TranceAddict Forums (www.tranceaddict.com/forums)
- Production Studio
-- How to create more headroom in your mix (with EQ)?
Pages (4): « 1 2 [3] 4 »

quote:

Originally posted by Subtle Yeah i know about the loudness war, and if i see spikes in my tracks i keep them, because i know limiting them will make the sound worse. I realise posting pictures was the worst idea ever. What I mean about bad mix is that the volume cannot be raised alot without loosing dynamics, while the good mix can have its volume raised without it sounding different than in the sequencer. Which means lesser processing to get an even volume overall, thus a more correct mixdown. The bad mix sounds good, it just cannot sound as high in volume as other tracks going with it.

hence you need to cut and control your transients on track level in order to optimize headroom and perceived volume

quote:

Originally posted by Raphie hence you need to cut and control your transients on track level in order to optimize headroom and perceived volume

Yes absolutely. That is why it is a bad mix, it is an old mix of mine.
I have noticed that my mixes are just getting flatter because i get better at mixing. But only by following the ears and nothing else, and i need the mixes to not spike, as i do not master my tracks afterwards, i simply compress once at 3 db threshold, slap a limiter on them and there they go.

And before anyone says "no no dude flat is bad" i have to also say that my kicks usually peaks between -3b and -6db which means room for transients and dynamics.

The whole notion that a mix needs to be loud [peaking at 0db] is the loudness war in action. The concept is basically "What is the best way to get maximal loudness and not distort?" And that is the approach DJ RANN is coming from I believe. This is accomplished by compressing audio all to fuck, flattening it out, and pushing it to the 0db wall. [Which, Subtle is what your "good" mix looks like, of course ive not heard it.]

As for the noise floor, he is referring to my maxing at around -5 to -8db ...I try not to let anything go over those.

quote:

Originally posted by Subtle Now, this is what i dont understand fully what has to do with the master channel. How can there be a difference between having having the master channel at higher volume and the individual channels set lower, as opposed to having the individual channels set higher and the master channel lower. The output should be exactly the same, no ?

i might've misunderstood you and dj rann but this is what i get from it (and how i've been doing it myself). you don't necessarily turn every individual channel down, only the problematic ones

you have a track but one or two instruments are clipping. you would turn down only the clipping instruments instead of turning down the whole track via the master. relative to the rest of the mix, the clipping instruments are too loud. turn down everything and the clipping instruments are no longer clipping, but still too loud when compared to the rest of the mix, which now has lost dynamic range since you cut out its maximum possibility of loudness by decreasing the master volume.

it's like cutting down the whole tree because two branches are in the way of your house. why wouldn't you just trim the branches instead?

quote:

Originally posted by cronodevir As for the noise floor, he is referring to my maxing at around -5 to -8db ...I try not to let anything go over those.

Okey, i see.. well that makes perfect sense, you should of course render with the MAX as close to 0 db as possible.

The thing with the loudness war is that you compress/limit so you LOOSE the transients, if your mix is maxing at 0db you arent loosing any transients at all (hence it got nothing to do with loudness war having the peak close to 0db), its when you limit/compress that you are taking transients away.

The loudness war is when you compress everything so hard that the waveform is just huge brick.

quote:

Originally posted by Omega_Blue i might've misunderstood you and dj rann but this is what i get from it (and how i've been doing it myself). you don't necessarily turn every individual channel down, only the problematic ones

Yeah i understand that, but in my case i does not make a difference. I use my ears to spot the problematic channels, and when i render a Test mixdown of the track i import it in a wave editor, and if I see spikes i can easily recall where the problem areas are with looking at the waveform, and then fixing them when i open the project again. I just seem to get a lesser and lesser amount of problem channels and that only by hearing (which is what mixing is ALL about)

quote:

Originally posted by cronodevir Except every piece of music or audio from a major studio sounds like shit. Highly compressed garbage. The things they teach are about making your audio stand out, not about the best possible quality. So sure, you may have went too school and you are a professional engineer, but the fact is 99.9% of "professionally" processed music sounds horrible. And this garbage is brought into the audio when ever a track is sent to be mastered. So somewhere along the line, something tells me these processes fail at some point. Wether the processes are bad, or they are implemented incorrectly, who knows. All of that said, as I mentioned before ive never had any problems with my mix from anyone, and I have had a few tracks released on compilations. Yes ive used limiters on channels before, but sometimes stuff still gets by.

