NeumannBerlin Microphone Digital Microphones For High Resolution Audio User Manual

SCHNEIDER  
DIGITAL MICROPHONES FOR HIGH RESOLUTION AUDIO  
DIGITAL MICROPHONES FOR HIGH RESOLUTION AUDIO  
MARTIN SCHNEIDER1  
1 Georg Neumann GmbH, Berlin, Germany  
Microphones with digital output format have appeared on the market in the last few years. They integrate the functions  
of microphone, preamplifier, and analogue-to-digital converter in one device. Properly designed, the microphone  
dynamic range can thus be optimally adapted to the intended application. The need to adjust gain settings and trim  
levels is reduced to a minimum. Dynamic range issues inside and outside the microphone are discussed. Advantages of  
digital microphones complying with AES 42, with a wide dynamic range and 24-bit resolution are shown.  
INTRODUCTION  
Current microphone technology thus focuses on the  
third category: microphones with integrated ADC. Here,  
a purist could further differentiate between  
One should first define the term “digital microphone” in  
the context of this article. A possible classification  
could comprehend:  
-
-
microphones with ADC output modules,  
microphones with ADC in closest proximity to  
the transducer,  
-
a transducer where the underlying acoustical-  
mechanical-eletrical transduction principle  
contains a quantization,  
where the first subcategory would describe a complete  
microphone, just with an added ADC module; the  
second subcategory represents transducers where the  
transducing element itself is closely integrated with the  
analogue-to-digital conversion process. In the context of  
high resolution audio it will be clear that the preferred  
transducer should be of the electrostatic (condenser)  
type, as this principle still yields the highest  
performance regarding parameters like linearity,  
dynamic range and frequency range.  
-
-
a combination of separate transducers, each  
responsible for certain quantization steps,  
a microphone integrating an analog-to-digital  
converter (ADC).  
The first category describes the “purely digital”  
transducer. The first microphone by Philipp Reis [1], a  
single contact transducer, represented such a transducer,  
albeit with very low quality due to the 1-bit resolution.  
This is the only purely digital transducer known to the  
author.  
In the second category we find e.g. an optical  
microphone, where the position-dependant displacement  
of a diaphragm is traced with distinct light rays. The  
reflected rays excite separate sensors, whose outputs are  
combined into a single signal [2]. Another, electrostatic  
transducer experiment shows the diaphragm as part of  
the ADC, as component for the electrical / acoustical  
summation in the feedback loop of a Σ∆-converter [3].  
To obtain dynamic ranges comparable to the 120-  
130 dB of standard analogue microphones, these  
principles would need to be scaleable over 6 orders of  
magnitude, a feat hardly achievable due to the extreme  
mechanical precision involved.  
1
HISTORICAL DEVELOPMENT  
Possibly the first realization, in 1989, incorporating an  
ADC in the same housing with an electro-acoustical  
transducer is mentioned in [4]. The corresponding  
electret condenser microphone by Ariel company was  
intended for use with the now defunct NeXT computer,  
with the then available 16 bit transducers and a stated  
dynamic range of 92 dB. A 1995 prototype by Konrath  
[5] put an ADC circuit inside the housing of a  
commercial microphone. It featured a 7-pin XLR-  
connector and dedicated supply, delivering a multitude  
of supply voltages to the circuit. A later commercialised  
version by Beyerdynamic (MCD100) simplified this set-  
up with the adoption of phantom power, similar in  
AES 31st International Conference, London, UK, 2007 June 25–27  
1
 
SCHNEIDER  
DIGITAL MICROPHONES FOR HIGH RESOLUTION AUDIO  
The essential requirement then for digital microphones  
remains to integrate A-to-D conversion providing  
dynamic range and resolution comparable to their high  
quality analogue counterparts.  
settings npre might be as high as -100…-80 dBu, and is  
seldom published in the specifications. One sees that  
preamplifier noise is higher or lower than the above  
mentioned microphone self noise of -120 dBu-A, and  
one main task for the recording engineer is then to  
optimise this sum, keeping preamplifier and ADC input  
headroom in mind. In analogue set-ups, the rule is to  
pull up the gain to studio reference level, trying to avoid  
clipping or distortion even with unforeseen very high  
sound pressure levels.  
