Digital Audio Analog

DIGITAL AUDIO ON TWISTED PAIRS

If you are heading to install AES digital audio, you have a range of selections.  The 1st option is to decide how to ship this audio about your facility, embedded or separate audio and video?. 

Embedded audio is easy.  A single coax cable will carry the video and also be carrying the audio.  This simplifies the labor by 50 percent, given that you are only utilizing one connector (a BNC) to carry equally signals. 

 

The 2nd advantage is lip sync.  If the audio and video have been in sync when they have been set on the cable, they will stay in sync until they come off the cable.  And loudness is also a non-problem: if the loudness level of the audio was best when set on the cable with the video, it will stay at that level until you get the signal off.

 

In reality, the only true disadvantage to embedded audio is that it is embedded.  You cannot do anything to the audio devoid of "de-embedding" it 1st.  This needs an pricey box.  But, if you will need to modify the audio, lip sync, volume or subject material, they you will have to de-embed the audio to operate on it.  Of course, once it is de-embedded, then the audio runs on a separate cable. And then you have far more choices…

 

In reality you have five selections (or possibly even far more) at that stage.  Table one lists all the basic selections you have.

 

CABLE TYPE                             Consists OF                                   Regular

Analog audio twisted pairs         common analog audio cable.             None

Digital audio twisted pairs          110 Ω, reduced capacitance digital cable   AES3

Class cable                          one hundred Ω Class 5e, Class six etc.   None for audio

Coax cable                                 75 Ω video coax                                  AES3-id

Ethernet                                     four-pair                                                  EIA/TIA 568B.2

 

Table one

 

Let's seem at each of  the option in Table one.

 

ANALOG AUDIO TWSITED PAIRS

 

Why cannot you use your aged cable to carry this signal?  Allow me display you what takes place if you run digital audio on analog pairs.  Figure one display the effect at 15 metres.

  

 

 

 

Figure one

 

The best square is the original digital bit stream, in this circumstance a 48 kHz sampled bit stream (bandwidth six.144 MHz). The dotted line is the stop of 15 metres of digital cable (Belden 1696A). The other trace is an analog cable (Belden 8451).  Of course, these two signal are not the identical but they are equally possibly recoverable.

 

But, when you get to 30 metres, in Figure 2, issues don't seem so great.

 

 

 

 

 

 

 

Figure 2

 

You can see that the signal voltage, far more than +/- 2 volts in the original bit stream, is now down to 50 percent that (+/- one volt) on the analog cable.  But even even worse than that is the rise and fall time on the analog cable. 

 

 

The reason why this is essential is simply because that rise and fall is the AES clock.  The clock is  extracted from the signal and employed to decode the signal by itself. The problem is that, if you seem carefully at the rise time of the waveform on the analog cable, wherever is the transition in between zero and one?  It is really distinct on the original waveform, and rather great, on the digital cable.  But on the analog cable, the receiving chip might have a hard time utilizing this as the clock.  This is known as ‘jitter', anything that has an effect on clock timing.

 

Analog cable is not created for digital audio.  It's not the right impedance (110 Ω).  It's far more like 30 to 40 Ω.  This is an impedance mismatch.  Such a mismatch brings about reflection ("return loss") on the cable.  As Figure 2 shows, 50 percent of the signal is reflected back to the supply and lost at the vacation spot (load).

 

The analog cable is also not right capacitance.  Digital cable is 43 pF/m.  Analog cable is far more than one hundred pF/m.  That capacitance is why we have than long charge-up and discharge, complicated the clock recovery and adding jitter. So the conclusion of this is, don't use analog cable to run digital.  If you don't have any option, keep cables quick, much less than 15 metres.

 

DIGITAL AUDIO ON COAX

 

AES3-id is the common for working digital audio on coax.  Television broadcasters insisted on this alternative simply because they could use the identical cable to run audio and video.  A single cable, one connector, one strip instrument, one crimp instrument.  Sounds rather effortless.  And BNC's are a whole lot less complicated to install than XLR's. 

 

These days, broadcasters typically set AES3-id audio on miniature coaxes.  These can typically go farther than twisted pairs.  In reality, if you want to have long runs of digital audio, set it on coax.  It can double or even triple the distance achievable on twisted pairs.

 

Class CABLES

 

Running digital audio on Class 5e or six is fairly common these days.  I suppose it's simply because designers and installers recognize this cable is meant to carry info (Ethernet) so utilizing to carry digital audio is not such a big jump.

 

The only problem is that Class 5e and six cables come 4 pairs to a cable.  If you want 4 channels of digital audio, I suppose this would be great. (And, in digital audio, you can also run two channels per pairs, so this could also be 8 channels of digital.) But what if you only want one channel, or at least one at a time?  This is the reason for Belden 1353A, one-pair Class 5e, especially manufactured for non-info applications with a single bonded pair 24 AWG, unshielded.

 

 

 

 

 

ETHERNET

 

This last option is considerably unlikely, but there are a whole lot of audio tools manufacturers who are now presenting Ethernet as a way to transport audio.  They can run 32 channels or far more (depending on the sampling fee), and you can use common Ethernet® switches and other assist devices, even though the payload is audio.

 

Then, of course, you cable is Class 5e or Class six undertaking what it was created to do, carry Ethernet. So it's interesting that Class 5e or six could carry AES digital audio in its "native" format, and also when it runs as Ethernet.  Plenty of choices!

