Can someone explain -18 db mixing?

[XxSicRokerxX]

So I read a few people say that recording and mixing at -18db helped them significantly. What exactly does this mean? Crank your monitors and mix w the master bus being at -18db? Or leaving the master bus alone, and having the individual tracks at -18db? I understand the concept, its too leave headroom because as you mix and process tracks loudness increases (potential risks of clipping) amirite? Can someone explain to me exactly how to do this and why?

Thanks guys

 

[colonel kurtz]

to keep it simple, -18dbfs in the digital realm is equal to 0dbfs in the analog realm. this means that if you record a track into your DAW that peaks at -6, it's +12 in analog...aka clipping like fuck

 

[Uros]

So I read a few people say that recording and mixing at -18db helped them significantly. What exactly does this mean? Crank your monitors and mix w the master bus being at -18db? Or leaving the master bus alone, and having the individual tracks at -18db? I understand the concept, its too leave headroom because as you mix and process tracks loudness increases (potential risks of clipping) amirite? Can someone explain to me exactly how to do this and why?

Thanks guys

No, leave your 2bus untouched. What it means is - have your tracks sitting around -18dBFS before you process them with plugins, and have all channel and bus (including 2bus) left untouched at zero.
In analog world idea is/was to have your tracks around 0 VU (+4dBU), which, depending how your converters are calibrated (two most often settings are -18 and -20dB), translates to digital realm to have your tracks around -18dB RMS (not peak, although quite a few guys here think it's like that; having tracks peak at -18 is not wrong per se, since noise floor in 24bit audio is so low, but it is not necessary, and in this case is misunderstood). That could mean that your tracks should peak around -6dBFS or something similar, it depends, it's not linear (for ex. a snare track will have higher peak level comparing to distorted guitar track, even though their RMS level is the same).

Every once in a while a thread like this yours pops up, but nevertheless - have your tracks peak between -12 & -6dBFS (if a project was not tracked by yourself, set all tracks to sit between these two values before any further processing and balancing; if you don't have trim knob on channels in your DAW of choice, use a trim plugin like FreeG for that). Try to maintain that level throughout whole plugin chain on every track. Meaning, if you for ex. hit a compressor with like -6dB, try to set its output so that it gives you similar dB value after the signal has passed through the compressor (use your ears for that). The same goes for ever other plugin on the track. Now you don't have to be anal about that, just don't clip input any of your plugins. I use FreeG plugin for very this purpose, if a plugin doesn't have output knob, like some compressors. Try and maintain roughly the same level throughout the whole chain.

Don't use compressors to increase loudness, use them to alter dynamics and envelope of your audio. While I am at it, don't use any of your plugins to increase loudness (we're talking about mixing here). There are channel faders in every DAW for a reason. Use them for level balancing between tracks.

Even though the only thing that ultimately matters is to not clip your DA converter, most of VST plugins work best around -18/-20 dB RMS, so try and use that to your advantage.

edit:

to keep it simple, -18dbfs in the digital realm is equal to 0dbfs in the analog realm. this means that if you record a track into your DAW that peaks at -6, it's +12 in analog...aka clipping like fuck

el colonelo, you are wrong. Read my post. You are mixing up RMS and peak. And there is no dbFS in analog realm anyways, it's strictly the digital 'thing'.

 

[Nimvi]

+1 to everything Uros said.

Another added advantage is that if every track sits around the same level with the fader on 0, you get to use all your faders in the region where they have the highest resolution, as in, you can make very tiny changes much easier around the 0-mark of your fader, than when it is at -30, where small movements make much bigger changes.

 

[Uros]

Yeah, that too, which is critical for fader riding when you do automation.

edit: I'd just add this to OP - buy Satson if you can. It has a VU meter. Load up an instance of it on every track, and set gain on every track to hit around 0 VU in Satson, and you'll see how you'll improve your mixes only from that.

 

[Tommy Evans]

^^^did I read correctly that, when mixing with this -18dbs rule in mind, you want the fader at -18? OR the fader can be anywhere as long as its audio signal is peaking around -18 on the 2bus? ...that's the only thing I'm hung up on.

 

[XxSicRokerxX]

Thanks guys but holy this is like reading chinese :/

 

[pikachu69]

^^^
no. the avereage RMS of your 2buss should be round -18dbs with the fader at 0.

 

[XxSicRokerxX]

^^^did I read correctly that, when mixing with this -18dbs rule in mind, you want the fader at -18? OR the fader can be anywhere as long as its audio signal is peaking around -18 on the 2bus? ...that's the only thing I'm hung up on.

I believe what he meant was that all faders should be left at 0, and ideally you want the source signal at -18db (or -6 to -12) ?