I agree that a lot of pro processed music is horrible but that's not a fault of engineering quality, it's down to the producers (in the real sense), composers and artists.

Bear in mind that engineers don't get to make artistic decisions about the content of music so don't blame them or their methods of providing the best possible recording, even if someone else fucks it up. It's like blaming the company that makes brushes for the current state of impressionist art.

The failure is lies with the creators of the content, not those that engineer it. Also, a couple of peaks actually don't matter on individual tracks, and sometimes when you clear the red on the mixer they don't appear the second time.

If you lower the master say 3db, then bounce, the track will have a dynamic range of 93dbfs instead of 96dbfs (16bit). So the noisefloor is introduced at that level of 93dbfs. SO when you turn it up to reach it's full potential 96dbfs (by 3 dbfs, which is twice as loud) you are effectively increasing the proportional noisefloor by a factor of 2 as well.

quote:

Originally posted by DJ RANN I agree that a lot of pro processed music is horrible but that's not a fault of engineering quality, it's down to the producers (in the real sense), composers and artists. Bear in mind that engineers don't get to make artistic decisions about the content of music so don't blame them or their methods of providing the best possible recording, even if someone else fucks it up. It's like blaming the company that makes brushes for the current state of impressionist art. The failure is lies with the creators of the content, not those that engineer it. Also, a couple of peaks actually don't matter on individual tracks, and sometimes when you clear the red on the mixer they don't appear the second time. If you lower the master say 3db, then bounce, the track will have a dynamic range of 93dbfs instead of 96dbfs (16bit). So the noisefloor is introduced at that level of 93dbfs. SO when you turn it up to reach it's full potential 96dbfs (by 3 dbfs, which is twice as loud) you are effectively increasing the proportional noisefloor by a factor of 2 as well.

Sure, but one is more likely to notice you lost a few hairs on your arm...than that extra noise. If in fact the noise exists in the first place. The effect of moving the master slider down is so insignificant it doesn't matter.

quote:

Originally posted by DJ RANN If you lower the master say 3db, then bounce, the track will have a dynamic range of 93dbfs instead of 96dbfs (16bit). So the noisefloor is introduced at that level of 93dbfs. SO when you turn it up to reach it's full potential 96dbfs (by 3 dbfs, which is twice as loud) you are effectively increasing the proportional noisefloor by a factor of 2 as well.

I don't know what DAW you're using but the master fader is usually within the 32-bit floating point part of the audio engine. So the bit depth of the export is irrelevant. (Unless you put dithering in pre-fader inserts of course). It does not matter if you attenuate the master fader or the individual channels. Of course 32-bit floating point has its limits but you'd have to do some stupidly large gain increases/attenuation to compromise the quality of the signal.
Personally I don't touch the master fader. But that's just because it will make the fader values of the channels meaningless.

Maybe, but if you cut your kick and bass above 500hz, there will be nothing decent of your kick or bass left. Kicks also contain vital higher frequencies, as bass (except for subbass ofcourse). [/QUOTE]

no one said ABOVE 500hz, I mentioned AROUND 500hz. (where you can play with the Q

Just slap 5 compressors 2 maximizes and 3 limiters on the master. Make sure you clicks the "Randomize Parameters" button if your DAW has one. This will give your music that nice high quality commercial sound.

quote:

Originally posted by Subtle Yeah i know about the loudness war, and if i see spikes in my tracks i keep them, because i know limiting them will make the sound worse. What I mean about bad mix is that the volume cannot be raised alot without loosing dynamics, while the good mix can have its volume raised without it sounding different than in the sequencer. Which means lesser processing to get an even volume overall, thus a more correct mixdown. The bad mix sounds good, it just cannot sound as high in volume as other tracks going with it.

Exactly - very well put. Mixing well and the loudness war two different things. You can have equal or closer relative levels on mix without it being labeled as a casualty of the loudness war. What I mean is to maximise the use of your available/potential gain by creating a well balanced mix with relative levels is good mixing practice and has nothing to with the loudness war.

Compressing the utter shit out of your mix to make the relative levels closer and reduce dynamic differences is a major component in the loudness war, but these two ways of working are not the same thing at all. That "high quality commerical sound" is';t anything random either. It's very carefully done to make crap pop songs fight against other crap pop songs when played on tinny radios or ipods. Technically though it does serve a purpose, and one that engineers specifically try to do for that medium and that particular genre.

Omega_Blue - you're description is exactly right. You don't cut the trunk of the tree because a couple of branches are too big. If your a couple of tracks in your mix are causing the master to clip, then you haven't balanced the track right, plain and simple.

