-18db messes with my head visually

Starting a new project and I'm really focusing on my levels. I'm working toward that average -18db recording level, but when I look at the track in logic the waveform looks so small. Visually I alway think I'm recording too low but I know I'm not.

Does anyone look at how small the waveform is and have is visually mess with your head?

Have you tried the waveform zoom?

I think the -18dB rule is way overrated.

I shoot for -12 to -10, which is where the Logic meters turn yellow. Gives me still enough headroom to avoid peaks.

As long as the interface is not peaking on the input, meaning the converters don't clip, you should be fine.

I think the -18dB rule is way overrated.

Overrated is an understatement.

There's no such thing as the -18 dB "rule". There's only a -18 dB "myth". And as you have realized, ManFromNappa, producing at those low levels comes with its share of inconveniences such as tiny waveforms and low meter levels with coarse metering resolution.

It doesn't really matter at what level you produce as long as you don't create distortion. So stay below 0 dB FS at all points in the signal chain. If in order for you to make sure you stay below 0 at all points it's easier to generally try to not peak above -6 dB FS then do that. If you peak above - 6 dB FS it doesn't really matter though.

In the final analysis, it's more important in my opinion that YOU understand how your material works, how it can potentially create distortion, and for you to deduct your workflow from those facts and your personal experimentation, than to blindly follow a workflow recommended by someone else (including me).

There's no such thing as the -18 dB "rule". There's only a -18 dB "myth".

I think of it as more a "helpful guideline" for those that aren't sure, more than anything. It gives a healthy signal, and removes the worries of overloading the recording in pretty much most circumstances.

When you are a bit more experienced, you can better make decisions about how loud your recording might be, and set your levels appropriately, still giving yourself decent headroom and removing the need to worry about the meters while recording.

And if your projects are super heavy on the track counts and you are constantly running out of headroom on the mix bus, then it might help to reduce your general tracking levels.

But yes, not a rule at all. The main thing is to not need to track like we used to in the 16-bit convertor days of the 90s, and tape before that, where these practices were standard - where you'd record as hot as you could to optimise the signal, without clipping, which is an outdated and unnecessary practice, and has been for a couple of decades now...

In the final analysis, it's more important in my opinion that YOU understand how your material works, how it can potentially create distortion, and for you to deduct your workflow from those facts and your personal experimentation, than to blindly follow a workflow recommended by someone else

Exactly.

I think of it as more a "helpful guideline" for those that aren't sure, more than anything. It gives a healthy signal, and removes the worries of overloading the recording in pretty much most circumstances.

To me it's akin to suggesting that when entering a building without knowing the height of the door opening, you close your eyes and bend down to lower your head. That removes the worry of hitting your head against the door frame in most circumstances. I don't like that kind of thinking because:

1. In most situations, bending down is unnecessary because most doors are tall enough.

2. In some situations where doors are really small, you may still hit your head.

3. Bending down all the time is uncomfortable.

4. A better workflow is to open your eyes, look how tall the door is, and determine if you have to bend down or not.

Going back to our levels in digital audio workstations:

1. In most situations, staying at -18 dBFS is unnecessary.

2. In some situations, you may still distort even when using -18 dBFS signals.

3. Working with small waveforms and lower level meters is uncomfortable.

4. A better workflow is to know your audio signal level at any point in the chain and know you're not clipping. For example maintaining a reasonably general level of, for example, peaking mostly below -6 dBFS and making sure that no process applies unnecessary gain.

Completely agree, people really should understand, and learn about what they are doing. But there are always people that don't want to understand, or learn, they just want a quick recipe to follow, so they can forget the engineering and just concentrate on the music. And I would tailor my advice accordingly in those situations.

I certainly don't follow a general -18dB principle myself, and nor I suspect do most people who know what they are doing - they are able to deal with the variables and set things to their needs correctly. And actually, I don't even think I offer that -18dB level advice that much, and it's usually qualified, for all the reasons David states.

However, I *do* use the Waveform zoom all the time, and that is a super useful feature I *do* recommend using to aid visual navigation around your projects...

And actually, I don't even think I offer that -18dB level advice that much, and it's usually qualified, for all the reasons David states.

I think the -18 dBFS comes from the analog world, which was a whole different ballgame To do the same.when producing inside a DAW is completely arbitrary. If you're going to go with a ballpark number to try to generally stay safe then -8 to -6 dB FS makes more sense IMO, but even if you do that, you have to understand that it's not a rule, it's just a cautionary measure with no guarantees, so it doesn't make much sense to turn it into a religion and spend hours creating a template for it or gain staging everything to reach these (or any) precise specific levels.

I can't find it right now but I remember a video with Andrew Sheps who was saying basically something like "Look, whatever makes you more comfortable, and if gain staging for hours to stay within a certain realm of levels makes you feel more safe and more comfortable and ultimately mix better because of that then by all means, go ahead, but if you spot one of your channels going to -13 dB and you think it should be -18 and start reworking the entire mixer to go to -18 then stop! Stop right now! What are you doing???" ... or something like that.

