gain structure and mixing aka THE MONEYSHOT THREAD

[Perfecture]

I see what your saying man, thanks for the advice, so in essence you are saying there is no difference between mixing as close to 0 as you can and then pulling the master fader down to -4 OR mixing to -4 and leaving the master fader where it is?

There's no mathematical difference. There are numerous other reasons to do it though - workflow, efficiency, repeatability between mixes, just not running the risk of clipping your track, learning how to maximise impact and power without just mashing shit against zero, working to a consistent ear level, blah blah...

Answer me this;

Why do you need to mix as close to 0 as you can?

I just remember someone saying mix to 0 so that all the elements equal up to 0 (so you know that the spectrum is filled nicely and not over done) and then bring the master down before mastering. But making sure that there is movement in the signal (dynamics) meaning when it gets mastered there is room for tweaks and also so that it doesn't sound brick walled to shit.

From what I can tell I have a load of dynamics in my track cause the signal on the master fader is sort of flowing smoothly up and down by a few db. what I gather from this is that it means I haven't slammed everything and there are lots of dynamics

[safeandsound]

It seems after 53 pages either some people have not understood what has been explained
or just have not read it through. I am gonna have a stab ...........

I just remember someone saying mix to 0 so that all the elements equal up to 0 (so you know that the spectrum is filled nicely and not over done) and then bring the master down before mastering.

"mixing to 0" is not related to filling up the frequency spectrum.

I will summarize:

1)Don't use a limiter on your master bus.
2)Operate your DAW at 24 bit.
3)Before you start mixing, peak your kick/snare drum at a reasonable nominal value i.e. -18dBFS to -12dBFS on the channels meter.
Use that kick or snare as a reference to balance your mix. (i.e. leave it where it is)

If you are a producer who mixes as you produce (I used to do this) part of your new work flow would be
to ensure the moment you start programming drums you peak them at the above nominal level. Then
bring your synth parts down to meet them. i.e. balance them.(if you have programmed a melody first).

It need not be over thought. So if you have got that and find it easy, cool.. you can read the next bit for context,
if not re-read the above.

------------------------------------------------------------------------------------------------------------------------------------

This is very similar to PFLing on an analogue desk. (Pre-Fade Listen) On an analogue console you would send any given channel to the PFL bus. (normally a VU or LED meter at the stereo bus out). By peaking your kick at 0VU or thereabouts you are optimizing the gain structure of the mixing console. Optimizing gain structure just means operating the mixing console optimally between the noise floor of it's analogue components (hiss) and the threshold of audible distortion (the point where the amplification gain of the electronics exceeds it's design capacity). Peaking an instrument with high transient energy such as a drum between -18dBFS to -12dBFS is the same but within the channels of your DAW.

All of this is not visually helped by the scaling on DAW metering -14BFS looks incredibly low relative to the top of the meter (usually around 1/2 way up or so) when in fact it is electrically not far from a signal level of about 0VU.

Look at the first and last meter scale on this diagram and see where they meet horizontally. (it is not perfect as it does not take into account ballistics or RMS/averaging nature of a VU meter, nor the fact that many VU's are not built to conform to spec) But it gives you a historical context from analogue to digital metering and why it matters.





cheers

[macc]

All of this is not visually helped by the scaling on DAW metering -14BFS looks incredibly low relative to the top of the meter (usually around 1/2 way up or so) when in fact it is electrically not far from a signal level of about 0VU.

This is a very good point - some DAWs allow you to adjust the of colour your meters which is a very useful visual cue.

Hang on, did this get mentioned in this thread already? I can't remember

[safeandsound]

Hang on, did this get mentioned in this thread already? I can't remember

It would not at all be surprising, the odds are in it's favour, lol.

[macc]

I was getting deja vu as I typed it

Ah well, as long as it's helping people!

