My car's system hasn't been touched in about a year, and I lost whatever notes I made when setting it up. So, I decided to take some measurements to record my gain staging for fiddling with it later. I noticed some interesting things while doing this that I thought I'd share.
****************************
Test 0dBFS
Dell headphone jack > miniDSP > 100kOhm pot > ASUS mic input > listen through headphones
All miniDSP faders @ 0dB, no filters
Audacity tone w/amplitude 1.0, Dell volume 90
(this was highest volume before miniDSP clipping)
(tested @ 1kHz, verified laptop not clipping)
(changing biquad, in/out faders had no effect)
freq Input Output Gain
Hz Vrms Vrms V/V
1k 0.920 0.876 0.952174
60 0.923 0.8785 0.951788
So, slight voltage drop. (-0.43dB power)
****************************
Test fader dB
[unspecified input] > miniDSP > Fluke
Fader Output Calc Calc
dB Vrms dB(V) dB(W)
0 0.5023 n/a n/a
-3 0.3553 -3.007 -1.504
-6 0.2516 -6.005 -3.003
So, fader units are dB(amplitude), not dB(power).
****************************
Test volume meters, static level
[unspecified] > miniDSP > [unspecified]
Input Input In_met Output Out_met
Vol Vrms dB Vrms dB
87 0.946 -0.3 0.897 -0.3 [should be clipping]
86 0.867 -1.1 0.823 -1.1
So, maximum input/output signal shows -1.1dB on the meter.
****************************
Test volume meters, change in level
[unspecified] > miniDSP > [unspecified]
(0.867 Vrms, whether input or output not noted.)
fader meter
dB dB
0 -1.1
-3 -4.1
-6 -7.2
-9 -10.2
So, meter dB follows changes in fader dB w/initial offset.
****************************
The tests were run with a miniDSP 2x4 Rev.B purchased 23 Sep 2012, running the 4-way Advanced plugin purchased 27 Jan 2015, powered through the USB port with my laptop.
With that said, I think this means my settings have been wrong this whole time. I had an EQ band set at +6dB, so to compensate I took the input fader to -6dB. The above tests verified that the dB units displayed on the fader are in terms of amplitude, not power. (I am not into pro audio, so I do not know if this is standard. I suspect it is, since the other common units dBu and dBV are amplitude-based.) The question is whether or not the biquad filters operate the same way. To my knowledge, filters are usually expressed in terms of power. (For instance, Butterworth filters are characterised in part by their half-power [-3dB] frequency.) If this is the case, then the appropriate fader position to allow for a +6dB filter would be -3dB. Unfortunately, I am away from home for the week and cannot verify.
Just food for thought.
* Edit -- Perhaps I was mistaken... Maybe while the shape of a filter is given in terms of dB(power), it's actual gain (what we would call boost/cut) is given in dB(amplitude). If that is the case, everything is fine and I just learned something new.
** Edit -- Crap... I mixed up the dB(V) and dB(W) columns in the fader test. That means -6dB = 1/2 amplitude (but 1/4 power), which is what we expect. So I guess I'm an idiot.
Sorry, guys.
****************************
Test 0dBFS
Dell headphone jack > miniDSP > 100kOhm pot > ASUS mic input > listen through headphones
All miniDSP faders @ 0dB, no filters
Audacity tone w/amplitude 1.0, Dell volume 90
(this was highest volume before miniDSP clipping)
(tested @ 1kHz, verified laptop not clipping)
(changing biquad, in/out faders had no effect)
freq Input Output Gain
Hz Vrms Vrms V/V
1k 0.920 0.876 0.952174
60 0.923 0.8785 0.951788
So, slight voltage drop. (-0.43dB power)
****************************
Test fader dB
[unspecified input] > miniDSP > Fluke
Fader Output Calc Calc
dB Vrms dB(V) dB(W)
0 0.5023 n/a n/a
-3 0.3553 -3.007 -1.504
-6 0.2516 -6.005 -3.003
So, fader units are dB(amplitude), not dB(power).
****************************
Test volume meters, static level
[unspecified] > miniDSP > [unspecified]
Input Input In_met Output Out_met
Vol Vrms dB Vrms dB
87 0.946 -0.3 0.897 -0.3 [should be clipping]
86 0.867 -1.1 0.823 -1.1
So, maximum input/output signal shows -1.1dB on the meter.
****************************
Test volume meters, change in level
[unspecified] > miniDSP > [unspecified]
(0.867 Vrms, whether input or output not noted.)
fader meter
dB dB
0 -1.1
-3 -4.1
-6 -7.2
-9 -10.2
So, meter dB follows changes in fader dB w/initial offset.
****************************
The tests were run with a miniDSP 2x4 Rev.B purchased 23 Sep 2012, running the 4-way Advanced plugin purchased 27 Jan 2015, powered through the USB port with my laptop.
With that said, I think this means my settings have been wrong this whole time. I had an EQ band set at +6dB, so to compensate I took the input fader to -6dB. The above tests verified that the dB units displayed on the fader are in terms of amplitude, not power. (I am not into pro audio, so I do not know if this is standard. I suspect it is, since the other common units dBu and dBV are amplitude-based.) The question is whether or not the biquad filters operate the same way. To my knowledge, filters are usually expressed in terms of power. (For instance, Butterworth filters are characterised in part by their half-power [-3dB] frequency.) If this is the case, then the appropriate fader position to allow for a +6dB filter would be -3dB. Unfortunately, I am away from home for the week and cannot verify.
Just food for thought.
* Edit -- Perhaps I was mistaken... Maybe while the shape of a filter is given in terms of dB(power), it's actual gain (what we would call boost/cut) is given in dB(amplitude). If that is the case, everything is fine and I just learned something new.
** Edit -- Crap... I mixed up the dB(V) and dB(W) columns in the fader test. That means -6dB = 1/2 amplitude (but 1/4 power), which is what we expect. So I guess I'm an idiot.
Sorry, guys.
Last edited: