gain staging, meters, luffs

paresh

Member
I'm trying to get up to speed with luffs. I've been doing old-school gain staging, - 18 DB RMS and - 12 peak. So far from what I'm reading, the integrated range for luffs on the master bus is somewhere around -9 - 20 to leave plenty of headroom for mastering. I know there's no exact rules, but do I use the same range in luffs, or whatever is the correct range for gain staging buses and individual channel strips as well? What about the very first stage of input gain for each instrument at the beginning of the signal chain? Is that about the same also? it's a little confusing because the manual gives all kinds of ranges for different streaming services but doesn't really say what's best for just CD quality. Any clarification is much appreciated! PS I'm using the Youlean meter but may also try others.

http://charleskasler.wix.com/sunmoon-yoga
 
I'm not even sure LUFs matter for the vast majority of recordings, as the old system of knowing your ears works pretty well - and popping onto spotify reveals how the resultant publishing levels get faffed with anyway.

When people started talking about LUFs, I discovered Cubase had meters that had the readouts, and I experimented with upping and downing levels and the dynamic range, and came to the conclusion that for my dull and boring music, what I was doing was sort of OK. Nothing has convinced me a happy LUFs readout is better than the ones where it looks less good? I even experimented with a few recordings I made. I deliberately wrote a piece that had hardly any dynamics, and another that had a really loud section and a really quiet section. Spotify left it pretty much alone. A piece with loud synth bass clearly got tweaked - so this sort of made me realise that even LUFs as a guide gets tweaked if Spotify, Apple and Youtube so desire.

The idea of a sort of constant watch of your levels, and a clever system to produce a magic number is just another system - like when we went from magic eyes to PPMs, but other people went from magic eyes to VUs and had to learn to interpret things differently. I think I shall continue to treat LUFs the same.
 
Last edited:
LUFS = Loudness Units Full Scale.

In recent years its used more often than dBfs.

Streaming platforms specify max levels in LUFS. If you are over their number, they compress/limit your track.

Since dynamic range is so important, consult the specs if you post to a certain streaming platform. YouTube, Facebook etc.
 
I lmfao everytime I hear people say they are leaving -6db or whatever to leave headroom for mastering. I lmfao even harder when I hear of mastering engineers demanding this
 
I suspect it's a bit pointless really, because where we had VU and PPM, this standard lasted a very long time. Now however, the streaming platforms kind of have movable goalposts, so the idea that a contemporary piece of music now, designed to not engage the compress algorithm will not trigger in in a years time is chasing a guess.

Make it sound good, don't go mad with the level, squeezing the last dB, and leave them, to it. You have instantaneous levels and averaged ones. Each system treats them differently, and you go with the flow.
 
LUFS is honestly just filtered and gated RMS. With most sources, unless the final balance is really weird or there’s a lot of gaps of silence, the two should measure pretty close the same. So if you’re used to RMS…

Average levels (RMS, LUFS or even VU) have nothing really to do with gain staging anyway. Gain staging is about finding the appropriate compromise between the noise floor and the distortion ceiling, and distortion doesn’t care about averages.

In the box, it’s mostly about how much distortion do you want out of whatever analog emulator you’ve added. You did want distortion right? That’s why you chose a plugin that distorts rather than any of the perfectly clean plugins available, right? Adjust the gain into that plugin until you get the amount of distortion you want. Meters won’t tell you if it sounds right, and in fact sometimes look very wrong when things are getting very good.

Also absolute levels are mostly meaningless until the final mastered distribution file. Up to that point all we care about relative levels. It doesn’t matter how loud the kick drum is. What matters is how its level compares to that of the snare, and how the drums overall compare to the bass and…. The thing I said about the gain staging compromise is also about signal level relative to noise floor and distortion ceiling, doesn’t matter what those numbers actually are. The integrated average all by itself tells us nothing, and neither actually does the level of the loudest peak by itself. What is informative is the difference between those two measurements. Once that crest factor is “right”, simply normalizing will get you to whichever loudness standard you choose.
 
Back
Top