I'm confident I don't have to inform you that the phrase is heading digital.  And, in the audio-video room, the 1st to go digital was audio.  The Audio Engineering Society (AES) and the European Broadcast Union (EBU) labored on the 1st digital audio requirements back in the 1980's, and the 1st common, known as AES3, arrived out in 1985, with continuing enhancements until the present day.

 

Digital audio is a really diverse animal from analog audio. And, for the last twenty-five a long time, the forces of analog have been resisting the forces of digital audio.  Now there will possibly usually be those who will use analog, just like there are nonetheless men and women who experience horses. A horse could not be as fast as a automobile, but it is a whole lot far more maneuverable, its pollution is a whole lot far more eco-friendly, and there are nonetheless men and women generating saddles.  In the identical way, there will be analog microphones and mixers and tape machines manufactured for a long time. The last of the large 24-channel multichannel expert machines, the Studer A827, was manufactured in 2004. More compact machines are nonetheless getting manufactured.

 

When digital 1st appeared there have been many comments about the quality of the sound, specially in comparison to analog.  But, as digital has enhanced around the last decade or so, the analog crowd has grown smaller and smaller.  A whole lot of this is possibly due to the MP3 and download crowd, who accept marginal quality as long as they get the subject material they want.  Compared to that quality, true AES digital audio is substantially higher quality.

 

So what is AES3 digital audio?  It is a program that enables the user to sample the analog materials and turn it into digital audio (info).  Table one shows the most common sampling prices.

 

SAMPLING Fee

In which IS IT Utilized?

Genuine BANDWIDTH

44.one kHz

Client digital, S/PDIF

five.6448 MHz

48 kHz

Broadcast radio/television

six.144 MHz

96 kHz

Film sound, recording studios

12.288 MHz

192 kHz

Best stop, highest quality

24.576 MHz

 

Table one

 

You will note that Table one commences at 44.one kHz, the sampling fee for customer CD's, digital audio track on customer camcorders, and many other customer applications. The quality of this sampling fee is considered to be somewhat better than the efficiency of FM radio broadcasts.

 

 

 

The sampling fee is not the bandwidth of the signal working down the cable.  The bandwidth is established by a range of elements.  1st is the digital ‘word' dimension, the range of bits in each segment of info.  The greatest authorized in the AES spec is 32-bit words.  The 2nd element is that the words can carry one or two channels of audio for every bit-stream.  This is why, for instance, on your house hello-fi receiver, it has a single RCA connector labeled "digital audio".  Than single connector will carry two channels of audio.  The last element is that, for twisted pairs, the AES selected a program known as ‘bi-phase' which mean, in the twisted pair, it doesn't issue which wire goes into which pin as long as you have the right two wires heading into the right two pins.

 

What this all implies is that you have to get the sampling fee and multiply it by 32 (phrase dimension) and yet again by 2 (2-channels) and yet again by 2 (bi-phase).  Or you can just multiply the sampling fee by 128 (32x2x2).  That method will then inform you the bandwidth of the signal.  And, as you can see in Table one, the genuine signal bandwidth on a digital twisted pair, is way beyond the 20-20 kHz of analog audio.  In reality, it commences at practically six MHz. The AES committee quickly realized that this bandwidth meant that the wavelength of even a customer digital audio bitstream was quickly attained as can be noticed in Table 2.

 

SAMPLING Fee

BANDWIDTH

WAVELENGTH

QUARTER-WAVE

44.one kHz

five.6448 MHz

53m

13m

48 kHz

six.144 MHz

49m

12m

96 kHz

12.288 MHz

24m

6m

192 kHz

24.576 MHz

12m

3m

 

Table 2

 

The committee realized that this mean they had to pick a particular impedance for these twisted pairs.  The excellent impedance, one with the lowest loss, would be 150Ω.  But this would make a large cable, one that would never fit into an XLR (the preferred connector).  So the AES committee lowered the successful distance and transformed the impedance to get the dimension down to a thing much closer to the dimension designers and installers have been employed to in the aged analog globe.  And this impedance turned out to be 110Ω.

 

AES3 also spells out how far one can go on a offered twisted pair.  If you feed an AES digital signal of 2 volts in, and a .2 volt (200 mV) signal out, that's as far as you can go.  This distance is also impacted by the resistance (i.e. dimension) of the wires, so diverse gage wires will go diverse distances.  But given that we're nicely inside of the quarter wavelength distance, the supply, the cable, and the vacation spot (load) impedance need to all be 110Ω.  This is now a "transmission line" wherever we need to match impedance.  Table three shows these distances.

GAGE

six MHz

12MHz

25MHz

26 AWG

248m

193m

145m

24 AWG

337m

267m

198m

22 AWG

469m

381m

309m

 

Table three

Table three might surprise a whole lot of men and women who have been usually informed "digital audio signals cannot go really far".  In some circumstances (if you go through the previous installment on analog audio), these digital signals can go farther than analog audio signals!

 

You have a range of selections to run these digital signals.  Of course, there are AES twisted pairs.  But there are at least three variations on that theme aloe to pick from.  Then you have another common AES3-id, that enables you to run these digital audio signals down coax cable.  We'll pay a visit to all of these….and more…in our next installment!