Headroom and Gain Structure [Advanced Topic]

Tell us how you got that killer synth effect, why you love your sequencer of choice or that you are sick of hearing the same lame preset every time you load up malstrom ¬_¬

Headroom and Gain Structure [Advanced Topic]

Postby Super Alex » Sat Jun 25, 2011 7:05 pm

Probably the most important thing I learned about producing music revolved around one question: how loud should everything be?

A quick story to demonstrate the importance of gain structure. When I first started producing, I didn't touch levels very much. I thought "Hey, I can hear each part, nothing sounds too loud, the levels must be fine!" and kept working. My tracks were alright I suppose, but the tracks would always feel squashed, so to speak, with no room to breathe. All of the space within the sonic world I was creating was completely full of sound.

But then I learned about how you should structure the levels of your music, and things changed for the better. Suddenly, each element of the track had more room to breathe, was clearer (generally, there are a lot of things that affect clarity), and I could make my music seem louder and more professional. Granted, my tracks are still nowhere near professional quality, but they sound much better.

So - what's the secret?!?!

Well, first, we have to look at an important concept called headroom. Headroom is probably the most integral part of this entire lesson, so listen up.

Headroom is the difference between the volume level something currently sits at, and the absolute maximum level it can sit at. The absolute maximum is, on most DAWs, the point where if you exceed that level, you will get clipping.

First, we should answer a basic question: why do we want headroom and proper gain structure? Put simply, your mixes will sound cleaner, bigger, louder (after mastering), and more powerful. If you use up all your headroom, your track will feel cramped and squashed, without room to fill the space in your head where the music plays. Why do your tracks sound so small compared to Kryptic Mind's enormous caverns, or next to Deadmau5's arena sized synths? Easy: they have headroom, you don't.

So. Imagine one, single, lonely volume fader. It runs from -infinity dB up to, lets say, 0dB, just for simplicity (I honestly have no idea what it would be, but I don't really care :6: ). You set your fader to -6dB. So, here's what we have: your fader is capable of pushing the level to 0dB, and you have brought it down to -6dB. This leaves 6dB more volume you can use safely, right? NO. By pulling the volume down on your one track, you reduce the chance of your track peaking and clipping. By having that 6dB of headroom, you open up more room for power, clarity, volume, all that good stuff.

Let's look at that idea some more, in real depth. How does pulling down the level help prevent clipping? Intuitively, you can already answer that; less volume = less clipping. At a basic level, that's true, but there's more going on than that. Let's go back to one fader, but this time all we have is a sample of a guy yelling "Hello!!". If your sample is quite quiet, then you can push your fader to 0dB without worry. If, however, the sample is quite loud, you should pull the level down. Here's why: any sound has a waveform. If you look at soundcloud, you'll see this. In a waveform, you can see the peaks and valleys that make up the sound. If the peaks (loudest bits) of a sound hit the upper limit of your DAW (in our case, 0dB), then you have two problems: you risk clipping, and more pertinent to this lesson, you have absolutely no headroom to work with. What does this mean? Simply, you want to avoid hitting the upper limit on your DAW so you can preserve headroom.

So, how to we make sure our tracks have plenty of headroom? Easy - TURN YOUR CHANNELS DOWN. This is going to be hard for you to do at first. Your tracks will be quiet. Too quiet. You'll want to turn it up. I know you will. RESIST. Volume does not come from mixing. Volume comes from mastering, which isn't something you should worry about. After you've mixed your track, with channels down and plenty of headroom, then slap a limiter on the end of the master signal chain to make up for a bit of volume. That's the best a bedroom producer without mastering experience can really hope for, and you know what? Your track will sound better. Try it out.
JannikR wrote:This is now a thread about Rminidicks womanliness.
Super Alex
Lord of Leads (Solo)
 
Posts: 1288
Joined: Tue Dec 01, 2009 3:40 am

Re: Headroom and Gain Structure [Advanced Topic]

Postby MnilinM » Sat Jun 25, 2011 7:38 pm

Two thumbs up!

I've been needing to know this stuff!
User avatar
MnilinM
Lord of Leads (Solo)
 
Posts: 1044
Joined: Fri Jun 03, 2011 6:23 am

Re: Headroom and Gain Structure [Advanced Topic]

Postby Super Alex » Sat Jun 25, 2011 7:45 pm

Ekol something out that I phrased a bit weirdly, and that I should explain a bit more. I said that volume comes from mastering, and he says it's from good mixing. I'd say we're both correct. If you do what I've been talking about and turn your channels down, then you will have a song that has a low apparent volume. Then when you master the track, there is plenty of room to boost the song's volume. If you have a song that's mixed too loud, then the mastering stage won't have any room to boost it, and the song will have the same 0dB level as a professional song, but it won't have the same apparent volume (which is a whole other topic haha).
JannikR wrote:This is now a thread about Rminidicks womanliness.
Super Alex
Lord of Leads (Solo)
 
Posts: 1288
Joined: Tue Dec 01, 2009 3:40 am

Re: Headroom and Gain Structure [Advanced Topic]

Postby Ekol » Sat Jun 25, 2011 9:01 pm

Super Alex wrote:Ekol something out that I phrased a bit weirdly, and that I should explain a bit more. I said that volume comes from mastering, and he says it's from good mixing. I'd say we're both correct. If you do what I've been talking about and turn your channels down, then you will have a song that has a low apparent volume. Then when you master the track, there is plenty of room to boost the song's volume. If you have a song that's mixed too loud, then the mastering stage won't have any room to boost it, and the song will have the same 0dB level as a professional song, but it won't have the same apparent volume (which is a whole other topic haha).


