Why digital cables are not used instead of analog cables?

[LOST TALE]

I dont get it.

[Sarcasm]

Because drivers operate on an analog sine wave.

[Aphexx]

then you'd have to cram a dac into the headphone enclosure increasing weight and restricting the space used to tune the sound

 

kind of like asking: why not put the graphics card directly into the monitor 

[MG2R]

What you need to understand is that speakers operate in an analog way, just as our ears operate in an analog way. This is because audio is a natural occurence and nothing more than differences in air pressure that follow eachother up rapidly (the more rapid, the higher the sound's pitch or frequency) and therefore there are an infinite number of possible pressure "levels".

 

A computer on the other hand, operates in a digital way. It only knows two states: either on (1) or off (0). This is its only way of representing data. This means that your song is saved as a stream of bits (with a certain bitrate, the number of bits per second).

 

When you try to get that analog speaker playing the digitally stored music, you need something that converts that digital stream to an analog signal. The device that does this is called a digital-to-analog converter or DAC. Basically this is the soundcard in your PC. That DAC outputs a small analog signal that needs the be amplified (by an amplifier) before it is powerful enough to actually get some movement (sound) out of speakers. This gives a complete audio output process that looks like:

 

audio bitstream -> DAC -> amplifier -> speaker

 

Now, you can either put the DAC and amp outside of the PC. This is basically an external audio card + amp. If you go with this approach, you connect the DAC to your computer by means of a digital signal (USB/optical/...).

Another option is to put the DAC inside the computer (internal soundcard) and the amp outside of it. Now, you have an analog signal that goes from your PC to the amp, thus you need an "analog cable".

The third option is to put both the DAC and the amp inside the computer. You often see this in the form of a soundcard with headphone amplifier.

 

The main reason (I think) that the DAC is put inside the PC is historic. A radio is analog, so people are used to having an amplifier that powers their speakers. When digital audio devices came to market, they had to be able to connect to those amplifiers, thus they had to build in a DAC.

 

Hope this clarifies your question and Aphexx's (correct) anwer

 

PS: this is mainly just semantics, but there is no such thing as a digital cable or an analog cable. All copper cables are capable of conducting either a digital or an analog signal.

[ShearMe]

The main reason (I think) that the DAC is put inside the PC is historic. A radio is analog, so people are used to having an amplifier that powers their speakers. When digital audio devices came to market, they had to be able to connect to those amplifiers, thus they had to build in a DAC.

 

The reason they put audio components in the PC is because the whole point of a PC is to have one self-contained computer that has everything you need inside. If they put GPUs in an case exterior from the PC, people would never buy it because that's another device to hassle with. Back in the day, on board audio didn't even exist.

 

Here's another cool history lesson.

[MG2R]

The reason they put audio components in the PC is because the whole point of a PC is to have one self-contained computer that has everything you need inside. If they put GPUs in an case exterior from the PC, people would never buy it because that's another device to hassle with. Back in the day, on board audio didn't even exist.

 

Here's another cool history lesson.

Well, yes, it's only logical to put it into the same machine. But they could just as easily put the DAC inside the amplifier. The reason they don't is (I think) the historic reason mentioned above.

 

The link doesn't seem to work. Nevermind, you fixed it already. Let's see

[ShearMe]

Well, yes, it's only logical to put it into the same machine. But they could just as easily put the DAC inside the amplifier. The reason they don't is (I think) the historic reason mentioned above.

 

Oh, I had a hard time understanding what you were saying, but after re-reading it I agree!   I must be tired or something.  :wacko:

[Piraetsu]

The reason they put audio components in the PC is because the whole point of a PC is to have one self-contained computer that has everything you need inside. If they put GPUs in an case exterior from the PC, people would never buy it because that's another device to hassle with. Back in the day, on board audio didn't even exist.

 

Here's another cool history lesson.

Nope. The point to have a PC is to have flexibility in terms of have everithing that you might want and need, independently of its an internal or an external component

[ShearMe]

Nope. The point to have a PC is to have flexibility in terms of have everithing that you might want and need, independently of its an internal or an external component

 

Well if you had a bunch of external components (thinking, like, back in the 80s) it wouldn't been considered much of a Personal Computer - more like a mainframe. 

[Piraetsu]

Well if you had a bunch of external components (thinking, like, back in the 80s) it wouldn't been considered much of a Personal Computer - more like a mainframe. 

One thing its a Mainframe, and other completely different is a PC with a few peripherals attached to it. Obviously you dont know that an external audio card is better than an internal one...

[Aphexx]

HOLY DE-fucking-RAIL batman

 

I only mentioned the gpu in a monitor because I had an image of an awkward block with fans attached to a slim framed light and elegant thing like my lcd monitor. 

[xByakko]

One thing its a Mainframe, and other completely different is a PC with a few peripherals attached to it. Obviously you dont know that an external audio card is better than an internal one...

