The biggest difference I hear between digital and analog

[LL21]

Making an analogy with film is interesting for sure - film also has a parallel to the modulation noise of analog tape, and that's film grain. Digital photography is grain-free, so the perceived contrast ratios are higher. This is parallel to how I perceive the sound of piano in (well-implemented) digital - its 'high contrast' whereas analog tape recordings of piano have some softening.

Analog lovers note that digital sounds 'grainy' though - what they're hearing is not the (inherent) quantization (pixellation) noise, rather the noise modulation resulting from poor digital implementation.

interesting expansion of the analogy...thanks!

 

[opus111]

I've been trying to think of a photo analogy to digital's modulation noise - perhaps its chromatic aberration. I have had cameras with chromatic aberration and these have generated false colour in scenes where there are many abrupt changes of contrast. The classic one is leaves of a tree against a bright sky. The effect is to make the tree's leaves seem 'grainy' from a distance, its only when you zoom in to the picture you notice that one edge of the leaf is one colour, the opposite side is the complementary colour. However this is a fault with the lens, not with the camera's being digital. That's analogous to noise modulation in the sense that its not an inherent problem with digital.

 

[Atmasphere]

I'm not saying that today's commonly used digital storage formats are the same as analog, but those of the immediate and easily foreseeable future are "analagous". 128x or even 256x DSD and 32/192 PCM and/or 32/384 DXD have data flow rates similar to if not greater than analog tape's data flow rate, because like it or not both analog tape and digitally stored music involve discrete "particles" which store data.

Even our hearing is "digital" in the sense that the hair cells in our ears only respond at specific levels, not at an infinite number of non-discrete levels. The auditory nerve and subsequently the auditory processing centers of our brains then convert those discrete signals into what we perceive as a smooth analog waveform.

Neither of these statements are true, unless you are able to assign a level of intensity to a digital bit, and of course that is not possible. In the ear, each member of the silia in our ears responds to a narrow group of frequencies, but expresses also the level of amplitude. The analogy does not hold up to digital at all!

To those that think that somehow at the lowest level of analog's recording abilities, its somehow digital: what you don't get is that the basic 'particles' (which is an inaccurate model to start with, drop the story of an 'analog tape's data flow rate' and this will be easier to understand) will have an analog value, not a 'yes or no' value! This is a feature that no digital recording system has. It really is that simple. But just in case there is still confusion, I will put it another way. Digital bits are either on or off. No inbetweens, IOW no value of the intensity of 'on' or 'off'. In this way digital is unlike analog regardless of the limits of either technology.

 

[rbbert]

Neither of these statements are true, unless you are able to assign a level of intensity to a digital bit, and of course that is not possible. In the ear, each member of the silia in our ears responds to a narrow group of frequencies, but expresses also the level of amplitude. The analogy does not hold up to digital at all!

What in the world are you talking about?

1) The whole point of a PCM digital "word" is to assign a level of intensity, and in DSD in fact the digital bits do correspond to levels of intensity

2) read again what you said about the hair cells in our ears (the cilia is only a part of that cell); each hair cell responds only to a certain small frequency range, and then responds to the intensity (volume) by releasing a specific number of neurotransmitter molecules, which is exactly what I said. First the auditory nerve and then the auditory processing centers of our brain use that neurotransmitter density (composed of individual molecules, remember) and "translates" that to what we perceive as sound. That is a D>A convertor; not necessarily binary digital, but nevertheless...

I understand that some people have an attachment to the idea of analog, but that's really all it is, an "idea" which doesn't accurately reflect reality when examined in fine enough detail. Other than 1-bit PCM, digital is never "yes or no", but rather is one of any "number" of specific values, and of the course the "number" is dependent on the specific digital format.

 

[Atmasphere]

For real?? It seems that you have something backwards. I take point 1) with no argument, save for the fact that you must not have read my prior post correctly. In short as you say, in digital, a word assigns intensity or amplitude but amplitude is not assigned to individual bits. Just as the ear does not behave in a digital fashion as you concede in point 2) above, neither does an analog system.

 

[rbbert]

To me, the ear clearly works in a digital fashion, so you and I (and neurophysiologists) must have somewhat different understandings of "digital" and "analog". A hair cell is "on" or "off", and "on" has specific integer values of "on", not an infinite or continuously variable amount of "on".

Likewise, the output of my DAC is an analog signal, listening to a PCM stream without a DAC sounds like modulated noise. Analog magnetic tape, vinyl LP and various digital audio formats are merely data storage media, waiting to be converted to an analog electrical signal and then an acoustic audio signal. That conversion process is different for each, and each has its faults and virtues, but in each case at some level there are individual "particles", whether it's a piece of iron oxide or a specific number. We don't listen to the individual particles, we listen to the analog signal generated from those particles.

 

[Atmasphere]

Here is what you said about the ear (entirely off-topic BTW):

That is a D>A convertor; not necessarily binary digital, but nevertheless...

Now since we only use binary digital, the ear is as you say above not the same process. I am very sure if the brain were opened up one would not see any memory chips either. We have a biological process that is not entirely understood. Because it is not well understood, to make up stories about how its really digital is not the best thing to do for one seeking credibility.

Per your second paragraph above I have almost complete agreement. However that paragraph does not state anything for or against your prior argument; its a red herring, so to speak.

 

[rbbert]

to make up stories about how its really digital is not the best thing to do for one seeking credibility.

Per your second paragraph above I have almost complete agreement. However that paragraph does not state anything for or against your prior argument; its a red herring, so to speak.

1) I'm not making up stories. Consult a neurophysiology textbook or a neurophysiologist if you don't believe me (and there's no particular reason you should believe me, unless you're not interested in doing your own research). As far as memory chips in our brains? Take another look in 10 years or so (maybe less).