[Electric_Head]

Use that kick or snare as a reference to balance your mix. (i.e. leave it where it is)

[mikeyp]

i either do it all wrong or it's just one of those things i get, mixing is (mostly) easy for me

anybody else ever felt like this?

tough luv but from a technical standpoint ur tracks have severe dynamics issues... altho aesthetically that might be just what you're after

tough luv.. no harsh feelings bro

care to elaborate?

snares dont hit. dynamic margin sounds like your drums lack impact power

yet the mix sounds pretty clean... anyone on this or am i really bad at hearing dynamics? and btw i love the tunes themselves!!!!! just wanted to stress the fact that a very very very few people can actually say mixing is something they dont need to improve on...

oh don't get me wrong, i didn't mean i'm a pro and it's the easiest thing in the world to do. I meant the concept is simple to understand. and I know that knowing how to do something and doing it are two very different things. but, the general idea and how to go about it are not hard. I always see a fuss about avoiding clipping. uh, simple solution - turn everything down? is it that hard? or like i originally said, am I the one not getting it?

[didi]

oh don't get me wrong, i didn't mean i'm a pro and it's the easiest thing in the world to do. I meant the concept is simple to understand. and I know that knowing how to do something and doing it are two very different things. but, the general idea and how to go about it are not hard. I always see a fuss about avoiding clipping. uh, simple solution - turn everything down? is it that hard? or like i originally said, am I the one not getting it?

I think it's more about getting repeatable high quality mixes by knowing roughly where everything should be dynamics wise for everything to peak where you want it.

[mikeyp]

oh don't get me wrong, i didn't mean i'm a pro and it's the easiest thing in the world to do. I meant the concept is simple to understand. and I know that knowing how to do something and doing it are two very different things. but, the general idea and how to go about it are not hard. I always see a fuss about avoiding clipping. uh, simple solution - turn everything down? is it that hard? or like i originally said, am I the one not getting it?

I think it's more about getting repeatable high quality mixes by knowing roughly where everything should be dynamics wise for everything to peak where you want it.

i didn't mean that was what it's about, just using that as an example because it's the first thing that came to mind.

[symmetricalsounds]

I meant the concept is simple to understand. and I know that knowing how to do something and doing it are two very different things.

yeah in general the concept is simple, balance all elements of your mix throughout a tune. how succesful you are is obviously a different matter.

[e-motion]

@macc what do you think of this:

AF: What compressor or limiter do you use on the final mixes? Because your tracks don’t sound so compressed compared to other electronic music songs.

Eisen: We don’t use compression at all in the final mix.

AF: Ok. Interesting.

Eisen: I mean, this is our- let’s say- our secret. Big secret. We do the mastering ourselves as well. We basically learned that if we push the gain really high and get everything distorted in a way that you don’t hear distortion, you just see the red really really - say 6db+ you don’t hear distortion yet and we record that analog to another convertor then we get much better sound. We tried even the Waves Ultramaximizer- I don’t find them sounding better than what we do. I find the Waves, for example, taking away the highs a little bit. And not only Waves, a lot of other companies that do the same things. The highs always go away, and something in the details goes away. This way, we record it analog, in our case we record it to the Prism Sound, and we get the Prism Sound distorted ['visually'. So the inputs are distorted and the meter shows that it cliping but my ears don't hear distortion]. We get really high gain and with a sound that doesn’t sound compressed.

Source: http://en.audiofanzine.com/sound-techni ... chemy.html

And thank you for your tips, they're really spot on!

[skimpi]

@macc what do you think of this:

AF: What compressor or limiter do you use on the final mixes? Because your tracks don’t sound so compressed compared to other electronic music songs.

Eisen: We don’t use compression at all in the final mix.

AF: Ok. Interesting.

Eisen: I mean, this is our- let’s say- our secret. Big secret. We do the mastering ourselves as well. We basically learned that if we push the gain really high and get everything distorted in a way that you don’t hear distortion, you just see the red really really - say 6db+ you don’t hear distortion yet and we record that analog to another convertor then we get much better sound. We tried even the Waves Ultramaximizer- I don’t find them sounding better than what we do. I find the Waves, for example, taking away the highs a little bit. And not only Waves, a lot of other companies that do the same things. The highs always go away, and something in the details goes away. This way, we record it analog, in our case we record it to the Prism Sound, and we get the Prism Sound distorted ['visually'. So the inputs are distorted and the meter shows that it cliping but my ears don't hear distortion]. We get really high gain and with a sound that doesn’t sound compressed.

