Apparently analog formats give a "full" and "seamless" picture of the soundwave. They are not quantized like digital, but have a "perfectly" smooth representation of the waveform.

Is this really true? Isn't the accuracy of their waveform representation limited to their very physical nature?

E.g. reel to teel tape...whatever the material is that makes up the picture on the tape (I'm not sure what it is :S), doesn't that, when you take it under the microscope, have holes as well (like the "space" in between each sample in digital audio)? Surely the vinyl lovers can't claim that "analog" is "perfect".

If this is true, this brings me to a question. Could a ginormous reel to reel tape, theoretically be more accurate than a normal sized tape (since it has many more "dots of dye" etc. to paint the waveform)?

This is a myth. It ignores the well-understood mathematics of the digital representation of a bandwidth-limited analog signal.

It's not the dots on dye that matters, it's the speed of tape against the head. R2R tape recorders had various speeds for recording, and were "fatter" than ordinary tapes in cassettes, and that elliminated most of hissing and crosstalk, when two layers og magnetic tapes influenced each other.
As far as I know, R2R and digitized signal from that tape can't be ABXed, I've read just a few test tests where result wasn't ABXable.

The output of an analog or a digital system is a 'smooth' ,analog wave. The digital one will be more *accurate* reproduction of the input, all other things being equal.

One could view magnetic tape as 'quantized' as well, by the particulate nature of the coating....LPs as well, by the molecular nature of plastic, all of whicin a totally theoretical view, disallow *absolute perfection* in the 'smoothness' of the reproduction. (Absolute perfection isn 't necessary, though; and other forms of distortion accompanying a technology that involves dragging a rock over a rotating plastic disc and converting stylus motion to voltage, are far more important to audibility! )

There's a website out there somewhere I've read, that actually (and mischievously) discusses how 'analog' is at least partially 'digital', and vice versa. Can't locate it right now, I must have it bookmarked on a diferent computer.

Its not perfect, its corrupted by noise that causes small deviations from the ideal waveform . . . same as digital.



Its not true at all.

No, they don't. All analog domain signals and signal processors add noise, linear distortion (frequency response, phase, and non-linear distortion(THD & IM).



Is that the good news or the bad news? Digital signals can have any desirable level of accuracy, just use more bits per sample, and more samples per second. Analog signals have definate limits, both practical and theoretical. Some of the dirtyist audio around is found in the analog domain.



Not so. Take a good look and you see that anlog waves are slightly mishapen, and have noise riding on them.




No.



That's true for both digital and analog signals. However, the accuracy of analog signals have definate limits (such as thermal noise). The potental accuracy of digital signals is almost unlimited. Just add more bits per sample and more samples per second.




Exactly. Analog tape is composed of microscopic entities known as domains, and domains represent a definate amount of magnetization, no more and no less. If you try to change their magnetization with a varying magnetic field, each domain will resist change until a threshold is reached, and then they flip and have exactly the opposite magnetization, but the amount of magnetization will be the same.




The smarter ones know better, but there are all those other ones that just graduated from the corner hifi shop school of audio technology.




Yes, the types of magnetic tape that used to be used by professionals were wider, thicker and moved faster than ordinary cassette or home reel-to-reel. tape However, you have to double the relevant parameter to get maybe a 3 dB improvement, so even making it much biger and moving it much faster had practical limits.

Some "audiophiles" believe that. Most of us here don't. We believe the science, the math, and the results of double-blind listening tests. And, some of us remember the "snap, crackle, and pop" of vinyl, the hiss of cassettes, and the (usually) poor frequency response of both.

Open reel tape was pretty good. I had a reel-to-reel tape recorder, but music wasn't released on tape.

With analog, you have to battle noise, distortion, frequency response variations, speed variations... did I forget anything? All of these things are (or can be) almost non-existent in digital (at least 10x better). With expensive analog equipment, you can minimize these defects, but after you acquire high-end equipment, then you have to find exceptional recordings. (i.e. Exceptionally good recordings on vinyl in exceptionally good condition.)

Some people do prefer the "sound" of vinyl, and some crackle doesn't seem to bother vinyl fans. Most people prefer CDs. The noise from records always bothered me! I'd be listening to a record where I knew there was a defect, and I'd be listening for it. Or, I'd hear a "tick" and listen for it again when the record came-around another revolution... Very distracting!

There is a filter to "smooth out" the raw digital output*. The Nyquist theory says you can fully-reconstruct a signal as long as your sampling frequency is more than twice the signal frequency.** This was all figured-out mathematically before digital audio existed.



* But, even without a filter (which some cheap soundcards don't have) digital usually sounds better than analog.... The harmonics in the stair-stepped waveform are ultrasonic, and probably filtered (mechanically) by the speaker.

** The signal is both time-quantized and amplitude-quantized. The Nyquist theory applies to time-quantization. But, 16 bits of amplitude quantization is better than any analog system... The quantization steps are smaller than the noise (uncertainty) in an analog medium.

P.S.
If you've ever heard good sound in a movie theatre with a good sound system, that's digital. Or , if you've ever heard a good home theatre system, that's digital. Some of the best recorded sound I've heard is "Dolby Digital Surround" on a DVD, and that's actually lossy compression!

