Raimund Specht wrote:
>--- In Walter Knapp <>
>wrote:
>
>>* The sensitivty of the Rode NT3 is 12dB(A), about 4dB less than
>>some other mics with the same self noise rating I regularly
>>use.Whenever the record gain is high, its easy to hear this
>>difference in sensitivity because its translates into 4dB? of
>>additional noise
>>when the sound is played back at final mix level.
> This is a whole 'nother aspect of the self noise equation. In what
>I said, mic sensitivity was more or less ignored. But it has a
>bearing in mic choice vs self noise. The better the mic sensitivity, the l=
ess
>youwill be increasing gain to get the meter reading you want, so the
>lessyou will be lifting the noise floor.
>
>
>Walt,
>
>I'm not sure if I understand correctly what you mean. The inherent
>noise of microphones is always specified in absolute sound level
>figures that correspond to a certain absolute environmental sound
>level. The absolute sensitivity of a microphone does not necessarily
>influence it's noise performance, as long as the additional noise
>introduced by the required pre-amplifier is low compared to that of
>the microphone. Most state-of-the-art professional pre-amplifiers
>introduce much less noise than the microphone. I recently tested my
>MKH 60 with the pillow method you suggested. After replacing the
>microphone by a dummy resistor, the remaining noise floor was much
>lower (even at high gain settings).
>
>Nevertheless, a high sensitivity is of course a good property of a
>microphone, especially if you use a relatively poor (noisy) pre-
>amplifier. When I compare my old Sennheiser K3/ME80 (5mV/Pa) with
>the new K6/ME66 (50mV/Pa), I can not find very dramatic differences
>in their noise performances. The lower output level of the old one (-
>20dB) can be compensated by adding more gain on the pre-amplifier
>without any loss of fidelity. Listen to this recording of a European
>Nightingale made with the outdated K3/ME80:
>htttp://www.avisoft-saslab.com/sounds/ntg.mp3
>
>The deviations between microphone sensitivities may be caused to
>some extent by different internal electronic circuits. It is often
>not very important, whether the amplification takes place within the
>microphone housing or in the pre-amplifier of the recorder. However,
>especially under critical circumstances, higher microphone output
>levels are less sensitive to small noise voltages (e.g. hum)
>entering the microphone extension cable.
>
>Regards,
>Raimund
>
Hi Raimund--
Part of our difference in whether mic sensitivity is a factor could
be our usual recording environments/goals. When isolating a bird call
with a mkh-60 or parabola, the noise introduced by sensitivity is
usually not audible in the sound file. The saturation of the
sustained body of the call might be 20-40% with percussive peaks in
the 60%-100% range. To contrast, in the middle of the night in a
rural setting, with an MP-2 cranked max and mkh-20's (10dB[A] noise),
the noise floor of the forest will produce 6% sound file saturation
with distant effects reaching ~12%. The later file has to be boosted
a great deal digitially to be heard upon playback and is a demanding
test of noise, from any source, including bit depth and resolution.
As to whether low mic pre noise can negate the sensitivity factor,
I'm going on some math Bob Cain did some months back about the
overhead in mic pres with very low noise mics:
MIC DIY mics:
>The MP-2 spec
>says that the input noise is -128 dBV so that the self noise of the
>mic is 12 dB above the noise of the channel. This isn't as big a
>difference as I expected but if you take a noise signal you won't
>hear much difference at all between it and it added to a noise
>source 12 dB lower so that it is safe to say that the mic self
>noise, low as it is, will mask the noise of the MP-2.
>
>The Rode NT-1A being 5 dB SPL is only 7 dB above than that of the
>pre which makes the pre's contribution to total noise more
>signifigant but I just did a quick test using white noise and I can
>just hear the transition between two segments, one containing noise
>and the other containing the same summed with another segment 7 dB
>lower.
Not as much room between mic noise and the noise floor of the preamp
as many would guess using gear that can actually test mic pre
performance. Sensitivity will not play a role with the very good mic
pres, but the pres most field recordists use don't fall into this
category. MD manufacturer's, for example, don't publish noise specs
on the mic preamps of their consumer recorders. The mic pres in the
TASCAM DAP1's I frequently borrow are not as quiet as the MP-2 and
they do start to introduce noise when cranked with mkh-20's to my
ears. I've not seen useful noise specs for the TASCAM either. Walt
says his Portadisk's pre in the sweet spot is on par with his MP-2's.
What I'm after is a better understanding of what leads to less noise
and more transparency in field recordings. The answers are likely to
be very different when assessing loud sound environments. If you get
a chance to listen to Evert's MP3's made with the SoundMan classic II
binaural mics, his recordings are beautifully intoned and very
clean. They were made in places like inside a moving train and next
to a go cart track. Gordon Hempton uses the Neumann KU81i for his
binaural work deep in the woods.
But many people find themselves in situations where mic pre gain must
be cranked. Here's what I'd draft as a working guideline: If a mic's
self noise is above 20dB(A) its going to introduce noise, period,
when the gain is cranked. If the mic's sensitivity is low to boot,
say < 10 mV/Pa, the sound file will also be undersaturated and the
bit depth will be compromised and the resulting sound will acquire
even more muddiness. When the self noise of a mic starts dropping
below 14dB(A), good mic pre will probably not introduce noise when
the gain is fully cranked, but other mic pres, like those in consumer
MD's, will introduce noise.
What's a "good" mic pre in terms of self-noise? If the manufacturers
provided apples and apples self noise specs that applied between mics
and mic pres (when cranked!), then anyone would be better able to
decide whether to invest in a better mic or a better pre. No surprise
we plod through a lot of confusion trying to share experiences and
relate them to specs. I'm sorry for the confusion that comes when
applying what I write to situations where sound levels are high and
mic pre gain is low. I've acted on others' advice like that from the
DAT HEADS FAQ pasted below and I'm interested in trying to adapt them
for field situations. Rob D.
[DAT HEADS FAQ]:
>Which Mic is Right?.....by Michael Cooper
>
>If you want dead-quiet recordings, a mic's self-noise and
>sensitivity performance should factor into your choice. These two
>specs go hand-in-hand in determining how much background hiss a mic
>will contribute to a recording. Keep in mind that specs are not
>always derived through using the same test methods, so they can be
>misleading. Still, you should look for a self-noise no higher than
>22dBA if possible. The quietest solid-state mics offer 17dBA
>self-noise or better. With an astonishing rating of 7dBA, the
>Neumann TLM 103 ($995 list) is one of the quietest mics available
>today. Many tube mics measure 20dBA or worse, but there are
>exceptions: The Neumann M 147 ($1,995 list) and Brauner
>Valvet($2,700) tube mics yield impressive self-noise specs of 12
>and 13dBA, respectively. A mic with low self-noise can still produce
>noisy recordings if its output level, or sensitivity, is weak. Using
>a weak mic on a quiet performance forces you to crank up your mic
>preamp to the top of its range, thus adding noise. Ideally, you
>want a mic that offers low self-noise and high output. Look for
>sensitivity of 10 mv/PA (10 millivolts per Pascal, a measure of how
>many millivolts a mic will output per a given sound pressure level)
>or higher.
[Non-text portions of this message have been removed]
________________________________________________________________________
________________________________________________________________________
|