gain structure and mixing aka THE MONEYSHOT THREAD

[Depone]

Im doing a series of dubstep / Logic tutorials
Make sure you watch them on youtube on full screen 720p HD for better clarity.
Hope you like em!

Heres episode 1 - Drum processing

Logic Pro - Dubstep Production episode 1 - Drums Processing Part 1


Logic Pro - Dubstep Production episode 1 - Drums Processing Part 2

[jsills]

this thread is gold

[icebagg]

Nothing to do with mastering. You're completely missing the point

This is mixing It is the simple fundamentals of gain structuring for building a good mix. Whatever comes later, comes later - level boosting and what not. But I 100% guarantee (and I don't do that very often) that your mixes will improve with consistent, sensible gain structuring.

As a side effect, mastering (or level boosting for DJ play etc) becomes easier, and will yield better results because the mix is better built. It's as simple as that.

FWIW 'If they don’t compress + limit loads then things will sound weak.' is so terribly wrong. Quiet yes, weak no. Do some level-matched comparisons of heavily limited tracks vs their clean counterparts and tell me which is weaker

POWER comes from a good mix. Get the mix right and the level will take care of itself.

ABSOFUCKINLUTELY! glad to see people know what theyre talking about, all to often people talk about mixing....as mastering, its hard to communicate with people cos you never know, i would recommend mixing at a low volume i do this and it really helps when you listen to the finished product, its just that much easier to hear what the levels are at, also making sure nothing goes over 0db (given) either way, much thanks guys!

[Oscar BrockWild]

Sorry if it isn't really answering your question, but -3 is way too high.

Remember that 6dB is half. So if you have one element at -6, that is half your headroom gone. Two elements at -6dB each = all your headroom gone. Having the drums at -3 will leave you fighting against clipping and struggling to keep everything down and under control.

Rather, set your drums for *around* -8 / -10 (ie, a bit less than half). The bass - if we are talking a pure sine sub - would probably sit best a dB or two below that, any distorted/fullband bass sounds should be effectively treated as different entities and mixed appropriately (due to Fletcher Munson).

This leaves you with a few dB headroom, and everything else is just parsley. No more fighting anything, you *will* get repeatable and consistent levels in your mixes, and better mixes as a result.

is it different for pressing to vynil then? Ive been told by the press I was planning to use that if i keep all my levels to -3db bthen the engineer who sets the master (as in the master vynil not the studio master) would have the right amount of levels to tweak....I'm really struggling to work out what the optimum masterimng is if i just want to get my music onto vynil.

cheers if u can help anyways

[macc]

They do tend to have level/volume controls, yes

If you want your music to sound good on vinyl then don't smash stuff with limiting, at any stage. Make it sound good, natural, leave dynamics natural etc etc. That goes for channels, the master, whatever. Do you what you need to do to get a good fat mix, but don't smash stuff way over the top.

One way of looking at it could be to say that vinyl is a better limiter than most limiters are

[zitanb]

This thread is gold and I'm still on page one Big Up.

Rather, set your drums for *around* -8 / -10

EDIT: Read through all 24 pages. Brilliant. Thanx Macc and Other contributors. Appreciate it!!!!

[CDR]

Hi
Im using cubase sx3 and various plugins..
When i export my tune as a 24bit wav it sounds fine but when i do as an mp3 there are click sounds at various points in the track...any1 no why???
If i open my 24bit wav in sony sound forge and export as mp3 it sounds fine....
Could this be a cubase problem?
Thanks
CDR

[therapist]

OK, I'm wondering very roughly what the spectral analyser should look like on the whole track at full flow. A basic mixing concept is that you want to fill all the frequency ranges yes?? Because my analyser tends to show about a billion db of sub-100Hz compared to everything else.

I realise that's a retarded question because a) every track is different b) they're probably not completely accurate c) use your ears you n00ble etc. etc.

But that assumes that my speakers and my room are perfect to judge my tunes on. I've never played anything on a club system, obviously I reference my tracks to similar ones and they sound ok to my ears. Is it worth taking the equal loudness curve into consideration? And would looking at a fully mastered track running through the same analyser be of much use for comparison?

[symmetricalsounds]

would looking at a fully mastered track running through the same analyser be of much use for comparison?

yes

[skimpi]

Hey, I started reading this about keeping all level low etc, and leaving the loudness for your mastering engineer and stuff.
Well I was just wondering what the difference between having all your levels low and then bouncing, or having them high but no channel is in the red, but the master is in the red, but if you pull the master fader down so nothing is in the red and then bounce is this just the same as having your levels low?