Excuse me for my rudeness but your ignorance (excuse me again if I'm using the wrong word ) is hilarious

But on topic: Sound is analogue (so it can be represented "in a way" with any real number), where as the the sound information coming from your computer is binary (represented only by 0s and 1s). My guess is that you got this question from the argument with how the analogue VGA connector is being replaced by digital interfaces such as DVI, HDMI, and display port. The only reason it works for monitors is because that you only need to tell the lights in your monitor to go on(represented by 1) or off(represented by 0) (of course you can represent it in the other way around, but the point is that there is only need for yes or no for lights). However sound is not a matter of on or off, thus you can't use a digital signal which only says 1 or 0 (or alternatively yes or no).

[ShearMe]

One thing its a Mainframe, and other completely different is a PC with a few peripherals attached to it. 

 

External GPUs and sound cards aren't peripherals.

 

 

Obviously you dont know that an external audio card is better than an internal one...

 

Obviously you're new here.

[mr moose]

 

Obviously you're new here.

Prove to me that internal sound cards are better than external ones, go on I dare you.  And don't go using fancy language neither, I don't want to hear about impedance matching and other fallacies like that, just down to earth reasoning why. :lol: :lol:

 

Psst, I thought anything that plugs into a computer was a peripheral, meaning on the side, related to but not the same, or constituting a boundary or outer limit with direct contact.

[ShearMe]

Psst, I thought anything that plugs into a computer was a peripheral, meaning on the side, related to but not the same, or constituting a boundary or outer limit with direct contact.

 

I guess it depends on definition, but I've always treated peripherals as non-essential devices. You can run a computer without a soundcard, keyboard, mouse, or monitor, but I doubt any modern GUI is going to load up without a GPU.

 

[MG2R]

It doesn't really matter as the OP isn't even monitoring this thread. Or are you, @LOST TALE ?

[mr moose]

I guess it depends on definition, but I've always treated peripherals as non-essential devices. You can run a computer without a soundcard, keyboard, mouse, or monitor, but I doubt any modern GUI is going to load up without a GPU.

 

When threads go like this I always feel like that bloke on the hand trolly.  No matter, at least this thread is different.  I was getting tired of the same old same old.

[LOST TALE]

Excuse me for my rudeness but your ignorance (excuse me again if I'm using the wrong word ) is hilarious

But on topic: Sound is analogue (so it can be represented "in a way" with any real number), where as the the sound information coming from your computer is binary (represented only by 0s and 1s). My guess is that you got this question from the argument with how the analogue VGA connector is being replaced by digital interfaces such as DVI, HDMI, and display port. The only reason it works for monitors is because that you only need to tell the lights in your monitor to go on(represented by 1) or off(represented by 0) (of course you can represent it in the other way around, but the point is that there is only need for yes or no for lights). However sound is not a matter of on or off, thus you can't use a digital signal which only says 1 or 0 (or alternatively yes or no).

digital signal sent to your screen is more then yes or no per pixel. there are 16.5M different collors (255x255x255) I think.

and yes thats what made me question this. ''how the analogue VGA connector is being replaced by digital interfaces such as DVI, HDMI, and display port''

 

why cant sounds be made with digital? thats what they are in the begining anyway.

the reaosn I stopped monitoring is because someone answered that the sound output has to come form an analog signal and the converters would be very big and it would come to the same thing in the end.

[LOST TALE]

Hope this clarifies your question and Aphexx's (correct) anwer

 

Thank you!

[mr moose]

 

why cant sounds be made with digital? thats what they are in the begining anyway.

 

 

Essentially that's exactly what happens.  When a device (dac) converts digital to analogue. All it does it add together the bits and presents the sum as an analogue signal. So the the more 1's you have in a digital byte the higher the signal voltage and the less 1's the lower the signal voltage.  The way binary works is that when you convert analogue (from a mic) to digital each byte (or series of bytes) of digital information has a direct ratio of 1's to the original signal shape.  If this makes sense.

[ShearMe]

USB headsets are digital. Problem is you end up paying a premium on the electronics in the headphone and get a lower cost headphone driver.

[xByakko]

digital signal sent to your screen is more then yes or no per pixel. there are 16.5M different collors (255x255x255) I think.

and yes thats what made me question this. ''how the analogue VGA connector is being replaced by digital interfaces such as DVI, HDMI, and display port''

 

why cant sounds be made with digital? thats what they are in the begining anyway.

the reaosn I stopped monitoring is because someone answered that the sound output has to come form an analog signal and the converters would be very big and it would come to the same thing in the end.

of course the signal sent to the screen is more than yes or no per pixel, hence i said "light" not "pixel". essentially what i meant is, information sent to the monitor is in binary where it tells the red,blue,green lights(i might be inaccurate in how back lighting works but should be close.) to turn on or not, and that the combination of R.G.B makes up the "colour" (dont question my spelling, i'm in british colony), however the fact remains that it's a matter of switching lights (the different coloured lights) on and off.