The working dynamic range of a typical microphone /  
preamplifier combination is shown in Fig. 3. The output  
level of the preamplifier Uout,pre is shown over gain v.  
ADC noise is left out, for simplification, and assuming  
that the preamplifier gain will be optimally set, so that  
microphone and preamplifier noise dominate. The  
limitations are then given by:  
3
DYNAMIC RANGE AND NOISE  
In order to be able to compare possible benefits of  
analogue and digital microphones, one has to look at the  
limiting factors, i.e. the behaviour at very small and  
large signal levels, corresponding to the noise floor and  
the overload characteristics, as well as the typical signal  
resolution, with a medium level signal present.  
As mentioned, the typical dynamic range of the output  
of a condenser microphone capsule can exceed 130 dB,  
with typical maximum levels at a surprisingly high  
+10 dBu (2.5 VRMS) and microphone self noise at  
-120 dBu (A-weighted). In the most noise free of  
current studio microphones this corresponds to sound  
pressure levels of 7 to 137 dB SPL, covering the needs  
of most applications. Only in excessively loud settings  
will there be a need to (manually) switch the pre-  
attenuation on, shifting the microphone’s dynamic range  
to higher levels.  
o
o
o
n200: -131.7 dBu-A thermal resistive noise as  
physical limitation,  
npre: preamplifier equivalent input noise (A-  
weighted),  
Maxpre: maximum preamplifier output level,  
here: +20 dBu  
o
o
nmic: microphone self noise, here: -120 dBu-A  
Maxmic: maximum microphone output level,  
A
D
here: +6 dBu  
One sees that the preamplifier noise npre reduces the  
maximum dynamic range of the microphone Dyn(Mic)  
by approx. 16 dB, to a maximum resultant working  
dynamic range Dyn(Max) of 110 dB. At the upper/right  
axis the diagonal curves of constant equivalent input  
sound pressure level are given values, for a microphone  
with sensitivity M0 = 12mV/Pa.  
Microphone  
Capsule Impedance  
Converter  
Preamplifier  
ADC  
Output  
Stage  
Figure 2: Simple analogue signal chain, with condenser  
microphone.  
The typical noise voltage nmic of a condenser micro-  
phone in Fig. 2 roughly follows a pink noise charact-  
eristic, whereas dynamic microphones, preamplifiers  
and AD converter inputs produce basically white noise.  
Preamplifier equivalent input noise npre (EIN) depends  
on the amount of gain v chosen. Concentrating all  
necessary gain inside the preamplifier, the sum of  
analogue equivalent input noise in an analogue  
recording chain with ADC will be  
nsum,ana = nm2ic + n(v)2pre + nA2DIn v2  
Figure 3: Dynamic range of a combination analogue  
microphone / preamplifier  
.
(1)  
The physical limit for preamplifier noise is determined  
by the thermal noise of the input load Ri  
The situation is different in the case of digital  
microphones with integrated ADC, as in Fig. 4. The  
capsule parameters can be chosen by the designer so  
that the capsule output levels are perfectly matched to  
the ADC input requirements. The noise sum then  
reduces to  
npre,min = 4kTRif  
(2)  
with k = 1,38*10-23 J/K (Boltzmann constant), T as  
temperature, and f as the bandwidth. For a typical  
microphone output impedance of Ri = 200 , nR  
calculates to -129 dBu (f = 23 kHz), or -131.7 dBu-A.  
At high gain settings many preamplifiers show noise  
figures close to this physical limit, but at low gain  
nsum,dig = nm2ic + nA2DIn  
.
(3)  
Accordingly, the curve for preamplifier noise in Fig. 3  
is replaced by the ADC noise nADIn. The noise over gain  
AES 31st International Conference, London, UK, 2007 June 25–27  
3
 
SCHNEIDER  
DIGITAL MICROPHONES FOR HIGH RESOLUTION AUDIO  
diagram for digital microphones is shown in Fig. 5. The  
dynamic range is vastly increased, especially for the  
small gain values often used with condenser  
microphones, and most importantly becomes  
independent of the chosen gain setting. It is now only  
limited by the microphone specifications, and by the  
digital processing limits, i.e. 0 dBFS level.  
noise nADIn approx. 7 dB higher, yielding -133 dBFS-A.  