Also it has been mentioned that you should have a good mix, and then give it your ME for loudness, but just limiting it afterwards for loudness is ok right? Because i normally mix it fairly loud, then bring the master down so its low, and then put it back in Logic a and gain and a limiter to make it louder. Is this good practice?

Cheers

[grooki]

Hey, I started reading this about keeping all level low etc, and leaving the loudness for your mastering engineer and stuff.
Well I was just wondering what the difference between having all your levels low and then bouncing, or having them high but no channel is in the red, but the master is in the red, but if you pull the master fader down so nothing is in the red and then bounce is this just the same as having your levels low?

I'm no mastering engineer, but as I understand it a good way to think about it is to think of the way a distortion or overdrive unit works. The first part turns the volume up until it distorts (so way into the red), the second part turns the signal volume down so that you now have a distorted sound at the same level as the other elements in the track. Turning down the distorted signal did not remove the distortion.

Also it has been mentioned that you should have a good mix, and then give it your ME for loudness, but just limiting it afterwards for loudness is ok right? Because i normally mix it fairly loud, then bring the master down so its low, and then put it back in Logic a and gain and a limiter to make it louder. Is this good practice?

so considering that the signal shouldn't clip before the master this probably isn't the best way to go about it. You should have a well balanced mix which is probably going to be quiet, and then give it to the ME or if you aren't then you can limit it yourself.

All this I learned from this thread. You did read the thread?

[macc]

Hey, I started reading this about keeping all level low etc, and leaving the loudness for your mastering engineer and stuff.
Well I was just wondering what the difference between having all your levels low and then bouncing, or having them high but no channel is in the red, but the master is in the red, but if you pull the master fader down so nothing is in the red and then bounce is this just the same as having your levels low?

This has all been answered in the thread already, but once again;

In a modern sequencer there's no appreciable difference between the two methods you describe. They work internally at 32-bit floating point, giving preposterous amounts of headroomabove 0dBFS. However when it hits the 'real world', ie your master fader/DAC, it has to be converted to 24 bit, meaning anything over that gets chopped off, ie clipped. So turning it down on the master fader brings it down under 0dB = no more problem.

BUT

Working the way you describe is, IMO!, symptomatic of poor gain structure and doesn't lend itself whatsoever to repeatibility. You're moving the goalposts all the time. Ending up with the master fader at -24, needing your drums a bit louder in the mix but the drum fader is all the way up already, opening the first plugin on that channel and turning it up, then realising it is hitting the compressor differently so you have to nudge that, etc etc etc. Sound familiar to anyone?

Get the sound at the right level from the start of the signal chain - in an ideal world the samples are at the right level and the tune is balanced without clipping, no plugins, no faders moved (see jazz band analogy on page 84686234). Everything else is a step away from that, involving more processing, and most importantly, more fannying about when you have better things to do.

To put the point I have been trying to make in this thread (yet) another way; why move 5 knobs when 1 - or even none - will do?

By minimising the variables you're dealing with, you get good mixes, increasingly often - ie repeatibility. This thread is, for me, about becoming better at mixing, not about making one single better mix. Better at mixing = lots of better mixes. Give a man a fish and all that.

[skyh]

Anyways, so I have a question about this. I know you said that some of the sub harmonics will be in the missing zone of bass from roughly 100 - 200 hz, but do you feel that the sub and main lines will still sound nice and beefy with this zone removed?

It depends.

It depends how you cross over, innit How you approach this area depends on the content you're dealing with. These super-rich distorto-basses everyone on this forum seems to love will have loads of content there that overwhelms the true sub content. So it will need more taming. An 808 might not need any.

Besides which, remember that filters aren't brick walls, and the steeper a filter is the more it rings, more phase distortion it causes etc etc. What I am getting at is that I'm not saying about using 58 HP filters to get a brickwall filter with no content in that area. Maybe a 12dB/oct on the LP and HP, gentle enough not to fuck stuff up and also shallow enough to allow some of that content through - not dissimilar to an eq. If need be you can make them sharper (I'd probably lean towards GlissEQ or something with a variable slope) to eliminate a bit more.

I used those numbers an example, but like I say, worrying too much about the numbers or approaching everything with the same technique leads to ruin. Personally I'd eq out that area a little if it needs it, and if it REALLY needs it (and I am doing other stuff to the upper band) I'll do the crossover thing. Or not. I'm not going to chop anything out if it doesn't need it. Whatever. It depends


.....


I'm going to apologise in advance here. You've stumbled on something that has been bothering me for a while, it's not your fault, and it' nothing to do with you.... but I feel a rant coming on

The point of this is to leave room for your kick drum, but if your kick is only on the one and three there are large spots in between the kicks where this frequency of bass will be left out...