Again,that's because your system (+ it's calibration) and the way you work should align with simple, proven, audio engineering techniques and theory such as gain staging, mixing to unity gain etc.

Eldrtich - you're right but when working with samples (and most of us do at some point with EDM) 32bit floating point processing only helps with them clipping, not reducing noise which is inherent in all samples. Also, if you're project is set to 16bit, then 32bit FP again only helps with closed system internal clipping, not the quality of your final mix audio which will be at 16bit. Yes, internally the softsynths will "compute" at 32 bit but again the moment you bounce them to audio files, they are truncated to 16bit which has a dynamic range 96dbfs and their noise floor relative to the samples is still maintained.

i don't think headroom is the word you are looking for.

Headroom is the amount of space you have before overloading a system.

Creating space in a mix has absolutely nothing to do with headroom. As far as the question you meant to ask , well i'm sure some others have already answered.

quote:

Originally posted by RichieV i don't think headroom is the word you are looking for. Headroom is the amount of space you have before overloading a system. Creating space in a mix has absolutely nothing to do with headroom. As far as the question you meant to ask , well i'm sure some others have already answered.

That's what I've been getting at.....

quote:

Originally posted by DJ RANN Finally, and not to be too pedantic - the title of this thread is actually a misnoma. You can't "create headroom" as such with EQ. Headroom in it's correct sense relates to dynamic range of the mix. EQ can;t by definition create more headroom. EQ can however provide frequency separation and there provide space between elements, but again it is technically impossible to "create headroom" with EQ.

It all depends on how you define headroom. Headroom implies that there is also a "ceiling" (which in general is 0dbfs) if your whole mix is continuously peaking close to 0dbfs, you've got little headroom.

If your mix is peaking @ -3dbfs you've got more headroom.

HOWEVER this tells nothing about the dynamics in a mix. the fist example can have less dynamics, because all transients are already shaved off then the 2nd one.

if the 2nd file has no transients, you can maximize it, or even limit it a bit, but you will hear squashing artifacts kicking in very quickly.

so if headroom is defined as amount of db's between peaks and 0dbfs EQ can help increasing that space. But once the dynamics are gone, you will never get them back. (you can get back something simmilar through a transient shaper, but it will never sound the same) some people also call the "inter instrument volume differences" headroom, though they really mean dynamics.

Do some reading on 0dbfs, K14 scales and RMS levels.

and then there is also a whole different dimension called "artistic aproach" which can be intented sqaushing to emphazise groove or to get a less distinct "knock knock" kick, sidechaining etc....

So i guess there is no "wrong" or "right" it all depends on the effect you're after and the techniques u use to get there. Not understanding the technique means you will not have targeted results....

quote:

Originally posted by DJ RANN Eldrtich - you're right but when working with samples (and most of us do at some point with EDM) 32bit floating point processing only helps with them clipping, not reducing noise which is inherent in all samples. Also, if you're project is set to 16bit, then 32bit FP again only helps with closed system internal clipping, not the quality of your final mix audio which will be at 16bit. Yes, internally the softsynths will "compute" at 32 bit but again the moment you bounce them to audio files, they are truncated to 16bit which has a dynamic range 96dbfs and their noise floor relative to the samples is still maintained.

How is that relevant to whether it's a bad idea to lower the master fader if the mix is too loud? All of the summing, including the master fader is done in 32 bit fp.

quote:

Originally posted by DJ RANN That's what I've been getting at.....

I disagree. Technically EQ is used to attenuate specific frequency ranges thus you can directly conlude you can create more headroom by using EQ. If you have a 250hz sine and eq 250hz down by 5db you have 5 dB more headroom, same goes for an entire mix (up to a certain extent). Especially if you attenuate dominant frequencies it can provide you extra headroom.

time correct everything so timing is as precise as a swiss clock, then shelve everything that's not bass or kick to 150hz. That should clean your mix in general.

12/24/36db/oct shelf?

quote:

Originally posted by Raphie 12/24/36db/oct shelf?

Whatever is the steepest you can get away with I would think...

In general while mixing you don't conciously EQ to create headroom, you do it through mixing relative track volumes. EQ'n gives the different parts in your mix space, a frequency spectrum to breath in, which is what the OP is probably looking for. I think if the OP used the word SPACE instead of HEADROOM, this thread would be a better place. EQ for Space, relative Level Mix for Headroom. But Storyteller did make a point, that EQ'n can overlap as an extra means for making Headroom. For example, if your cutting sub 40hz frequencies on a Bass sound that was adding an extra 1 to 2db of volume to the channel, then you'd be adding some space AND headroom by cutting off those frequencies. Gotta twist to shout.