And what does plugins expect? Like amplitube and other simulators of analog gear? I’ve read somewhere that they have been optimized for -18 dB, but if that’s also just a myth then…?

Well first what do you (or the people who wrote what you read) mean by "optimized for -18 dB"? Is that an average level? A peak level? RMS level? VU-meter level? How does that work with highly dynamic signals? Signals with heavy transients?

Every plug-in is designed differently. Most plug-ins don't expect anything specific, and in Logic's 32 bit floating point, can handle levels way above 0 dBFS (not that I would recommend that as a workflow, but it's important to know that it would not create distortion).

Amplitube is a guitar amp emulation, so the amount of gain created in the preamp stage varies depending on the level of the signal you feed it. Other simulator of analog gears don't all work the same, so it all depends on how each specific plug-in was designed. Some engineers like to push a little more signal into an analog mic preamp, a compressor, an EQ, a guitar amp, etc... in order to get a nice saturation going. Some analog plug-ins may be modeled after that as well, so that you won't get the desired sound if you feed them a level that's too low. So with any level-dependent designed plug-in, it becomes a matter of experimenting, listening, and dialing in the sound you want, but certainly not sticking to any specific level as a rule.

Yeah I think we can safely conclude that it’s not an exact science Some people religiously stick to hit around -18 dB. Others state a different target and others say that as long as it sounds good…

I think I’m just aiming for less degrees of freedom. So any rule of thumb helps. Every time you have another gain stage/knob in your signal chain you have more to tinker with. If it was all just linear it wouldn’t matter, but since it’s not it’s like stacking 7 billiard balls.

But obviously if the rule of thumb is just a myth then I might as well just design my own rule and fix that knob there.

And actually, I don't even think I offer that -18dB level advice that much, and it's usually qualified, for all the reasons David states.

I think the -18 dBFS comes from the analog world, which was a whole different ballgame To do the same.when producing inside a DAW is completely arbitrary. If you're going to go with a ballpark number to try to generally stay safe then -8 to -6 dB FS makes more sense IMO, but even if you do that, you have to understand that it's not a rule, it's just a cautionary measure with no guarantees, so it doesn't make much sense to turn it into a religion and spend hours creating a template for it or gain staging everything to reach these (or any) precise specific levels.

 

I can't find it right now but I remember a video with Andrew Sheps who was saying basically something like "Look, whatever makes you more comfortable, and if gain staging for hours to stay within a certain realm of levels makes you feel more safe and more comfortable and ultimately mix better because of that then by all means, go ahead, but if you spot one of your channels going to -13 dB and you think it should be -18 and start reworking the entire mixer to go to -18 then stop! Stop right now! What are you doing???" ... or something like that.

Does all of that (unnecessary?) -18dB (rule/myth) you are debating also apply to Dolby Atmos?

WOW! Thank you to everyone who commented. I'm going to need to read this thread a few times to absorb everything in it. I never worried too much about my levels until I decided to test out UA's Luna. I have an Apollo interface and a friend who was RAVING about Luna. (Spoiler Alert: I wasn't won over.)

Luna has an "OL" light that goes off if ANYTHING goes above -18db. Track or plugin. So I really had to spend time digging into gain staging when I was playing with Luna. My impression was that it's a nice feature, but the implementation of it doesn't really apply in the digital world. This seems to be supported by this thread.

Luna has an "OL" light that goes off if ANYTHING goes above -18db

I have an Apollo myself, but wasn't too hooked on Luna as well. And that OL light would drive me crazy, haha!

Does all of that (unnecessary?) -18dB (rule/myth) you are debating also apply to Dolby Atmos?

A Dolby Atmos mix/master must be at or below –18 LUFS (Integrated Loudness).

The True Peak level must not exceed -1 dBTP

A Dolby Atmos mix/master must be at or below –18 LUFS (Integrated Loudness).

 

The True Peak level must not exceed -1 dBTP

Thanks Simon.

So for the OP keep in mind that even the Dolby Atmos levels mean your tracks and your master output channel would be peaking way above -18 dBFS, so your Luna overs would light up all the time anyway.

I think there's some confusion here about the -18dB "rule".

The theory goes that your average levels in analog-land used to hover around 0dBVU, which is equivalent to -18dBFS in digital land. AVERAGE levels. And many plug-ins are designed to emulate hardware from anlog-land. So having your AVERAGE levels (not peak!) around -18dBFS - and therefore your peaks will vary from -14 to -8 or so depending on how dynamic your signal/instrument is - makes a lot of sense, I think.

So having your AVERAGE levels (not peak!) around -18dBFS - and therefore your peaks will vary from -14 to -8 or so depending on how dynamic your signal/instrument is - makes a lot of sense, I think.