SO FUCKING WHAT?!?? Why does everything have to be full all the time? What the fuck happened to people's concept of SPACE in their fucking music?

So basically I'm asking do you find this method better than side chaining your kick to the sub line?

This shit, I mean, I think I went on holiday or just didn't read enough forums or something, but suddenly sidechaining is this saviour of all shit. Bollocks. It's a tool, and it can be handy. But sidechaining all this shit to all everything else to shoehorn another element that doesn't NATURALLY fit into the mix in... I mean, it has a place in certain styles and all that but FFS, it's a last resort or a creative effect. Pick the right sounds and they coexist, naturally, easily, beautifully.

Sidechaining leads to this awful situation where you have no SPACE in the mix, everything is constantly pushing other things out of the way, everything rushing in around everything else and back and forth like some sort of sonic fucking quicksand and all at the cost of SPACE. Or, rather, because of the absence of space. They're not coexisting, they're fighting for room. It's not natural. I say again, it can work, and it can be useful creatively, yes. Sometimes, just occasionally, it is the only option. But I think before I started hearing about it absolutely everywhere I had used it on two mixes in about 86 billion.

It's bandied about as an easy way to make things work together in the mix, but the fact is, plain and simple, it's sticking plaster, a bandaid or whatever you call it where you come from. It increasingly being used as a cover up for lack of sound selection/arrangement skills and/or the lack of eq ability. It makes up for the fact that these people - newbies a lot of the time - don't even know how to choose sounds that work naturally, and they're compounding their sonic misery by throwing more processing after bad decisions. There's too much shit there, and not enough room.


Those bits where there are 'gaps' (isn't it terrible!) are what lets your tune breathe, have air or space or whatever. Everything has its own place, and they all fit together to fill the spectrum and soundstage. Everything does its own job, they all fit around and together and work with coordinated independence, and when they are resting there is space, and not all this other shit rushing in to fill it because of some misguided notion that everything has to be on all the time. Everything all the time is tiresome. As Thelonious Monk said, what you don't play is just as important as what you do play.

Now go and listen to Kind of Blue, and think about this while you're listening.




Right... sorry about that. Back to polite mode.

Hope this helps

GENIOUS. I'm one of those people that normally squashes the shit outta my sounds. Crap EQ, crap mixing.. blah blah blah. But, today when I get home from class I'm going to use your first post to make something natural. So fucking brilliant, man. Lovin' this thread right now.

[macc]

I must have been getting loads of uber-sidechain mixes that week Bit harsh

[Metaphor]

Okay, okay.

SO, I've been reading this thread for a while, kinda stopped keeping up over the last few pages, but, I may have a new question for you here:

I write and mix in 2 stages:

1-I write in Renoise or Ableton and then Bounce to.....
2-Logic for mixing.

WHEN/WHERE should I be making sure the faders are at -10? In renoise/ableton before I bounce ( I suspect this is so), or is it better to have a greater volume (without clipping) going in to logic and then render from logic at -10 ?

Your help will be greatly appreciated!

[stappard]

Okay, okay.

SO, I've been reading this thread for a while, kinda stopped keeping up over the last few pages, but, I may have a new question for you here:

I write and mix in 2 stages:

1-I write in Renoise or Ableton and then Bounce to.....
2-Logic for mixing.

WHEN/WHERE should I be making sure the faders are at -10? In renoise/ableton before I bounce ( I suspect this is so), or is it better to have a greater volume (without clipping) going in to logic and then render from logic at -10 ?

Your help will be greatly appreciated!

I'm not sure where you've got this magic -10 from... well you got it from this thread but Macc has been insistant that this isn't an exactly science, rather a set of guidelines.

Anyway, given that Macc does actually have a life outside this thread (presumably) I'll have a go at answering in the hope that he then won't have to.

Your post was slightly unclear so there are two possible scenarios that you may be referring to, but luckily my advice is the same for both.

Firstly, if you mean you're bouncing a master from renoise/ableton and mixing/'mastering' in logic, it would be good practice to have the master fader (set at 0) in ableton/renoise peaking at around -6 or so, which would be indicative of a good mix with plenty of room for individual sounds to breathe and nothing running all over each other or whatever. At which point you can whack the master bounce in logic and spunk all over it with compressors until you sound like Justice.