Source: http://en.audiofanzine.com/sound-techni ... chemy.html

And thank you for your tips, they're really spot on!

That sounds like a really cool idea!

[ChadDub]

This shit got unstickied.... Good job

[macc]

There's nothing new in mixing that way - it's no secret or anything like that. We've had conversations about the same thing on here in the past (possibly with Rob Sparx if I remember rightly...?)

There's a few points to consider here;

- First things first: Clipping's not magic. Often carefully controlled clipping sounds cleaner than limiting in the way the interview mentions, but it can also give awful problems downstream if there are changes required. Not hearing distortion when you mix it down doesn't mean it isn't there. Making changes later can (arguably, will) bring it out, though it seems these people (I have no idea who they are but people seem to like them) are doing the final steps themselves. More below.

- It's nice to know some people don't use compression on their master. You just don't need to. Forget the internet. There's been a spate of it recently, like it's a competition to see who can compress things the most. Blurgh.

- The Prism Sound they refer to (presumably either the Dream or Orpheus) can take a beating. Lots of producers use clipping in the mix to maintain a better impression of punch. Lots of mastering engineers (myself included) use clipping as a way to get louder masters at times. I've seen my favourite engineer clipping a Prism Dream and it sounded good, I let my Mytek clip a little bit sometimes as it sounds better than straight limiting or whatever (especially on 'pointy' snare drums, FWIW). As always, it all depends on the circumstances/sounds concerned. There have been many debates about the merits of clipping expensive convertors; I don't really give a toss - it's not magic. The absence of a release stage (or indeed, any time-dependent behaviour at all) is both what makes clipping at times more transparent than limiting (as mentioned in the interview), but also means it has a more specific distortion that can raise its head later on - either on playback on other systems, or when other work is done to the track.

- Clipping like this is no different to limiting or just clipping a master in that it bakes the cake - if any changes are required later you're likely to be in severe trouble. Ideally it should be absolutely the last thing done to a bit of audio. If you do anything to a flat-topped waveform, you don't have a flat top any more. What you are left with is the distortion without the 'benefit' of the reduced dynamic range. Then, once the required changes are made, you are obliged to re-limit or re-clip those same peaks AGAIN to get the level back to where it was before. This is not good for audio


With all this in mind, I (personally!) don't recommend this as a way of working for people in general for a few reasons.

- Your mix has to be absolutely SPOT ON, and I mean, spot on. If you're going to do anything else to the audio after, if you have any doubt, then you're playing with fire.

- If you don't have the courage or faith in your mixes to commit to mixing down a heavily clipped track, but have mixed it with heavy clipping, if you turn down your track to provide headroom for subsequent processing then you will completely change the sound and shoot yourself squarely in the foot. Most often this ends up with insanely spiky transients that have been massively over-emphasised in order to counter-balance the clipping and maintain the impression of punch. I see things like this a fair bit; it's pretty obvious. It's also a pain in the arse


To make my position clear;

To an extent, smashing the absolute balls out of a track can help to inform your mix decisions, and personally, I heartily recommend doing this for informative purposes. I've detailed one or two methods for this on here before. You can make decisions based on how the mix holds up under extreme pressure, and then remove it to see how things are in the real world.

This way you know you have the best of both worlds: you know your mix can take a proper spanking, but you haven't cooked irremovable distortion into the mix that is likely raise its head later on.


Standard disclaimer - this is my opinion, based on my experience and my day-to-day work. There aren't any rules, other things may work for some people, blah blah blah

[Kaslo]

^^ Spot on explanation. I often use a little "pseudomaster" chain on my master, and have it bypassed while mixing, just so I can toggle it and see if my mix will hold up to being slammed, or mildly boosted, as it would be (more professionally) at the mastering stage. Good to have both frames of reference handy.

