dBFS=20*log10(volume/max volume)
Hi hgo58 , thank you for your comments.
Yes, I agree that the "VU meter" is not a "VU meter". And 0dB VU was used confusingly by me.
But I keep my opinion that the meter is now a proper and useful digital peak meter, with 0dB display = 0dBFS digital signal.
Sorry for some lengthy well known definitions below, I just want to make my point clear, maybe just loud thinking for myself.
Precisely, a dB level L is just a definition of a logarithmic power ratio:
L = 10*log(P/P_ref), where P_ref can be anything with the meaning of a power, like a reference electrical or optical power or acoustic power.
As power P is proportional to amplitude squared A^2, (A = voltage in electronics at a defined resistor, dimensionless amplitude in digital, or sound pressure level as amplitude-like quantity in acoustics) the same level in dB reads by the laws of logarithms log(x^2) = 2*log(x)
L = 20*log(A/A_ref), A_ref being any reference amplitude.
Why saying this?
"0 dB" could be anything, defined only by the respective P_ref or A_ref, if P = P_ref or A = A_ref.
Depending on the reference, there are many dB definitions. For dB in acoustics A_ref = 20 µPa sound pressure, for dBm in RF electronics or fiber optics P_ref = 1mW. And here the culprit starts, as we all know.
Now to true VU:
“The reading of the VU volume indicator shall be 0 VU when it is connected to an AC voltage equal to 1.228 Volts RMS across a 600 ohm resistance (equal to +4 [dBu]) at 1000 Hz.”
And referring to dBu, the reference A_ref is 0,7746V RMS which is 1,095 V peak of a pure sine, equal to 1mW at 600Ohms ... let's stop here.
My point is:
These dBVU as well as dBu, relevant for analog signal levels, are pretty useless in digital audio, having dimensionless numbers as signal representation.
Unless WiiM claims that the "VU meter" is related to the absolute DAC output voltage and to that VU definition from the 1940ies they shouldn't call it "VU meter".
Maybe we could agree on this.
To be more precise, lets now talk about instantenuous signal amplitude, either in Volts or a dimensionless number in digital audio.
Using double precision 32bit floating point math, for internal processing, these digital signal amplitudes are virtually limitless.
For a peak meter to avoid digital clipping at an input or output device, i.e. ADC or DAC, with limited voltage range (full scale, FS) V_FS or limited digital number range N_FS, dBFS is used, as we all know:
L [dBFS] = 20*log(A/A_FS).
This is used in many if not all professional digital audio devices to avoid digital clipping, both DAW or mastering tools like Wavelab or hardware audio interfaces like from RME for music production I'm working with. Besides dBFS, there are other usefull digital audio metering dB definitions, average power, loudness based metering units and so on.
For a next check:
I generated a 100% full scale amplitude sine wave and square wave in Wavelab, both properly metered at 0dBFS, there. Of course the square has double = +3dB average power, but we are talking about peak amplitude here. My RME audio interface also shows 0dBFS for both waves. Optical out SP/DIF to the WiiW ultra, 0dB input gain, EQ and RoomFit of, Fixed Volume 100%:
And voila, with both sine and square, the meter reading is 0dB, as it should for a digital peak meter with a 0dBFS signal.
Adding extra gain above 0dB generates audibla distrortion with the sine. Not audible with the square, of course.
Conclusion:
The WiiM Ultra meter is now a correct and useful "Peak Level Meter" in dBFS.
It is confusing from WiiM to call it still a "VU meter", to claim any analog heritage, unless they really mean it. I think not.
