests.

Ask yourself, when was the last time you sat down and actually listened critically to music without conversation, computers, books, TV or some other distracting activity?

99% would be virtually never. So, iPods and MP3 sounds as good as ever to them.

Actually SACD is alive and well, and so is DVD-A. My 2011 Dodge Challenger with NAV plays my DVD-A disks pretty good. Here's a site to look at for more information on hi-def digital music:

http://www.elusivedisc.com/default.asp

Pioneer, Oppo, and others have universal blu-ray players to play these discs as well, and they're new models. The only thing that really hampered hi-def digital music was Sony playing their game with releasing a proprietary format (SACD) to compete with DVD-A discs. That created a format war that blocked progress.

I think there's a theme developing in these comments, and it's one I was hoping my readers would confirm.

There's no appreciable difference in performance between vinyl and a CD or hi-def digital format. We can debate specifics, and there are longevity issues with vinyl, but ... they both can produce music at a fidelity level, given the right playback equipment with dynamic range and distortion specs to reproduce it, that sounds the same to human ears.

The problem with digital formats is in the compression and dynamic range leveling that's going on with mobile players in mind.

Somebody said to me via Twitter yesterday that what they really want is a hi-def downloadable master, uncompressed, that they can grab using a lot of bandwidth and all the time in the world. They'd then choose to play that on something at home, or transcode it and compress it as needed to fit on a mobile device.

Producers/labels aren't making that choice available by mastering music at the clipped roof of loudness, and the examples in the blog post are evidence of that trend. I think that's where Neil was headed without so much as calling out Rick Rubin.

This is a great discussion!

You may consider going the vinyl route over CD because the original recording on the vinyl most likely will have less compression than its remastered CD counterpart or its remastered MP3 counterpart.

Or, if you are like me, the vinyl album has all those pretty pictures that are absent in the CD format. ;-)

A crude analogy would be reading. One's eyesight and vocabulary affect the ability to fully experience reading any printed material. We are limited, first by the frequency response of our hearing. Secondly, given that we each, at one time or another in life, had a "normal" representation of instrumental sound stored in our minds, with the ability to compare an electronic reproduction to compare it to, we are predisposed to what is "should sound like." A live vs. recorded version can be arranged. This, in fact, is the criteria most audiophiles use to judge a sound system. That, and a spatial aural "picture."

The ultimate reason any of us cares, is the feelings and/or emotions that come from music. Being somewhat of a critical listener as AH has alluded to, I have to have a reasonable representation of what I know is in the music's frequency content to "enjoy" it. Bass is an obvious category. But I have to be able to hear high frequencies, too. I just can't enjoy a piece of music on a system that doesn't reproduce the bass and treble content of music for my taste. When bass is deficient, "tinny" is usually the way this gets characterized. I find that my mind is actually filling in the missing content, to some degree, so the enjoyment, as an expectation, or anticipation, still exists. "Taste" is very key to this. Maybe some can be fully satisfied without "high-end" reproduction. I really don't believe it, though. I think if you played a selection of music to anyone on a transistor radio vs. a "decent sound system, who really believes the transistor radio version would be preferred? But my tone control settings might be an anathema to people around me, because of my hearing "curve." The subjectivity component can't be denied.

Most of us (especially as we get older) fool ourselves into thinking that we can hear subtle differences between systems. I think the argument about compression not being able to preserve the dynamic range is the most demonstrable weakness in MP3 music. But most music doesn't require a large dynamic range.

I have always enjoyed listening to music with a good set of headphones. But, to me, that is spatially different from listening to it with open ears. I still enjoy it, though.

If we all had our ears tested every year to see where we stand on the listening "curve" it would still come down to what reproduction system allows us to most freely experience the emotions created by the piece of music we are listening to, rather than lamenting the missing components or distortions we perceive.

... the average listener? That includes a whole generation of people raised on iPods and MP3s. Some of them have never had their music played for them, side-by-side, on what would be termed, a "high-end" music system, so they would know any difference. They are content with the iPod version. So I would say, "No. The average listener won't care, because they don't know any better." It is similar to what my generation would call "the dumbing down" of society.

That last comment on not knowing better is dead on. We've grown used to the differences in a mobile call versus a wired call (i.e. delay, clarity). A whole generation of users doesn't even know what a wired call sounded like. The experience is "good enough" now, and there's no perceived need to go back.

I would further charge that most listeners don't even know how to listen to music, even if it was reproduced on a good system.

Or what constitutes a good system (beyond the bass rattling the hub caps off their car).

I read your blog it's quite good. I really like to listen to music so I used to visit Live Tv Kodi Add-ons. it has a large number of videos collections. mp3, videos and much more in one place.