 

[Uros]

Yeah, faders at zero before any mixing, your track levels should sit between -12 and -6 dbFS peak.
Eh, I honestly don't see any way I can make it any more concise than this. :/

 

[XxSicRokerxX]

Wait, so is it recommended that they are at -18 db? or around - 6 to -12 db?

 

[Uros]

-18db RMS (average), which in effect will have your tracks to peak between -12 and -6dbFS. Since DAW meters are mostly peak meters, tweak levels so meters read between -12 & -6. Got it?

 

[XxSicRokerxX]

Oh okay, I completely understand now, in other words try to maintain around -6db to -12db (-18db RMS) before and even after mixing "Try to maintain that level throughout whole plugin chain on every track. Meaning, if you for ex. hit a compressor with like -6dB, try to set its output so that it gives you similar dB value after the signal has passed through the compressor"

And do the same for every track.

 

[Uros]

 

[Nimvi]

Hehe welcome to the world of decibelscale confusion But looks like you've got it now SicRoker.

If all of this seems confusing now, just give it some time to sink in and read about it again in a while. At some point it will just click in your head, and you will develop an understanding of the reasoning behind it naturally.

The thing with all these little "rules" is that they don't come out of thin air. They are often the result of trial & error of many people over long periods of time, and people tend to reach similar conclusions after these phases, because they find stuff that generally works well as a startingpoint for everything. It is important to keep in mind that none of these things are set in stone, but a deep understanding of their origin can help you to bend them to your will and break creative barriers.

 

[colonel kurtz]

el colonelo, you are wrong. Read my post. You are mixing up RMS and peak. And there is no dbFS in analog realm anyways, it's strictly the digital 'thing'.

guess i should have specified between dbfs and dbvu...but yea, i stand corrected. this is one of those areas where i learned the shit years ago, verified my converters were properly calibrated, and now just make sure the preamps don't hit the red and all is good.

it's kinda like when someone asks you to recite a phone # that you call on a daily basis, and your fingers know what buttons to push but your brain goes :zombie:

 

[The-Zeronaut]

ok...so...correct me if im wrong...

with every fader at 0 the signal must peak at -18dbfs (average) (in the recording process...and if not...using trim gain plugins before even start mixing)
and it must be at that level before and after every plugin in the fx chain...

But how am is suppose to do it with busses and the master fader (aka 2-buss)?

 

[colonel kurtz]

(in the recording process...and if not...using trim gain plugins before even start mixing)

if the levels of the tracks are hitting way above -18 already with the fader itself at 0, you aren't going to gain anything by turning down the trim gain

honestly, the title of this thread - "-18db mixing" is misleading, as the entire issue really applies to -18db TRACKING. pretty much all preamps/consoles/hardware/whatever are made to perform their best at 0dbvu, which roughly translates to -18dbfs in the digital realm. if your shit is peaking close to 0dbfs in your DAW with faders at 0, it means the audio's been recorded way too hot, which is likely going to decrease dynamic range and raise the noise floor in a best case scenario, or just straight up clip the fuck out of the waveform. either way, your mix will be fucked from the start, regardless of the quality of the signal chain or your mixing capabilities.

it must be at that level before and after every plugin in the fx chain...

i could be wrong, but i personally don't trip on the levels of the track coming out of and going into plugins, or the level of the 2-buss. as long as you aren't coming in too hot on the way in - thus pushing preamps and converters beyond their optimal range - you should be able to have whatever track peaking at whatever level, providing it isn't being clipped to shit. again, what's really important is that the RAW audio is at the correct levels!

 

[Uros]

ok...so...correct me if im wrong...

with every fader at 0 the signal must peak at -18dbfs (average) (in the recording process...and if not...using trim gain plugins before even start mixing)
and it must be at that level before and after every plugin in the fx chain...

But how am is suppose to do it with busses and the master fader (aka 2-buss)?

Use channel faders (channels that feed a group) for that. If you want to 'save' them for automation needs (meaning so they stay close to zero so you could have the highest possible resolution for fader riding), use trim plugin (in Cubase there are trim knobs on group tracks too). The idea is to have your 2bus around -18dB RMS too.
___________________
edit: @colonel kurtz:
well, If I don't track a project (if I am, I try to have good levels from the get-go), all I ask from clients is that tracks don't peak over 0 dBFS. There is no damage to audio if it peaks at say -1dBFS, you just lower it with the trim knob. That's in my experience.

Having plugins on unity gain has one more benefit to it aside from maintaining good levels - it is also good because your mixing judgments are not badly influenced (for ex. louder=better sounding).

 

[colonel kurtz]

i've never had a 2-buss in the digital realm come in anywhere near -18 with faders at 0. on the contrary, IME a bunch of tracks recorded to peak around -18 will sum to somewhere between -3 and 0 on the 2-buss, depending on the # of tracks and whatnot.