[e-motion]

@ macc: yeah that's what I thought, do it last in the mastering stage. I was just curious because I found some EDM producers that like dynamics and don't compress the mastering, rare thing this days I was also curious to know if the analog clipping technique was common in mastering as well as not using a compressor. But yeah I forgot this is a mastering thread, sorry and thanks a lot

[legend4ry]

Analog clipping isn't as dangerous as digital clipping though, right?

Ive seen desks being pushed to the red and still sound brilliant.

(or is this the point? I didn't get the wording totally..)

[Ldizzy]

analog saturation is often regarded as "tasty".. as the saturation curve is gradual and often described as "more musical"

digital clipping is an abrupt phenomenon and its effect isnt beautiful sonically imo

[macc]

Analog clipping isn't as dangerous as digital clipping though, right?

Ive seen desks being pushed to the red and still sound brilliant.

(or is this the point? I didn't get the wording totally..)

Clipping converters isn't analogue clipping, it's very very much more like digital clipping, but not 100% either.

Remember you have an analogue stage at the front of the ADC, which feeds the actual converter chip itself. So exactly how the audio gets clipped when going A>D (or rather, what happens to it just before it gets clipped and therefore, exactly how it is clipped) depends on the converters being used, the design of their front end.

Expensive converters are sometimes renowned for handling massive clipping nicely ('You have to have a Lavry Gold to get a loud master', etc), and apparently all sound different. I think I'm actually glad to say I haven't spent an awful lot of time worrying about it. I clip the Mytek sometimes, sometimes it sounds rubbish. So it's just like anything else.

EDIT: Should be a bit clearer: You end up with a straight-up clipped waveform; it's how the analogue stage affects how it hits the converter that makes the difference.

[hookey]

So you have this dnb tune peaking at -6db so the engineer has a decent amount of headroom to push it to 0 the right way. The question is: How do you get the most of those -6db? I mean, there are tunes by big name producers that sound louder in the intro when it's not yet peaking at 0, more like at -3db (the classic dnb intros with the drums and some effects in the background before the actual mid-sub action starts) that are sounding louder than your tunes at the most bussiest 0db moments. How do they do this? I guarantee all the non-needed frequencies are low/high passed. I high pass the sub at 30hz (btw, it's ok to just put a 30hz high pass in the master so you guarantee there are not frequencies under 30hz instead of adding an EQ on every channel wasting CPU resources?) I low pass everything beyond 22hz (same, is it ok doing this on the master channel?) then I individually threat my reecee/bassline it individually on 3 parts (low-subs 30 to 140hz, mid 140 to 5000hz, high 5000 to 9000hz) are those ranges making sense? I tend to cut at 9000hz because the high frequencies from the reeces start clashing with the hi-hats and cymbals which start at around those frequencies (8000 to 15000hz), is this ok? It's one of my big questions aswell: How do they manage to get those reeces/growls/whatever sounding bright at the same time leaving space for the hi hats for that nice crispy touch? I always tend to think "this reece is lacking top end..." and I end up cranking it too much beyond 9000hz and it starts mudding the hi-hats and top end percusison sounds. I can't seem to find a balance there. I high pass my kicks at around 80 hz, snares at around 120hz, and EQ wise don't do much more really. I used to over do EQ'ing by putting notches on everything and I endeed up with a thing sound with no midrange. So to not extend this a lot more: -What are some strategies to get your shit sounding phat maintaing a -6db overall peak? (I end up putting a limiter on everything to control the peaks, so im sure it's not peaking beyond I don't want to, and try to get as much sense of loudness within that range, but I usually end up over-using those "enhancer" features like on the Sony Limiter one. Is this ok?) -How do you go about having a reesee/growl/whatever sounding full and without being forced to cut the higher frequencies too much so it doesn't collapse with your hi-hats stuff?