To master, the following were used in this order -

1. Multi-band Compressor
2. EQ
3. Limiter

Pictured at peak close to volume, here:



To obtain the following result:



In such a way, I used a multi-band compressor and subtractive EQ, sparingly, to take out frequencies that would allow for maximization of over-all volume without squashing the dynamics. In that regard, EQ was used to maximize the available headroom so that other frequencies could have more prominence.

do yourself a favour: http://www.pleasurizemusic.com/en/download
install and start (re)mixing.....

That looks sweet. Thanks!

quote:

Originally posted by Raphie It all depends on how you define headroom. Headroom implies that there is also a "ceiling" (which in general is 0dbfs) if your whole mix is continuously peaking close to 0dbfs, you've got little headroom.

This unfortunately is not a point for subjective discussion. Headroom is the space before you exceed 0dbfs. If you want to use the term headroom for the differences between loud and quiet peaks then you're using the term incorrectly. That is called relative dynamics.

quote:

Originally posted by Raphie HOWEVER this tells nothing about the dynamics in a mix. the fist example can have less dynamics, because all transients are already shaved off then the 2nd one.

....exactly why it has nothing to do with relative dynamics becuase that's not headroom (as above).

quote:

Originally posted by Raphie if the 2nd file has no transients, you can maximize it, or even limit it a bit, but you will hear squashing artifacts kicking in very quickly.

But the second file does have transients they are just closer together in relative terms - it doesn't mean they don't exist.

quote:

Originally posted by Raphie so if headroom is defined as amount of db's between peaks and 0dbfs EQ can help increasing that space. But once the dynamics are gone, you will never get them back. (you can get back something simmilar through a transient shaper, but it will never sound the same) some people also call the "inter instrument volume differences" headroom, though they really mean dynamics.

Well the space between peaks and 0dbfs is (not "if") headroom, but that's only one part of it becuase headroom also takes in to consideration the overall RMS level as peaks alone don't give a true representation of the level of the track. They both are components of headroom and both affect it's value.

quote:

Originally posted by Raphie Do some reading on 0dbfs, K14 scales and RMS levels.

err...How about the thread that I posted a while back called "what 0dbfs actually means" which also happens contains the tutorial I made with a step by step guide to calibrating your monitors to the K system (including the download link for the claibration noise files)... (I also already posted the link to this tutorial at the beginning of this thread)....

http://www.tranceaddict.com/forums/...threadid=512627

quote:

Originally posted by storyteller I disagree. Technically EQ is used to attenuate specific frequency ranges thus you can directly conlude you can create more headroom by using EQ. If you have a 250hz sine and eq 250hz down by 5db you have 5 dB more headroom, same goes for an entire mix (up to a certain extent). Especially if you attenuate dominant frequencies it can provide you extra headroom.

Actually, you are completely technically right that if you eq a frequency (especially a bass frequency that has more result on final dbfs) down then you could create more headroom but this would be the least significant/effective method of creating headroom - relative mixing, ,limiting, compression are so many more times effective at creating headroom that it's not really the right way to go about it. It's like trying to kill someone by stabbing them with toothpicks - you'll get there eventually but there's a lot more effective ways of doing it. A
Generally speaking (and there are obviously exceptions) eq is a precise tool, not really used to significantly like a blanket tool to lower the overall DB to create headroom. Yes you could use many separate instances of EQ to acheive this, but you would just normally use things like band specific compression etc.

That's why personally, I don't consider EQ a way to really create headroom (in the correct definition of headroom as stated above).

Laurent - that's actually very good advice. Even slight timing issues can really fuck up things in a mix and so many sounds have lower harmonics that eventually add up to just make everything muddy.

Eldritch - I don't like to touch that master fader because that's an integral part of my speaker calibration gain staging. By lowering the master I'm affecting the accuracy of the calibration, and therefore my ability to mix accurately. It's really a minor consideration in fairness, but I just don't see the need to touch the master. I just adjust the tracks themsleves and problem solved.

Finally at Kismet - I disagree with a lot of what you've posted on here recently but what you're saying here is totally spot on.

phew......

Powered by: vBulletin
Copyright © 2000-2021, Jelsoft Enterprises Ltd.