please explain

 

Are we talking peaks or RMS? Because peaks at -18 dB does sound a bit scary.

 

Graham Contradict himself a little bit by saying RMS and then talking about "kissing the 0VU"; I say this because the way you "kiss" it is the shortest transient that can mostly be achieved by db peak... Now, I understand that real VU have a slower than db response, but a faster than RMS response... and that is also completely dependant on the on the buffer that particular emulation use or in the tweak the VU meter has (at least In the ones I had in hand [my father is an electrician and used to fix some] there was a screw to manipulate the response, with a fastest close to nanoseconds)... I also don't know the method for averaging the signal that is used in VU's so this is as far as I can say to this. I would bet "RMS" is the sane route...

Now, when it comes to samplepacks, they are often fixed at a particular RMS or a particular peak as maximum and consistent fluctuation in the sound (be it by compression or straight up limiting). So let's say the peak of a loop is around 0db and the RMS is closer (depends on the density on the sound) with little to no fluctuation whatsoever, I would go with db... If the fluctuation in the RMS is higher, I would average the center of the fluctuation closer to -18 and don't bother with the peaks...

But take all that I've said with a grain of salt, I'm no pro, just a years long hobbist

It is scary indeed, but the low volume levels will free up your perception a lot, and your master headroom (You can always crank your output up if if its too low)

 

Flux- Pure Analyser, for me has everything you need for metering, true peak and LUFS meters and way more.
However you will need a screen dedicated for it.
TC LM6n is the next best thing, more focused to level-energy metering.

 

Well, I'm asking for a clarification because many of my audio tracks -- back when they were all normalized to -0 dB before mixing -- have had RMS's in the neighborhood of -12 dB to as low as -19 dB RMS, so saying "they should have an RMS in the neighborhood of -18 dB" sounds reasonable. I don't compress or limit on the way in and I'm recording or bouncing from VIs without going to tape or riding levels (unless there's some weird sort of one-time peak that throws off the whole track, then I'll drop in a Waves L2 to catch it). Obviously, peak levels of -0 dB are waaaaaaaaay too high, so I've recently decided to experiment with peak levels of -6 dB, which makes sense to me because most DAW and analog faders go up to +6 dB. I've read articles stating they should peak out at -12 dB, in deference to analog and analog meters …. but that seems too low to me. As it is, after mastering, I'm averaging between -14 dB and -10 dB, and the mix downs are plenty loud. So the mixing seems like it's on the money, it's the tracking stage where I think I'm having problems. The audio files are too loud (all peaking at -0 dB before mixing) and need to be reduced, just can't figure by how much.