At 10:43 AM +0000 3/30/08, Raimun d Specht wrote:
>Rob Danielson wrote:
>
>> I don't understand why the mics with lower sensitivity appear to
>have
>> lower effective noise on the Rane chart. I asked about this some
>time
>> ago; maybe someone knows.
>
>Hi Rob,
>
>The dBu noise floor values on the Rane chart are absolute noise
>voltage levels that depend on both the equivalent SPL noise level
>(the "dBA" specification) and the sensitivity (the "mV/Pa"
>specification) of the microphone. See also the Rane table 3 at
><http://www.rane.com/note148.html>http://www.rane.com/note148.html
>
>Imagine you had two different microphones that have both the same
>noise floor specification of 14 dBA, but different sensitivities of
>20 mV/Pa and 50 mV/Pa. The 20 mV/Pa model provides a noise voltage
>floor of -112 dBu. The 50 mV/Pa model produces a higher absolute
>noise voltage of -104 dBu. The 50 mV/Pa model could just be
>interpreted as a 20 mV/Pa microphone with an additional ("zero
>noise") 8 dB preamplifier.
>
>So, what does this mean to us? If you already have a recorder with a
>mediocre preampilfier, you should select the microphone with the best
>(highest) "mV/Pa" specification. In other words, a more sensitive
>microphone already includes a "zero noise" preamplifier that lowers
>the noise performance requirements on the recorder side.
>
>Regards,
>Raimund
Hi Raimund-
It seems the intended use of Table 3 http://rane.com/note148.html is
for determining whether the noise output of a preamp is at least 10dB
lower than the noise output of the mics. It creates dBu, A-weighted.
equivalents for the mics and your A-weighted, EIN measurements on
http://www.avisoft.com/recordertests.htm plug in with no additional
adjustment.
My question is whether it is appropriate to use Rane Table 3 to
compare fully adjusted noise output between mics? When I try to do
this, it seems I should interpret higher negative values as lower
effective noise output, like preamp numbers. For example, a Rode
NT1-A (5.5 dB(A) and 25mV/Pa) computes to -117.5 dBu and a MKH-20 (10
dB(A) and 25mV/Pa) computes to -113.5 dBu (all numbers are
A-weighted). Assuming all noise is equal for the moment, and 1 dB is
the smallest unit of difference that one can hear, the table suggests
the MKH-20 has about 4 audible increases or "steps" in noise above
the noise produced by the Rode NT1-A. For this example, the Table
seems to be in agreement with my experience with higher negative
numbers resulting is less, cumulative noise output. There are other
Table 3 computed mic comparisons, however, that come out contrary to
experience.
Are you using Table 3 to compare noise output between mics? Are you
saying that the Table 3 grid numbers are not really noise output
equivalents between mics and pres-- that mics compute to "noise
voltage floor" numbers where lower negative numbers effectively mean
less noise output? [I'm going on your example, "The 20 mV/Pa model
provides a noise voltage floor of -112 dBu. The 50 mV/Pa model
produces a higher absolute noise voltage of -104 dBu. The 50 mV/Pa
model could just be interpreted as a 20 mV/Pa microphone with an
additional ("zero noise") 8 dB preamplifier."] I can see how -104 dBu
is -8dB hotter than -112 dBu, but I thought that Table 3 attempts to
show equivalent noise output, apples and apples, for all mics and all
pres.
I'm with you regarding the general relationship between higher mic
sensitivity and improved noise performance with moderately noisy mic
pres. Rob D.
--
|