Neumann.Berlin Digital Microphones For High Resolution Audio Manuel D’Utilisation

Page de 6
SCHNEIDER 
DIGITAL MICROPHONES FOR HIGH RESOLUTION AUDIO  
AES 31st International Conference, London, UK, 2007 June 25–27 
 
2
principle to P48 defined in IEC61938 [6], but adapted to 
the lower voltage and higher current requirements of 
ADC components. It already featured a gain ranging 
ADC, to be discussed later, and limited remote control 
functions (pre-attenuation) but yielded sub-optimal 
noise figures, compared to standard analogue micro-
phones. Another proprietary solution was presented by 
Milab [7]. 
Although the mentioned developments could not fully 
compete technically with state-of-the-art analogue 
microphones, they were helpful in starting discussions 
amongst manufacturers on the future of digitisation in 
microphones. It was found that, before presenting 
microphones with digital output to a wide public, all 
questions of power supply, interfacing, connector types, 
remote control etc. should be put into a public standard, 
to allow future products to interconnect between 
manufacturers. Accordingly, the German DKE 742.6 
committee served as a starting basis, then handing over 
to an AES standardization committee to publish the 
AES 42-2001 standard [8,9], currently revised to the 
2006 edition. Almost ten international microphone 
manufacturers were actively or passively involved, 
guaranteeing a common consensus. First microphones 
complying with the new standard were presented in 
2001, as a full-feature large diaphragm microphone 
[10], later followed by a measurement microphone [11] 
and small diaphragm capsule systems [12,13]. 
In contrast to the professional audio approach, trying to 
provide highest possible audio quality, recently other 
solutions have been presented, driven by computer 
technology, i.e. mainly USB-powered microphones, 
with currently in comparison very limited specification 
ranges [14]. 
2  REASONS AND REQUIREMENTS FOR 
DIGITAL MICROPHONES 
Analogue output condenser microphones are now, 90 
years after their invention by E.C. Wente, certainly a 
mature technology. In a professional set-up, with 
appropriate cabling and limited outside interferences, a 
very high dynamic range of up to 130 dB-A can be 
transduced [15,16]. To reduce effects of cable length 
and interferences on the comparatively small 
microphone output signal, preamplifiers are often 
located in close proximity to the microphones. In any 
case, proper level matching of all analogue components 
is necessary to guarantee optimal signal transmission, 
allowing for sufficient head-room and foot-room in the 
process. On the other hand, digital technology provides 
potentially loss-less transmission, once the analogue-to-
digital conversion has taken place. Accordingly, the 
interest for microphones with digital output arose when 
high quality ADC technology became available, 
allowing conversion only minimally affecting micro-
phone specifications. 
Some of the requirements on digital microphones [8] 
later realized in the AES 42 standard [9] were 
physical layer interface & protocol 
compatibility: AES3 protocol with overlaid 
phantom power, using 3-pin XLR connectors, 
control information from the microphone: via 
user bits in the AES3 data stream, 
control information to the microphone: via low 
frequent modulation of the phantom power 
voltage. 
With the chosen interface, loss-less transmission can be 
performed over approximately 100 m also with high-
quality “analogue” microphone cable, approximately 
300 m with AES3 “digital” cable. This compares well 
with typical values for high-quality analogue set-ups. 
An essential point in digital technology is proper 
synchronization of all audio streams to a reference 
clock. In a minimal set-up a receiver can synchronize to 
a single microphone, although this would be in contrast 
to typical studio procedures, where either the mixing 
console, or a dedicated reference clock provide the 
clocking reference. But, with multiple digital micro-
phones one needs to either work with sample rate 
converters in every channel at the receiver side (AES42 
mode1), or preferably synchronize the microphones to 
the reference clock (AES42 mode2). High quality 
sample rate converters do increase the cost, and even 
though in their current embodiments [17,18] they might 
not influence the signal much, they will increase 
processing time and thus add to the overall latency, 
which can become prohibitive in some applications, e.g. 
where direct monitoring is called for. 
Sending the clock signal directly to the microphone 
would imply multi-lead cables, incompatible with 
standard 2-wire+ground/return studio wiring. The 
solution adopted by AES42, after extensive tests, was to 
integrate a voltage controlled crystal oscillator (VCXO) 
inside the microphone, yielding an already very stable 
data stream but where the frequency is dynamically fine 
tuned from the receiver side via the control information 
sent to the microphone (Fig. 1). 
 
Figure 1: Connection of a digital microphone, with 
synchronization using AES42 interface specification. 
Microphone sample rate is controlled (CTL), comparing 
extracted microphone rate and external word clock.