radiolistener:
This is complete bullshit, contrary to actual usage, supported by a comically failed attempt in appeal to authority. You could at least read the backlog, where people working in the industry for decades share their experience or give actual examples.
"This is complete bullshit" - this is what you will hear when you try to provide some signal which has 4 levels on digital GPIO pin of some controller and ask firmware developer to handle all 4 levels because this is "digital" signal... And I will agree with him, because this is really complete bullshit.
If you're talking that the signal is digital, it should have two possible state 1 or 0.
And when you're talking that analog discrete signal with possible 4 levels is "digital", this is complete bullshit. Because there is no digital logic circuit that supports signal with 4 possible states.
You can name digital signal as analog and analog as digital for yourself, but if you're want so other people can understand you, you should use term digital for signals with 2 possible levels 1 and 0. And use analog term for signals which can have more than 2 possible levels.
This is an invalid argument[/b], but since you considered it ok and it may be worth waking you up from your denial of reality, here is what your own sources say about word “analog”
I don't needs to walk. Analog signal means that signal IS NOT DIGITAL. So, any signal which has more than 2 possible levels is analog. This is pretty easy...
In order to process analog signal in digital domain you're needs to convert it from analog domain to digital domain. And in order to generate analog signal from digital domain you're needs to convert from digital domain to analog domain. There is no way to process analog signal in digital domain without converting. And you're needs to learn the difference between analog domain and digital domain to avoid confusing them.
This is very simple:
- Digital signal has two possible states: 1 and 0.
- Analog signal has more than two possible states.
Grouping of several signals don't change their type. If you have 8 lines and all are digital, then this is digital bus, not analog. The same if you use 8 lines and all are analog, then this is analog bus, not digital.
You are dead wrong, sorry.
As others have said: binary is a subset of digital.
Analog means that a continuous value changes in proportion to the information. For example, amplitude of a sound waveform: this modulates the amplitude of the carrier wave in AM radio, or the carrier frequency in FM.
In contrast, in a digital system, the values are discrete (they can only take one of a fixed set of values), and those values need not correlate directly to the information being encoded. In digital audio, even if uncompressed, the individual digits do not correlate to a particular amplitude. In typical PCM encoding, the entire word needs to be a) identified, and b) interpreted, in order to derive an amplitude value.
It’s true that in electronic signaling, most digital signals and circuits are binary. But not all are, and telecommunication signals frequently use non-binary representations of digital data. The fact that they encode informatikn using multiple levels (and in many, phase shifts) doesn’t make it analog, because the encoding doesn’t change in proportion to the information itself. Other than memory, I’m not aware of any digital logic that uses multilevel signaling, so in the context of practical digital circuit design
other than telecommunications, yes, digital means binary. But not because they’re synonyms, but simply because “digital excluding telecom” is a set containing only one member in practice.
All digital signals live in the analog domain, insofar as the physics of the real world are analog. So what we really mean when we say that a signal is digital (whether binary or otherwise) is that we are interpreting it in a digital way. This layer of abstraction is one thing that makes digital so useful, because noise can be ignored to an extent, since the physical signal is not an analog of the information.
And let’s not forget that there are quasi-analog signals that take only two states, like pulse width modulation and pulse density modulation. And indeed, with simple analog filtering, true analog signals can be extracted from them. (You can’t take a PCM bitstream and analog filter it to get audio.)