Adding a real condenser capsule, the thermal/acoustical  
capsule noise ncaps adds another 3 dB (-130 dB-A). This  
means that the thermal/acoustical noise ncaps of the  
capsule and the electrical noise nADIn of the combined  
impedance converter and ADC are roughly at the same  
level. To achieve even lower values, one would thus  
have to work on optimising both electronics and  
capsule.  
Note: The gain shown in Fig. 5 is performed after the  
ADC, i.e. in the digital domain.  
As a side effect, the benign noise of the analogue  
components, capsule and impedance converter, with its  
largely gaussian distribution serves as an efficient dither  
on the ADC quantization noise [19]. With typical  
capsule parameters of small and large diameter cond-  
enser capsules, the summed noise nsum,dig can be at a  
level of -122 dBFS or -130 dBFS (A-weighted),  
respectively.  
A
D
Microphone  
Capsule Impedance  
Converter  
ADC  
Figure 4: Condenser microphone, with integrated ADC  
4
ADC CHARACTERISTICS  
Fig. 6 shows an ADC with dynamic range of 140 dB-A.  
ADC circuits matching such a vast dynamic range  
would be of the gain ranging type, combining two or  
more ADCs working at different signal levels. This is  
one realization of a floating point converter, with  
exponents of 20 and 24 [20,21].  
Figure 5: Dynamic range of a digital microphone  
Figure 7: Simple gain ranging ADC circuit [10]  
400mV  
Critical switching  
Path-1 clipped  
0V  
Path-1  
Path-2  
Very precise signal matching required  
to avoid glitches and amplitude errors  
-400mV  
400mV  
V
Figure 6: Noise spectra (16k samples, 32x averages) of  
an ADC with a. input short-circuited (-140 dBFS-A,  
lower curve), b. impedance converter and equivalent  
capsule capacitance (-133 dBFS-A, middle curve),  
c. impedance converter and real capsule  
0V  
-400mV  
0s  
0.5ms  
Path-2)  
1.0ms  
1.5ms  
Time  
2.0ms  
2.5ms  
3.0ms  
V
-
(-130 dBFS-A, upper curve).  
Figure 8: Signals in combined ADCs of Fig. 7, with  
audible “glitches” in the summed signal [10]  
A more detailed perspective of the noise components is  
presented in the spectra of Fig. 6. With the input short-  
circuited, the ADC shows a roughly white noise  
characteristic nADC, typical of today’s Σ∆ –ADCs, with  
slightly increasing noise above 20kHz, due to noise  
shaping algorithms. Reduced to a single value, the  
shown noise is in the region of -140 dBFS-A. The  
analogue impedance converter, loaded by a typical  
equivalent capsule capacitance, overlays this with a  
As is well known, switching directly between ADCs  
working at different levels can lead to artefacts like  
“glitches” (see Fig. 8), or noise modulation [20,21],  
when signal levels pass the switching level. The noise  
floor of an ADC is typically wide-band white noise.  
This white noise then becomes most audible when it is  
modulated by a low frequent signal, not masking the  
AES 31st International Conference, London, UK, 2007 June 25–27  
4
 
SCHNEIDER  
DIGITAL MICROPHONES FOR HIGH RESOLUTION AUDIO  
higher frequent white noise components. One possible  
way to reduce this effect is a non-linear network,  
keeping both ADCs always in operation, and summed,  
depending on the signal level, as shown in Fig. 9 & 10.  
unavoidable physics, and the maximum allowed sound  
pressure levels cover the vast majority of applications.  
For very loud signals, the dynamic range of the capsule  
output and thus of the complete digital microphone can  
be shifted by e.g. 6, 12, or 18 dB with the same  
mechanisms as in analogue microphones (shunt  
capacitance, negative feedback, or reduced polarization  
voltage). For safety purposes, an additional very fast  
look-ahead peak limiter (see Fig. 11) implemented  
inside the microphone takes care of unforeseen  
excessive sound pressure levels.  
