Topic: Request for help - calculating mic output and Pre input (crazy math stuff)

[landshark]

Hi all -

Ok, I'm trying to work the math around figuring if my mics are overloading my preamp.  Of course, I don't know how to go from the measurements of mic efficiency (-34 dBV/Pa & 20MV/Pa) (Senn MKH 8040's, also AKG 480's) and the input maximums on the pre (-16 dBu mic / +16 dBu line [-26 to +4 dBu nominal]) (Korg MR-1000). I'm assuming I need to have an assumption on the ambient sound I'm recording, which I'll assume to be 120 dB.

OK, I found what I need here: http://www.sengpielaudio.com/calculator-transferfactor.htm (scroll down to the table)

So, at 20 mV/Pa, at 120 db the output is -6 dBu.  More than enough to overload the mic pre inputs at -16 dBu.  Crud.  Anyone know the formula used to create the table values?  I'd like to be able to figure out the values for 110 db down to 80db or so as well.


PS:  I found a good site for sound-related calculators, if anyone's interested:  http://www.sengpielaudio.com/Calculations03.htm

[landshark]

Hi Ms -

I'm not following, probably because this stuff makes my head hurt.  What I'm trying to do, is work from a known mic sensitivity, figure out the dBu of output given an SPL measured in db (or Pa).  Or:

Given a sensitivity of X mV / Pa (or X dBV/Pa), an SPL of Y db, calculate the max output as Z dBu. 

So I have sensitivity and sound, trying to calculate Mic output.

Mike

[Todd R]

MSH has the formulas you want, though I think he's got a mistake -- the formulas should be in Volts V, not millivolts mV.  A dbu is referenced to 775mV, or 0.775 volts.  A dbV is referenced to one volt (1000mV).

So for instance, starting with the second formula and the AKG mics:

dBu = 20 * log (V/0.775) + dbSPL - 94 (since the sensitivity of the mics is referenced to 1 Pascal = 94db)

For the AKGs at 20mV per Pascal (20mV = 0.020V), and a SPL in the club of 120db:

dBu = 20 * log (0.020/0.775) + 120 - 94

dBu = -5.76   or ~= -6db as the site you referenced says.


The Senn mic is more difficult since you will need to solve for mV/Pascal from the first formula:

-34 dBV/Pa = 20 * log (V) + 94db (1 pascal) - 94
log (V) = -1.7, using inverse log function, V = 0.020 per pascal, or 20 mV/Pa  -- same as the AKGs

So for either mic, use the dBu = 20 * log (0.02/0.775) + dbSPL - 94 formula and plug in the different dbSPL values you want.

FYI, on Windows machines, the calculator function under Pgms>accessories has an option for scientific calculator functions which include LOG.

Also, I'd assume you'd have ~118db SPLs at a loud club as a guess.  Probably more like 110-115, but you should probably plan for 118-120 as an upper end.

[dactylus]

Very helpful - posting for a reference point!!

Thank you.

[landshark]

Awesome.  Now I understand.  Well, actually, I don't "understand" but I get it enough to do the math.  That gives me:

Db   dBu
0   -125.76
10   -115.76
20   -105.76
30   -95.76
40   -85.76
50   -75.76
60   -65.76
70   -55.76
80   -45.76
90   -35.76
100   -25.76
110   -15.76
120   -5.76
130   4.234

Which I'm thinking is the db to dBu table for the Senns and AKGs.  Is that right?  Is it a linear relationship where a 10 db change in sound gives a 10 dBu change in power from the mics?

My guess is these are measured at a specified reference frequency such as 1khz?  The reason I ask is the Senn's have crazy low end pickup and seem to therefore output more energy when exposed to low frequencies than the AKGs.  Or maybe I'm not really understanding sound and energy (which is probably more likely).

What this also tells me is that the mics are putting out enough energy to be in the "nominal" band for Line in when they're in the 100 to 130 db range.  Hmm.  If the low end is putting out more energy, causing the effective dBu from all frequencies to be higher than that at the test frequency, then maybe I cold do line in from the mics.  Too bad the Korg only accepts line in through the TRS, and doesn't provide phantom on the TRS.  Oh well.

Mike

[DSatz]

Yes, microphone sensitivity is measured at 1 kHz on axis in a nominally free-field environment (anechoic chamber or the like). That figure then serves as the basis for the usual on-axis frequency response diagram, which will normally show exactly 0 dB response at 1 kHz. That 0 dB represents the nominal sensitivity of the microphone.

To figure out the maximum SPL I can record with a given combination of microphones and preamp or recorder, I find it convenient to think in terms of voltage rather than dBu. If a preamp overloads at 500 mV and my microphones put out 10 mV/Pa, then I know that the preamp will reach its limit when my microphones are exposed to about 128 dB SPL, since 1 Pascal (1 Pa) is about 94 dB SPL, and 500 mV/10 mV is 33.98 dB; 94 dB SPL + 33.98 dB = 127.98 dB SPL.

Simpler yet, I know that my best microphones can put out just under a Volt at their limit, so I use preamps that can handle at least 1 Volt; then I never have to worry about preamp input overload at all.

--best regards

[busterr]

Not to stray too far off the "crazy math stuff", but is there a reason why the use of attenuators would not be a good solution to this?

That is what I have been doing with my mr-1000 after trial and error of my on-mic -15db pads not being enough to keep it's pres from brickwalling. Just curious if I have gone the wrong route, or if there are any issues I should be aware of...been more than happy with the results FWIW.

[DSatz]

busterr, as long as the attenuators have well-matched series resistors in the two modulation leads (so as not to disturb the input balance, which is what creates the common mode noise rejection), and as long as the attenuators aren't loading down the microphones, then they're generally the best solution.

They are specifically preferable to using the "pad" switch on the microphone, since they reduce the microphone's inherent noise, and any hum or noise pick-up in the cable, by the same degree as they reduce the signal levels.