 

whereas sound is a wave, which can be represented by "infinite" decimal points (if you like) of numbers (eg. 3.141592653...) at a certain point of the wave, which is very hard to represent in binary (using only 1s and 0s, the only form of information a computer can "use"). thus when sound is converted into digital information which the computer uses, it's only and can only be an approximation of what the original sound is (which is your bit rate or sampling rate i believe) using multiple digits of 1s and 0s to represent different values at different points of the wave. this means if you leave it in digital form, the sound is only on or off, instead of high or low. thus when your computer outputs sound, it needs a DAC to sum up the 1s and 0s and approximate what the original wave is. (im not very good at explaining, and this is only what i learnt from a first year university paper).

[mr moose]

of course the signal sent to the screen is more than yes or no per pixel, hence i said "light" not "pixel". essentially what i meant is, information sent to the monitor is in binary where it tells the red,blue,green lights(i might be inaccurate in how back lighting works but should be close.) to turn on or not, and that the combination of R.G.B makes up the "colour" (dont question my spelling, i'm in british colony), however the fact remains that it's a matter of switching lights (the different coloured lights) on and off.

 

whereas sound is a wave, which can be represented by "infinite" decimal points (if you like) of numbers (eg. 3.141592653...) at a certain point of the wave, which is very hard to represent in binary (using only 1s and 0s, the only form of information a computer can "use"). thus when sound is converted into digital information which the computer uses, it's only and can only be an approximation of what the original sound is (which is your bit rate or sampling rate i believe) using multiple digits of 1s and 0s to represent different values at different points of the wave. this means if you leave it in digital form, the sound is only on or off, instead of high or low. thus when your computer outputs sound, it needs a DAC to sum up the 1s and 0s and approximate what the original wave is. (im not very good at explaining, and this is only what i learnt from a first year university paper).

 

It's not an approximation, It's a very accurate sample. The signal is converted to a series of numbers (stored in binary) that each represent how high or low the voltage is for any one given point. The number of samples per second and the bit depth dictate the accuracy, currently we have 24bit (which gives us an extreme amount of headroom and a very low noise floor) and sample rates at 96Khz, given that we can't here frequencies above 20Khzs this means that the digital signal In 1's and 0's is far superior to any analogue signal we can hear. Thus no approximation, only specifics.

[xByakko]

It's not an approximation, It's a very accurate sample. The signal is converted to a series of numbers (stored in binary) that each represent how high or low the voltage is for any one given point. The number of samples per second and the bit depth dictate the accuracy, currently we have 24bit (which gives us an extreme amount of headroom and a very low noise floor) and sample rates at 96Khz, given that we can't here frequencies above 20Khzs this means that the digital signal In 1's and 0's is far superior to any analogue signal we can hear. Thus no approximation, only specifics.

Yes, approximation is the wrong word to use, my bad. And yes, it's possible to be accurate to the point where we won't really notice any difference, however it's still only sampling, not the exact curve. so if we really wanted to get really specific it's still not the original graph because you are essentially getting points of the graph and if you "visualize it", it's just connecting the dots with lines, rather that the exact curve.

an image to illustrate:

For those who don't know, the "smooth" wave is the sound, and the jagged "graph" is the sampling. Of course, in actual sound application, the sampling rate is much higher and we would need to zoom in the see the jagged edges, but the fact remains that it will never be "exact", but it's enough for us humans.

When I was thinking of digital information, I was think of the raw sequential input of (010111000101001, etc) rather that the "chunk" which for example 8 bit integer numbers 10(00001010) or 42(00101010).

However this is getting off topic I think, the answer remains that we need to use analogue cables to connect to a "sound source" (speaker, headphone) because we can't directly use digital information to recreate sound without an actual DAC process.

[ShearMe]

Yes, approximation is the wrong word to use, my bad. And yes, it's possible to be accurate to the point where we won't really notice any difference, however it's still only sampling, not the exact curve. so if we really wanted to get really specific it's still not the original graph because you are essentially getting points of the graph and if you "visualize it", it's just connecting the dots with lines, rather that the exact curve.

an image to illustrate:

For those who don't know, the "smooth" wave is the sound, and the jagged "graph" is the sampling. Of course, in actual sound application, the sampling rate is much higher and we would need to zoom in the see the jagged edges, but the fact remains that it will never be "exact", but it's enough for us humans.

When I was thinking of digital information, I was think of the raw sequential input of (010111000101001, etc) rather that the "chunk" which for example 8 bit integer numbers 10(00001010) or 42(00101010).

However this is getting off topic I think, the answer remains that we need to use analogue cables to connect to a "sound source" (speaker, headphone) because we can't directly use digital information to recreate sound without an actual DAC process.

 

Erm, i think you should watch this: 

 

http://www.youtube.com/watch?v=cIQ9IXSUzuM