Jesse,
I'd like to try to answer this by introducing the concept of "natural recor=
ding", which means that the recorded sounds are reproduced at the same soun=
d pressure level (SPL) as they originally were created in their natural env=
ironment. Now this concept may break down shortly, but it's a starting pla=
ce.
When we make a recording and then reproduce it, there is the level of the o=
riginal sound, the sensitivity of the microphone, the noise floors and sens=
itivities of the microphone preamplifier and recorder, and then the gain of=
the power amplifier and the sensitivity of the speakers that are used. Wh=
at is typically done is that the listener turns the gain of the reproductio=
n system (amplifiers and speakers) up until it seems about right. Hopefull=
y this will be about the same level as the original sound, but it seems as =
if quiet sounds get turned up more, and loud sounds are not played back qui=
te as loudly as reality.
How much gain it takes in the playback system to reproduce the sound at the=
original level depends upon how much gain is applied in the recording syst=
em, including the sensitivity of the microphone. The sensitivity of a micr=
ophone is typically specified as dBV/Pa (short for Pascal). 1 Pascal is eq=
ual to 94 dB SPL, which is a moderately loud sound. The gain of the microp=
hone preamplifier is usually adjustable over a very wide range, and the gai=
n of a modern recording system is likely to be dBV for 0 dBFS (dB relative =
to digital Full Scale).
Assume that we all use microphones that have -40 dBV/Pa sensitivity, set ou=
r microphone preamplifiers to have 40 dB of gain, and use recorders for whi=
ch 0 dBV causes a FS signal. Then 94 dB SPL at the microphone will be full=
scale in the recording. A louder signal will clip and sound distorted.
On the reproduction end the signal will be fed to a DAC, which almost alway=
s has an output of 2 V for 0 dBFS, or +6 dBV. Then it is applied to a prea=
mplifier which contains the level control used to set the replay acoustic l=
evel, to a power amplifier which probably has about 30 dB of gain, and to a=
loudspeaker. The average loudspeaker has a sensitivity of 88 dB SPL/1 Wat=
t, where a Watt into an 8 Ohm speaker is 2.83 Volts, or 9.04 dBV. If you s=
um all of these individual gains together, you get 115 dB SPL. You need to=
have a gain of -21 dB in the preamplifier in order to get our desired 94 d=
B SPL.
One would hope that a good microphone preamplifier will have lower noise th=
an the microphone does, but that isn't always the case. Some microphones h=
ave lower noise, especially at high frequencies, than typical preamplifiers=
. Luckily noise in the top octave doesn't usually contribute to the audibi=
lity of the microphone self noise.
OK. What does all of that have to do with your question about hearing the =
self-noise from your Rode NT-1A? If you turn the gain in the replay system=
up high enough, you will always be able to hear the self-noise of the micr=
ophone, unless some other noise masks it. If your system is calibrated as =
a "natural recording" system, then the mic noise probably will be only bare=
ly audible. If the noise from the NT-1A had a white spectrum, which it doe=
sn't, then that self noise would be right around the threshold of hearing. =
Without looking at the spectrum of the noise, it is difficult to estimate =
exactly where it lies relative to the threshold of audibility.
If the system isn't adjusted to "natural recording" levels, then all bets a=
re off.=20
It is also possible that one might wish to record things that are extremely=
quiet. Perhaps even sounds that are naturally below the threshold of (hum=
an) hearing. I had planned for some several weeks to address that question=
in this forum, but haven't done so yet.
What most people are interested in doing is to record sounds that have sort=
of average intensities, but which may be occurring too far away for the re=
cordist to get reasonable access. And that is the reason to use a directio=
nal microphone. Which is a whole subject in itself.
I hope this helps.
Eric
jessehammons <> wrote:
Out of curiousity, if a microphone has 5 dB of self-noise, and you
turn the gain on the pre-amp up to 50 dB, how loud is resulting signal
(assuming the microphone is place in a sound-proof chamber of some kind).
What if the mic has 10 dB of self noise?
I'm trying to understand how quiet a place can be before one has to
give up and use some sort of parabolic dish to collect more sound.
Even the very expensive microphones talked about on this list have a
self-noise rating of 5-6 dB or so. A quality pre-amp won't introduce
much extra noise on top of that, but I am assuming it will amplify the
mic noise just as it would any other signal.
I used my new "pro" recording setup (rode NT1-A mics + mic2496
pre-amp...It's "pro" for me anyway) to record a location the other
night with the gain turned up to around 50 dB or so. I haven't fully
analyzed the test recordings yet, but I think I can hear the mic
self-noise in the recording using that amount of gain on the preamp.
I'm wondering if I can calculate whether that is possible or whether
the self-noise value wouldn't even be audible if all my equipment is
running within spec.
Thanks,
-Jesse
[Non-text portions of this message have been removed]
________________________________________________________________________
________________________________________________________________________
|