Figure 9: Gain ranging ADC circuit, with non-linear  
network [10]  
100mV  
Path-1  
“No signal”  
controls noise gate  
via DSP  
0V  
Path-2  
-100mV  
V(Path-2)  
V(Path-1)  
100mV  
0V  
-100mV  
Figure 11: Signal flow in a digital microphone, with  
0s  
0.5ms  
V(Path-1) AND V(Path-2)  
1.0ms  
1.5ms  
Time  
2.0ms  
2.5ms  
3.0ms  
compressor and peak limiter  
Figure 10: Separate signal paths and re-combination  
result in circuit of Fig. 9, with non-linear crossover  
topology [10]  
All this holds of course only true for the described  
professional digital microphones with very wide  
dynamic range, which the AES42 standardization  
committee had in mind. Other recent microphones with  
digital interface, powered by USB, show a very limited  
dynamic range, often with a noise floor consisting of  
undithered ADC quantization noise plus power supply  
artefacts, and thus offer no advantage over their  
analogue counterparts, other than simple connectivity to  
PC environments [14].  
One side note has to be included, regarding current  
digital recording and monitoring equipment: Often,  
these devices are so designed as to expect only digital  
input signals aligned close to reference studio level, and  
accordingly only offer limited gain manipulation, e.g.  
+10dB, of such digital signals. As has been shown in  
Fig. 5, digital microphones can be recorded directly  
with the widest dynamic range if they are operated with  
no or small digital gain and do not require pulling up the  
gain as high as possible. Still, and be it only for direct  
monitoring purposes, those perfectly recorded low-level  
signals need to be made audible. It would be helpful  
then, to find more digital recording equipment offering  
amplification of digital input signals, and not only the  
analogue ones, over a wider gain range.  
As mentioned, the gain ranging ADC shown in Fig. 7 is  
a floating-point processor with exponents of 20 and 24.  
Combining them does widen the dynamic range by  
4x6 dB = 24 dB, but does not improve their specific  
resolutions. Such a simple switching circuit will then  
modulate from the lower range ADCs noise to the  
higher range ADCs noise whenever the signal passes the  
crossover point, producing a distinct noise peak. A non-  
linear crossover network smooths this transition region  
out, making it inaudible. Properly designed, the result  
can then be a digital microphone with a dynamic range  
of up to 130 dB-A, with all noise components 80 dB  
below the signal over a wide dynamic range.  
5
APPLICATION BENEFITS  
From the above, some benefits for the user become  
immediately clear. With up to 130 dB-A, the dynamic  
range of the conversion covers the complete dynamic  
range of the analogue microphone counterpart. There is  
no need anymore for setting the gain controls in order to  
match input and output levels, as needs to be done with  
standard analogue recording set-ups. When recording to  
an appropriate 24 bit medium, the digital microphone  
can be connected and recorded directly, any gain  
levelling taking place after the recording, or just for  
monitoring purposes. The lower limit for the signals is  
determined by the self-noise of the capsule, thus by  
6
OUTLOOK AND CONCLUSION  
Microphones with digital output are a comparatively  
new concept. Still, they show clear advantages  
regarding gain settings and dynamic range handling, and  
AES 31st International Conference, London, UK, 2007 June 25–27  
5
 
SCHNEIDER  
DIGITAL MICROPHONES FOR HIGH RESOLUTION AUDIO  
they are bound to find wide spread use. As the signal is  
transformed with a high-quality AD conversion to the  
digital domain, it is now also possible to obtain high-  
quality recordings with comparatively inexpensive,  
semi-professional recording equipment, if it does allow  
24 bit word length, with the chosen sample rate. Digital  
microphones will make the job of the studio or location  
sound engineer simpler, reducing the probability of  
errors, thus keeping his mind free to concentrate on the  
acoustical and artistic aspects of the recording.  