2) That paragraph was/is my argument, and tomelex's (who made the post which led to this tangent)

 

[Atmasphere]

Uhm, the point is that the ear is not digital as we know digital. You will be contradicting yourself to deny this, per the quote I just put up.

2) is not the argument you were making. However, if that is now the point you are trying to make I'm not going to argue it.

 

[rbbert]

Technically, analog and digital are processess that convey information and actually are more the middle ground than one or the other:...

Yes, your LP is a digital process at the lowest levels, and so is air movement, etc....really its all quantization...there aint no smooth analog signals nowhere in audio....but just to remind that processes can be called analog and digital but when you get down to the details things go quantization route...ie tiny information "thingies", whether its oxide tape particles assuming a direction (digital like) or bumps on a LP at a particular spot (digital like)....aint it funny....just sayin.

I believe this represents the pertinent parts of the original post to which you responded?

As far as hearing goes, I don't really see a difference between a 24 bit digital "word" (for example) sampled x/sec and a group of hair cells sending "on" messages along the auditory nerve at x/sec, but if you do then there's not much more I can say.

 

[Atmasphere]

OK. How about let's not talk about the ear in this thread and we will be OK.

With regards to the tomelex post, it was not that one, but this one:

tape theory (in analog tape decks loved by many) states that the iron oxides or whatever, take on a magnetic alignement, and as such eash particle becomes orintated one way or the other, and that is a digital process my dear fellows.

 

[rbbert]

Ah. Well, I'm not sure I regard analog magnetic tape as a digital process per se, but I do think it can be accurately modeled by a digital process, which was how I read the post I quoted.

 

[opus111]

Analog magnetic tape, vinyl LP and various digital audio formats are merely data storage media, waiting to be converted to an analog electrical signal and then an acoustic audio signal.

There's a significant difference - its possible for mag tape and vinyl LP to be used to store data, but that's not how they are used in audio. If they were data stores then the data on them could be perfectly copied, which isn't the case - each copy incurs a generational loss which can never be zero. Digital and analog systems interpret the media differently.

 

[NorthStar]

-- BTW, tapes, have more to do with chemical reactions, over time.

...Magnetic heads and oxide formulations. ...I don't see the 'digital' equation here.

 

[rbbert]

There's a significant difference - its possible for mag tape and vinyl LP to be used to store data, but that's not how they are used in audio. If they were data stores then the data on them could be perfectly copied, which isn't the case - each copy incurs a generational loss which can never be zero. Digital and analog systems interpret the media differently.

The fact that they are imperfect does not mean they aren't data storage.

 

[opus111]

My argument is not about them being imperfect. What do you mean by 'imperfect' here?

 

[NorthStar]

-- One; that's a good post Tom. ...I understand what you're sayin', and I'm good with that.

Two; there is no two. :b ...Only zeros and ones.

________________

- Direction: both analog and digital have directions.

- Contact: laser (optical) for Discs, and 'rock' (diamond, ruby, sapphire, ...) for LPs (direct physical contact with the grooves).

- Noise: dithering noise, quantization noise, digital jitter. Wow 'nd flutter (analog tapes), mechanical noise (turntables and arms).

 

[NorthStar]

I'll just add that ambient info sounds more compelling if the master is originally analog.

---- Very interesting quote, from Roger. ...But, what is 'ambient' sound, or is it some type of distortion?

* BTW Roger is gone, wonder why ....

 

[Atmasphere]

OK, so I need to do some more explanation as not everybody is thinking about this at the level I described. So, lets take the groove in the LP, we all KNOW what that looks like…if you said at a microscopic level it looks like a bunch of grains of sand then now you are with me. Each grain of sand on that groove is now going to be in contact with the diamond stylus. The diamond is therefore smoother (compared to sand). But, as in digital, we now have these little bumps, ie peaks and valleys, ones and zeros my dear fellows on the side of the groove, and our smooth by comparison diamond stylus drags along those bumps and wow, if we had a very precise play back stylus then we would be recording NOT a smooth say sine wave, but a sinewave that had all kinds of little bumps and valleys in it.

My point was and is, that as you do down in level, you lose “analog” and get to the real world, where we are quantizing those bumps, we are doing a digital to analog conversion. While we play that groove, the stylus will bump up and down, at that level, in steps, for each grain of sand that protrudes up from the vinyl as say in a half a sphere. We don not get a smooth variation.

The same thing can be said by tape at that level, as tape, no matter what theory you ascribe too, at its lowest usable level is composed of particles that take on a north and south pole, ie they are each a little magnet, a little digital magnet, a grain of sand so to speak.

Lets say that, back to the LP, that there are 1000 sand unit variations (bumps) for one sinewave, then we have quanticized it AND dithered it too due to the randomness of those grains of sand.

The point is, that as you go down in level, and I hope my picture words above took you there, we see the world is more digital than we ever knew.

Analog and digital, as I said before are processes, one is smooth and continious…at the proper level of investigation as in with the naked eye…and the other level, it is a matter of quantized, ie digital stuff…with microscope…that is what is happening at the micro level on the LP groove wall.

That smooth continious sinewave, when viewed with more precision, becomes very unsmooth, and has random effects (dither) too!

That’s all I was sayin my dear fellows.

Sorry, but this analogy just does not hold up! The bumps you refer to are random noise. You can't count on them to be ones and zeros (lol! ) and they most certainly are not! In fact the closer you look under the 'scope, the more you will find that there are even smaller 'bumps' as the term 'fractal' is a much closer description of what is going on.

 

[mep]

I'm still trying to understand why some people feel compelled to come up with round peg/square hole analogies to try and convince people that analog is really digital. Maybe next we can start in on why transistors are really vacuum tubes because they both amplify signals and get that thread rolling.