Perceived loudness is not the same as actual loudness. It's not just about volume which makes a track seem loud, it's how it's mixed and how the elements work together in the actual track that give the track it's perceived loudness :)
User avatar
Ekol
Squire of Sound (Surround)
 
Posts: 500
Joined: Tue Dec 01, 2009 2:36 am

Re: Headroom and Gain Structure [Advanced Topic]

Postby petermennitirecords » Sat Jun 25, 2011 9:29 pm

Protocell wrote:Perceived loudness is not the same as actual loudness. It's not just about volume which makes a track seem loud, it's how it's mixed and how the elements work together in the actual track that give the track it's perceived loudness :)


A lot of this has to do with proper EQ and arrangement - if you EQ out unimportant parts of your sounds, they won't really sound a whole lot less loud in a mix, but you'll actually cut the volume substantially. Likewise putting a slight bump at the part of the sound you want most prominent can bring that bit out without raising the volume of the whole sound and swamping your track with a bunch of unnecessary noise. Just don't go overkill (6-12dB cut is usually good, and I rarely boost by more than 4 dB with an EQ) or you'll get a nasty, unnatural sound.

Arrangement is important too - you can save a lot of headroom by spreading the different elements of your track out across the frequency spectrum. Basically (I think), if you stack a bunch of stuff at say 800Hz, you can only have everything so loud because if you push it too much, you'll clip at 800Hz. But if you have your elements at 600, 800, 1k, and 2k Hz, you can have everything much louder because the loudness will be spread out across the different frequencies. If you're encountering this problem try moving bits up or down an octave or choosing different instruments. Panning things out left and right works on basically the same principle and can also help save headroom.

Feel free to correct me on this because I realized after starting this post I forgot a lot of the really technical stuff I learned after a semester not taking any music courses :P
User avatar
petermennitirecords
Knight of Noise (Pink)
 
Posts: 695
Joined: Sat Oct 30, 2010 9:42 pm
Location: United States

Re: Headroom and Gain Structure [Advanced Topic]

Postby Super Alex » Sat Jun 25, 2011 9:34 pm

Protocell wrote:
Super Alex wrote:Ekol something out that I phrased a bit weirdly, and that I should explain a bit more. I said that volume comes from mastering, and he says it's from good mixing. I'd say we're both correct. If you do what I've been talking about and turn your channels down, then you will have a song that has a low apparent volume. Then when you master the track, there is plenty of room to boost the song's volume. If you have a song that's mixed too loud, then the mastering stage won't have any room to boost it, and the song will have the same 0dB level as a professional song, but it won't have the same apparent volume (which is a whole other topic haha).


Perceived loudness is not the same as actual loudness. It's not just about volume which makes a track seem loud, it's how it's mixed and how the elements work together in the actual track that give the track it's perceived loudness :)


Exactly. Thanks for putting it more eloquently than I :lol:

And petermennitirecords, you're right on the money. Panning can help, EQ helps a lot when you need it, and making sure you don't stack frequencies is VITAL. An article talks about those ideas here: http://www.dnbscene.com/article/88-thin ... tutorial/1 EVERYONE SHOULD READ THIS.
JannikR wrote:This is now a thread about Rminidicks womanliness.
Super Alex
Lord of Leads (Solo)
 
Posts: 1288
Joined: Tue Dec 01, 2009 3:40 am

Re: Headroom and Gain Structure [Advanced Topic]

Postby Ekol » Sat Jun 25, 2011 10:17 pm

You can stack frequencies, or else the music would sound HORRIBLE. What it means is that if something does not need a certain frequency that clashes with others (e.g a hat with a lot of low end) then it should be removed.
User avatar
Ekol
Squire of Sound (Surround)
 
Posts: 500
Joined: Tue Dec 01, 2009 2:36 am

Re: Headroom and Gain Structure [Advanced Topic]

Postby Rmonik » Sat Jun 25, 2011 10:53 pm

Exactly. Even if it doesn't have that frequency (or you think it hasn't) - cut it out if you don't need it. The more room you create, the better. I do kinda oppose to the idea of having a lot of headroom left for mastering. You create headroom in the process, to be able to put more stuff in there without it clashing with eachother and making stuff sound muddy. Mastering is really just about filling the last tad of unfilled headroom. The most important thing is giving every element it's place in the song. (in the box, if you read the EQ tutorial) Some of the bigger names don't even master at all.
Image
User avatar
Rmonik
Le Facetious Lemon Key Lord
 
Posts: 4429
Joined: Sun Mar 20, 2011 12:59 am
Location: Ghent, Belgium

Re: Headroom and Gain Structure [Advanced Topic]

Postby JannikR » Sun Jun 26, 2011 12:00 am

Protocell wrote:You can stack frequencies, or else the music would sound HORRIBLE.


This.
I'm getting tired of hearing that you can't stack frequencies, of course you can. As long as you don't overdo it so that it clips.

Also, I'm pretty sure the reason that artists don't master is that record companies usually have a sound engineer who does that. :)
http://soundcloud.com/phaselabs/your-explosion-my-mind-e
Phaeleh wrote:Yeah man, I can write in all other software (except logic, which is the work of satan)
JannikR
Lord of Leads (Solo)
 
Posts: 1255
Joined: Thu Aug 05, 2010 4:36 pm
Location: Skövde, Sweden

Re: Headroom and Gain Structure [Advanced Topic]

Postby Ekol » Sun Jun 26, 2011 12:14 am

JannikR wrote:Also, I'm pretty sure the reason that artists don't master is that record companies usually have a sound engineer who does that. :)


This.
User avatar
Ekol
Squire of Sound (Surround)
 
Posts: 500
Joined: Tue Dec 01, 2009 2:36 am

Next

Return to Computer Music

Who is online

Users browsing this forum: No registered users and 1 guest