And don't forget, each time you play a vinyl or tape, you wear it down a little. Might not mean anything a first, but play it a lot and you'll see...
So these guys with open reel would purchase a vinyl, copy into reel and store the disc until the reel tape degraded enough to make another copy.

Different times indeed

I don't remember why I didn't do that! Probably because I thought I was getting better quality from the "1st generation" vinyl. Or, maybe it was the cost of the blank tape, or that tape wasn't quite as convinient as vinyl... Or maybe because it's harder to play just one song. I wish I would have done it... I tried to take care of my records, but most of them are in rather poor condition.

If that is true than the digital format still wins because perfect copies can be made. The analog copy always looses a bit of quality and the analog original will deteriorate with time and use (the digital one can be backed up without losses.)

Been there, done that on a quarter track revox A77 in the days of.

Also, with large reels you coud record several albums on one side of a reel of tape for more continuous music.

I would like to know about these ABX tests. The only test that I am aware of* is between a recording done to analogue tape and then converted to digital and a straight digital recording. That makes the test much easier administer, but doesn't really compare the two mediums properly.

Any arguments for the use of analogue tape are not for it accuracy, but that it's inaccuracy or character is pleasing an may even help the recording illusion. Another argument for analogue tape is that the recording process is different and creates a good 'vibe'. (i.e. you can't use digital tricks to fix mistakes so you better have a good performance)


*I am not very aware of tests in general

found it

http://www.st-andrews.ac.uk/~www_pa/Scots_...rt12/page1.html
http://www.st-andrews.ac.uk/~www_pa/Scots_...rt12/page2.html

Something you have to remember about analogue vs digital is that no matter what esoteric technical or audiophile voodoo arguments you can find, digital always wins out at a given price point, and particularly in a reasonable price point.

Sure, in theory a very well designed expensive analogue system might offer better sound than a cheap CD player. However comparing a cheap CD player to an expensive analogue system is silly. You'd want to compare an expensive digital system to it.

This is the biggest problem I see. People talk about the merits of their record player, and compare it to a normal consumer CD player. Ok well if you are spending the kind of money it takes to get a good turntable, head, stylus and so on, you are in the arena where you could buy a DVD-A/SACD player and perhaps a quality external DAC if you wanted. So that's the kind of setup you need to test against.

Any time there is some esoteric or outdated technology that is being claimed to be better than the current state of the art, ask yourself if it is an even comparison. Are you talking systems that are the rough same amount of money? If someone says a $50 cassette player sounds better than a $50 CD player, ok, well that is something to check out. However if someone says a $5000 turntable with a $3000 tube amp is better than an iPod, well that isn't meaningful. Even if the claims are true, it proves nothing more than that you get better sound when you spend more money.


Also digital has an advantage that analogue can't touch no matter what: No degradation of quality through playback or copying. With any analogue format, every copy is imperfect. There is a bit more noise, more distortion, more wow and flutter added. No matter how good the equipment is, the copy is never 100% of the original. Now this is particularly problematic in that to mass produce something, you have make copies to many levels. You don't make all copies form an original, you instead have to copy the original, and use the copies to make copies, or perhaps even copies of the copies to make copies. Then if you the end user want a copy, there's another generation and so on. Also, analogue media degrades a bit when used. Each time you play a tape or record, it loses a little bit of information, gets a little worse in quality.

Neither of these apply to digital. All copies are perfect, and verifiably so. a 1000th generation copy is indistinct from the original. Also digital formats don't lose information as they get played. They can get damaged, but of course a copy can be made any time before that, and there is no slow reduction in quality due to wear. They work or they don't.

Hi,



Why is it so important to state that "in theory" analogue is "not bad" but the real thing is just digital?

Reminds me of those politicians when they say "I think you are absolutely right but you're wrong!".

This discussion is pointless.

Regards, Pantofaron (Who happens to be a fan of both: analogue and digital)

If you want to get really silly about this, quantum theory implies that there are fundamental limits to both space and time. In other words, there is an absolute minimum dimension something can be and a minimum interval of time. So at the lowest levels the universe is 'gritty', so analogue waveforms are an illusion.



Rust for the most part or Iron oxide, though chrome rust Cro2 and cobalt and have been used as well as thin metal films.



No, the particle sizes on the tape are tiny while the recording head is comparitively massive. Imagine a huge road with 1m wide pot holes but the tyre rolling over it is 4km in diameter..




I don't think they would claim perfection. Tape hiss would be one problem.

Digital encoding can PERFECTLY reproduce an analogue waveform, it is a fundamental tenet of the Nyquist-Shannon sampling theorem.



Studio tape decks could use 50mm wide tape, but that was mainly because multiple tracks were layed down on it. the speed the tape moved past the heads was more important than sheer physical size. Have a read:

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

Speaking of that - imagine quantizing with 64-bit range. Theoretical dynamic range would be bigger than Big Bang itself - problem is, you won't find any equipment capable of playing that back in the full range

Modern digital audio recording of, say, 24-bit accuracy is significantly more "precise" than any other analog medium could possibly hold. And once you "quantize" it - it stays the same as long as the digital representation is intact.

This cannot be said for analog magnetic media - not to mention various distortions in the analog domain.

~ Development of an Ultrawideband Magnetic-Tape Recorder