Basically the question. Should sum of all tracks be 0 Vu -18rms or should each channel hit that number?

 

This could become an interesting thread.

Neither, why should it at all? I'm doing neither of this, simply because it's not necessary (at least not for the way I'm working) and I'm very very rarely using a separate VU meter plugin.

Be aware that
- percussive sounds can easily peak at several +dB when set to this,
- not all analogue emu pluggies are calibrated to -18dB = 0VU (these three compressors are fed with the same signal)


Oh and BTW, it's 0VU = -18dB FS peak, not RMS.

 

Okay. At what level do you gainstage then and how?

 

Neither one, nor the other.

VU meters are useful if:

• you are an old fart who has switched from analog to digital and is used to working with VU meters
• you are a young fart who, like me, has switched from analog to digital and is used to working with VU meters
• you need an RMS meter that displays the RMS level around a reference level of your choice in high resolution

BTW there are a lot of threads about VU meters on this platform where you can read how @No Avenger and I (and many others) argue about the usefulness and purpose of VU meters. So there is already enough reading material for you. The search function is your friend

 

-1 dB True Peak. Meters set to pre-fader and peak hold. This way my DAW monitors all the time if I accidentally raised the output of plugin.
When I need to lower a signal for an input sensitive plugin, I either use the plugin's input knob or use a volume plugin before it. But the latter is rarely the case.

 

VU meters on daws are the stuff of nightmares and youtube tutorials. You're probably better off without them, but if you really need to pick one of those options you should be fine with 0dB per track at that calibration.

 

What these VU meters are really useful for is to determine the momentary loudness of individual tracks. For example with vocals. Put a VU meter on a vocal track after a compressor (or during clipgain automation), calibrate the VU meter so that the needle hits at ~0VU for most of the time, and then keep track of the overall dynamics of the needle when compressing.

But for general level decisions or the stereo sum, a VU meter is unsuitable. Here is a mix bounce from me. Most of the time, the VU meter needle is just stuck at +3:



More useful is a loudness meter with LUFS and TP detection, as you can see.
When limiting your sum, just make sure your limiter is oversampled with x8 and leave 0.2dBFS TP headroom for possible underreads.

 

Those of us who used analogue magnetic recorders in their youth learnt early on that VU meters are crap. I'll repeat that: VU meters are crap! They're just a slow moving-coil meter fed via a full-wave rectifier. As far is recording is concerned, what matters is the peak value of the signal, not a slow average (misleading called RMS; VU meters don't measure true RMS, they measure rectified-average, like any other moving-coil meter).

The right way to do this job is to use a peak reading meter. This is true for both analogue and digital recording. Many years ago the BBC developed a meter called the Peak Programme Meter, or PPM, which has been used throughout the UK, in broadcast and recording studios, and which has latterly spread to Europe. This device gives a pretty accurate reading of peak values. It has a logarithmic scale, rather than the VU meter's linear one, so it covers a bigger dynamic range. As you might imagine, the hardware ones with all their electronic gubbins were rather expensive compared to the cheapo VU meters used in America.

Unfortunately, the term "VU meter" seems to be used today for arbitrary level monitoring devices, such as the ones provided by DAWs. These generally don't conform to the VU meter standard (ANSI C16.5), although you can get plugins which claim to.

The reason why a VU meter is crap is that it's useless for monitoring peak values. You have to guess, depending on the peakiness of the audio material, and/or allow a large safety margin.

Anyway, to return to the original question, the purpose of any audio meter is ensure that overloading (clipping) doesn't occur at the one extreme, and that the signal-to-noise ratio isn't awful at the other extreme. With modern digital systems it's only clipping you need to worry about. But with 18 dB headroom, this shouldn't be a problem.

Meters aren't useful for balance or mixing. This is always done by ear.

So my argument would be: you need a peak meter for digital recording (via microphones etc) and for the digital master track, in both cases to avoid clipping. You don't need them anywhere else, except as confirmation that some signal is present or that it's getting a bit close to the clipping level. If you want to put a VU meter on every track then by all means do so, but it's not necessary as DAWs have their own metering, which is good enough and probably less misleading than a VU meter.

 

Except for visual pleasure, I never understood why anyone would want to simulate analog elements that are actually inferior to digital or software versions.
VU meters are one such element. It may look neat but they are inferior in performance.
Analog modeling is a different matter, of course, and pleasing GUI is a different matter as well.
So i can see having a VU meter just for looks that does NOT simulate a coil-driven thing but is not quite instant in visual response, but is augmented with digital readout and peak-hold markers....
But to simulate a coil-action device makes no sense except to look at it.

 

100%
Whether they work accurately or not, and regardless of whether they are or arent necessary, the individual VU meters on my mixing board look good as fuck. Clients love them, more so if they are high or drunnk. They are wallet openers