Second scenario, is that you're bouncing individual tracks from renoise/ableton and want to mix them in Logic. In that case there wouldn't be a thing wrong with going about it in the same way as method 1 - ie making sure your master fader (set at 0) is peaking around -6, thereby basically ensuring you a slammin' mix, bounce those stems, whack it in Logic and mix away - bearing in mind that if you're still planning to send it to a mastering engineer they're more likely to be your friend if your mixed version is still peaking at less than -4 or so (set at 0, dont go moving that pesky bugger).

However with the second scenario you could also just make sure the individual channels aren't peaking and then whack them in logic, mix to less than -4 and you're away again, either to the ME or to whack Ozone on the end with an exciter bigger than the shit they sell in ann summers.

Basically the answer to your specific question is - it doesn't matter: Just follow the mantra of ending up with a final pre-mastering mix that's peaking somewhere between -7 and -3 (roughly, of course, nicht scienceo hombre!) without CHEATING (no moving the master fader, no squashing everything with compressors) and youll be dandy.


tl;dr: read the thread again.

[macc]

It isn't about the faders being at -10 (or whatever). If your input signal is at plus 10 zillion dB, -10 on the fader won't even touch the sides.

It's about making sure the signal is at the right level as early as possible in the signal path. Ideally, you don't even have to fanny about with the faders. If the samples, bass, pads, tambourine, drums and cheesey 'woah yeah baby' sample are all at the right level, why would you want to touch the faders? Tune's mixed, right? Get the sounds at the right level at the earliest opportunity.

If you're in a band and the drummer plays too loud, you don't play the gig knowing the dude is too loud and it sounds rubbish, then go home and bitch about it. You tell him to STFU there and then FFS. Same thing. Sample CD drums are almost always normalised to 0dB - turn the sample itself down (the audio file, or use your sampler) before you even bother with faders. Synth patches are usually way too loud too - turn the synth down, on the synth GUI. Bollocks to messing about with faders. Ideally, it's already right by that stage.

If all your faders are at/near unity and you never even touch them, IMHHHHHHHO you are doing things the right way. The sound was right before it even hit the mixer, why do you need to touch it? Saves time, saves messing about, less faders all over the place and wonky gain structure to mess up. Then if everything is absolutely perfect but your Chas and Dave 'Rabbit' sample can come down half a dB, then it is easily done with a fader. That's not so easy when the fader is at -28dB, and the smallest step downwards takes it to -36dB (this is often called fader resolution).

Mixing - technically speaking - isn't intrinsically difficult. People just make it hard for themselves by giving with the left hand and taking with the right.

I can't believe I am still trying to think of new ways to explain this I do have a life outside of this thread.... honest

[Metaphor]

I'm not sure where you've got this magic -10 from... well you got it from this thread but Macc has been insistant that this isn't an exactly science, rather a set of guidelines.

well, on page 1 of this thread, Macc says something to the effect that things should be around -8 to -10 for drums as a guideline. I'm guessing this meant each individual channel as a start, and not the master channel?

Your post was slightly unclear so there are two possible scenarios that you may be referring to, but luckily my advice is the same for both.

My Bad.

Scenario 2. I bounce individual tracks out of renoise/ableton and mix in Logic. Master is always @ 0 on everything. I never touch it. Faders are wherever they need to be. So, you're saying adjust level in abe/renoise and then start @ 0 in logic and use the faders as needed?

Basically the answer to your specific question is - it doesn't matter: Just follow the mantra of ending up with a final pre-mastering mix that's peaking somewhere between -7 and -3

I'm not sure I agree here, hence my questions. I mean, if I clip inside ableton and render that for use in logic, it's going to sound shit no matter what I do, right?

[macc]

Check my post above chief

[Metaphor]

It isn't about the faders being at -10 (or whatever). If your input signal is at plus 10 zillion dB, -10 on the fader won't even touch the sides.

I get that. Input level, of course. I said faders, but I meant input level. I guess, since I'm mostly using samples, many of which have been limited to 0db, turning down the fader has this effect, right?

If all your faders are at/near unity and you never even touch them, IMHHHHHHHO you are doing things the right way. The sound was right before it even hit the mixer, why do you need to touch it? Saves time, saves messing about, less faders all over the place and wonky gain structure to mess up. Then if everything is absolutely perfect but your Chas and Dave 'Rabbit' sample can come down half a dB, then it is easily done with a fader. That's not so easy when the fader is at -28dB, and the smallest step downwards takes it to -36dB (this is often called fader resolution).

Gotcha. Simple answer is: turn it down at the gui level. On a sampler, turn down the sampler, not the mixer. Okay, I think this is really clicking. Sounds like I turn it down in renoise/ableton. Thanks, guys for contributing the evolution of my art.