Yes, there is definitely a lot of confusion which is why I am trying to bring precision to the topic in indicating that the goal certainly is not to follow a rule of keeping your peak levels below -18 dBFS. So that's a good start. Still, we can go further with this.

Now that we agree that the rule has to d with keeping your average levels around -18 dBFS ... well I still do not agree with that. You suggest that peaks will vary from -14 to -8... well that depends highly on what you're recording or producing and I can think of examples where the peaks could stay below -14 or go way above -8 and end up clipping. But even if in your situation, that is what you're experiencing, then what is the reasoning for never peaking above -8 dBFS?

Why should we use the same rule for all instruments, and all music genres, independently of each instrument or mix's crest factor? In digital recording, should highly dynamic classical music be recorded at the same average level as highly compressed pop music or heavy metal? And if yes, why? What is the reasoning? Should they be recorded at the same peak level? And if yes, why? What is the reasoning? Does it even matter what peak level they reach, as long as it stays below 0 dBFS?

i have recorded classical orchestras and jazz bands using digital recorders. In some cases if I had set my average level to -18 dBFS, I would have completely clipped the recordings. In fact I had to set the average level below that. In fact in my young years, I remember doing a session with Archie Shepp once. I had carefully set my recording levels to ... whatever was deemed to be the correct average recording level according to standard gain staging practices back then. The engineer was happy. I was happy. The engineer walked out for a phone call. I sat down behind the console and pressed record. The band started to play. And play. And play. They were completely improvising, nothing was written, no one knew for how long they would play (and I would record), not even Archie Shepp. At some point the levels start going up. And up. The band got excited, the climax was approaching, they pushed it further, and further... and I started seeing the meters go up and up, ever closer to the red... what was I to do? I took both my forearms, placed them over the top part of all the faders, and slowly pulled them all downward together by a good 10 dB or so. Not exactly a technique they teach you in recording school. But hey, I saved the day, the recordings did not clip, and the engineer was happy. The band never noticed anything.

So in that spirit, in my opinion, the most important rule when it comes to levels or gain staging (or just about anything) is, use common sense. And if you decide to follow a rule, make sure you know exactly why that rule exists and why you decided to follow it.

In conclusion:

1. There is no reason for blindly following a -18 dBFS average level rule when recording music in the box with a DAW today.

2. There are multiple reasons for not following that rule. One was raised by ManFromNappa in his original post in this thread, which is that you'd be working with tiny waveforms making audio editing impractical, and low level metering which is uncomfortable and imprecise. Another was pointed with highly dynamic material that could very well clip when recorded at that level.

At some point the levels start going up. And up. The band got excited, the climax was approaching, they pushed it further, and further... and I started seeing the meters go up and up, ever closer to the red... what was I to do? I took both my forearms, placed them over the top part of all the faders, and slowly pulled them all downward together by a good 10 dB or so. Not exactly a technique they teach you in recording school.

Priceless!

Man, talk about some cold blooded decisions. But that's the reality of recording real musicians.

Good tips here aswell

and this is the problem with musicians in general; they have no built-in limiters...

David Nahmani has already said all that is needed to be said in "It doesn't really matter at what level you produce as long as you don't create distortion. So stay below 0 dB FS at all points in the signal chain."

For 20 years I worked as a recording engineer, as well as a mix engineer which I still do, and I learnt that with digital recording sticking to an absolute peak level is pointless.

What is much more beneficial is learning at what levels individual tracks need to roughly at, so when combined the master output has a decent amount of headroom. So when recording a band for example, I want the master output to reach approx -6 dBFS at the densest point of the arrangement. This means recording individual tracks in the region of -10 dBFS.

This gives several advantages

- visible waveforms for easy editing and easy to read metering

- any insert plugins will be in a good operating range

- mic/line preamps, EQ, compressors have a healthy amount of headroom when recording

- headphone mixes are super easy to achieve

- you can very quickly get a rough mix with minimal effort

This last one is something I learnt from top engineers whilst working as an assistant engineer in London. Create your mix as you record, so with faders at, or close to zero you get a really good balance - set levels at the input - faders close to zero are in a sweet spot for operation. Keeping to the suggested levels absolutely helps with this, and if a recorded part is meant to be lower in the mix, don't be afraid to commit to a slightly lower recording level so it is in context with the fader around zero..

If things are getting a little hot on the master as you progress, apply clip gain to pull everything down a dB or 2, or 3.

This isn't brain surgery, so in addition to above, if something feels good by slamming the input to create clipping distortion or recorded at a super low level to be gained up in the DAW to create a lo-fi effect, then do it.

The best quote yet, and another from David which I confess I am going to steal To paraphrase, digital recording is like walking through a door, as long as you don't bang your head all is good. I would add, unless you want to smash some motherfucker on that door frame : P