[11] Microtech Gefell, Brochure: “MV230 digital -  
Digital  
measurement  
microphone”,  
[12] Schoeps (2006), Brochure: “CMD series”,  
[13] Robjohns H (2006) “Neumann KM184-D”, Line-  
Up (Nov/Dec 2006); also: Brochure: “KM-D”,  
REFERENCES  
[14] Funke R, “Untersuchungen an Mikrofonen mit  
digitaler Schnittstelle auf ihre Einsatztauglichkeit  
in Rundfunkproduktionen“, Dipl.Arbeit, Fach-  
hochschule Deggendorf, to be published (2007)  
[1]  
Reis JP (1861) Über Telephon durch den  
galvanischen Strom, Jahresber. d. Physikal.  
Vereins zu Frankfurt am Main (1860-1861)  
pp. 57-64  
[15] Peus S, “Measurements on studio microphones”,  
preprint no. 4617, 103rd AES Conv., New York  
(1997)  
[2]  
[3]  
Keating DA (1994) “Optical Microphones”, in:  
Gayford ML (1994) Microphone engineering  
handbook, Focal Press, Oxford  
[16] Schneider M, “Eigenrauschen und Dynamik-  
umfang von Mikrophon und Aufnahmekette“,  
20. Tonmeistertagung, Karlsruhe (1998)  
Yasuno Y, Riko Y,“A basic concept of direct  
converting digital microphone”, Acoust. Soc. Am.  
J., vol. 106, pp. 3335-3339 (1999)  
[17] Harris S et al., “A monolithic 24-bit, 96-kHz  
sample rate converter with AES3 receiver”,  
preprint no. 4965, 106rd AES Conv., Munich  
(1999)  
[4]  
[5]  
Paul JD et. al. (1991) “Digital output trans-  
ducer”, Patent US 5051799  
Konrath K (1995) “Konzeption und Entwicklung  
eines Prototyps des digitalisierten Mikrophons”,  
Diplom Arbeit, FH Düsseldorf, Fachbereich  
Medien  
[18] McLaughlin KJ, Adams R, “An asynchronous  
sample rate converter with 120 dB THD+N  
supporting sample rates up to 192 kHz”, preprint  
no. 5191, 109rd AES Conv., Los Angeles (2000)  
[6]  
[7]  
IEC 61938 “Audio, Video and Audiovisual  
Systems – Interconnections and Matching Values  
– Preferred Matching Values of Analogue  
Signals”  
[19] Vanderkooy J, Lipshitz SP, “Resolution below  
the least significant bit in digital systems with  
dither”, J. Audio Eng. Soc., vol. 32, pp. 106-113  
(March 1984)  
Almeflo PO, Johansson M, “Suppression of  
switch mode power supply noise in digital  
microphones”, preprint no. 5341, 110th AES  
[20] Blesser BA, “Digitization of audio: a compre-  
hensive examination of theory, implementation,  
and current practice”, J. Audio Eng. Soc., vol. 26,  
pp. 739-771 (October 1978)  
Conv.,  
Amsterdam  
(2001),  
see  
also  
[21] Fielder LD, “The audibility of modulation noise  
[8]  
[9]  
Harris S et al., “Towards a digitally interfaced  
microphone standard”, preprint no. 4518, 103rd  
AES Conv., New York (1997)  
in  
floating-point  
conversion  
systems”,  
J. Audio Eng. Soc., vol. 33, pp. 770-781 (October  
1985)  
AES42 standard for acoustics – “Digital interface  
for microphones”  
[10] Peus S, Kern O, “The digitally interfaced  
microphone – the last step to a purely digital  
audio signal transmission and processing chain”,  
presented at 110th AES Conv., Amsterdam  
AES 31st International Conference, London, UK, 2007 June 25–27  
6
 

MTX Audio Car Amplifier TA5601 User Manual
National Instruments Network Card NI 9401 User Manual
NEC Projector XL 6500 User Manual
NETGEAR Network Hardware XAV5501 100NAS User Manual
Nilfisk ALTO Vacuum Cleaner BK 800 User Manual
Nortel Networks Network Card NN44200 313 User Manual
Olympus Camcorder E 5 User Manual
Olympus Carrying Case PT 017 User Manual
Olympus Microscope Magnifier AX70 User Manual
Panasonic Cordless Telephone KX TGA641 User Manual