Dark Ride: 'Hammer Down'

[SonicTramp]

"As I see it - you don't actually have to hear everything all the time.  It's a mix, not a collection of individual, perfectly frequency separated parts as such. I think a mix would sound a bit odd if you had the kick perfectly bracketed to fill only 30-70hz, the bass 70-150, the guitar 200-3k, the vocal 3k-7k, the hi-hat 8-12k, etc." 

I agree, and there's a good reason not to work like that. Except for percussion and drums, which are constant pitch, for the most part, the rest of instruments will not stay at the same frequencies since they all constantly moving from chord to chord, note to note, rendering the "bracketed" method unusable since pitch (note)=frequency.

Unmasking, to my understanding, is to get rid of a certain frequencies from another instrument that interferes with the intended instrument. For example, I notice there's a little snare in almost all the guitars, vocal, strings, etc. A little cut from 3-4/6-8 kHz, from almost any of them, but I'll stay with the guitars bc it's top end rich, will allow the snare's top to poke out a bit most of the time. And if it does damage to it, I can always put it back and try on another instrument. The key is to go right to the right one and that relies on how keen one's hearing is. I am still working on it.

I get it that mixing is quite a complex process bc there are just too many ways to...skin a cat. I find I can cut some top end of a, say, guitar by lo-shelf, LP using an EQ, but I also can do that with a faster attack on the compressor or slower attack on a gate or even a limiter. Each of them can do the same job to an extend. So which one? My answer is: whatever one makes the mix better. Here, too, the key is to go right to the right tool and still with even keener ears. I am still working on it.

From your comment, and thank you for such a detailed one at that, on my last mix, I also noticed level balance changes in the mix. Louder snare, kick, bass. Lower synth, guitars, etc, tho I never touched faders. I knew I had to re-balance levels but resisted bc I just want to make sure what I heard would be what you'd hear. It helps me understand more about the relationship bet frequency and loudness. I kinda understand it but much more now that I hear it in action. That's always more convincing to me. I'll work on it a bit more and try to apply what you suggested and see if I can do it. I still have the old file as a safety net. Thanks, man.

[mikej]

Hi,

Regarding unmasking - yes exactly there are many possible solutions to each problem you may have.

As an aside - I think one of the issues I had, and still have to some extent I guess, with mixing is the feeling that there is only one correct answer, that I am doing it wrong, and that it is completely obvious to everyone that I am an idiot. All arguably true I suppose - however I think that this is where the art comes in to the science. No one really knows what you do in a mix. As I think Andrew Scheps once said, really the only thing that matters is what comes out of the speakers.

Rather than starting off with specifics, In my previous post I thought it would be more helpful to try to attempt to explain my overall philosophy and how I think about it, as such.

I must admit I do feel that I am talking way above my pay grade here, but of course, I'm happy to share my thoughts if it might help:

For your snare example, I suppose my thinking would be:

- Get the initial volume balance right. I use volume to help the frequency balance too.

I know we are talking snare here, but perhaps it would be easier to use kick and bass as an example to explain. If I am wanting the low end of the kick to cut through, but the bass is masking it - initially I will listen to that low end of the kick I want, whilst I bring up the bass. As soon as I hear the bass overpowering the low end of the kick that I am wanting to keep, I will back off the bass volume. Depending on how present and which frequencies of the bass I want in the mix, as you say I will then start thinking about high pass, low pass, dipping out frequencies from the bass, or boosting others, through to compression, parallel compression, etc.

I suppose the route I choose really depends if I want the kick and bass to be equal players in the mix, or whether I want one or other to dominate at that point in the mix. Whether I need one or the other to fill out a particular frequency area in the mix, if I feel the mix as a whole is lacking low end, or midrange or whatever. I think that I usually aim for the kick to handle the real low end -say 40Hz ish, and the bass around 80Hz. I also might want, say 5khz of the bass, and 3Khz of the kick. It all depends on the mix and what is going on.

Maybe that might help with your question?

Back to that snare.. Where were we...

- High and low pass to remove low end rumble and get a smooth top end.

- Eq out any resonances or issues with the sound, odd room tones, or whatever

- Eq boost the freqs I want to hear more of, in context of the mix. Personally I like to boost more than cut. (Yeah I can already hear the shouts from the gallery. Whatever. You do you )

- Oh, also I suppose darker sounds tend to sit back in the mix, and brighter sounds tend to sit more forward. Certainly for me that 4k area really seems to bring things closer.

- Panning - One way of getting good frequency separation is to pan things wide. Have to take in to account mono compatibility too of course. I really only check this at the end of the mix. Speaking of which, sometimes listening to the mix in mono can help with eqing too. So I might have kick, bass, snare and vocal in the middle, and pan everything else around that. One thing I find is that sometimes panning something just a few notches off centre can make a difference. Oh, and also sometimes you can find a spot for something by panning around when the whole track is playing.

- Compression. Yes, you made a good point. Slow attack, fast release for the snare to give it punch to cut through. For other sounds yes, you might do the opposite to take the initial edge off, or perhaps play around with the release to sit a track back in the mix. Transient sounds do tend to cut through more static or sustained sounds and notes. Especially that initial attack.

- Reverb for sound (tone), impact and depth. For snare, I suppose I kind of think along the lines of a gunshot, or maybe tapping a pen on a desk (for kick). Or both. Not sure if that actually makes any sense though.

- Automation. If an instrument, or in this case the snare is getting a bit drowned out at any point, I will go through and automate the volume up (or perhaps down if it is too loud). How loud something is perceived depends on what else is going on around it.

- Eq cuts of other instruments, or busses that are masking the snare. I tend to do this during the latter stages of the mix as a sort of polish and fine tuning when I've got everything going in the mix. Yeah, as you say I might possibly dip around 8k or whatever. I also might dip around 200Hz too. For unmasking vocals I might dip 4k ish and 200Hz. I feel unmasking the lower mids as important as the upper. I spend a lot of time listening to what I feel I want to cut through more and make little boosts and cuts here and there. I'm talking about a db or so, or less. I probably tend towards making such moves on the buses, rather than channels. I suspect most might make the moves on the channels though.

- Master bus processing. I guess you are 'supposed' to tackle this in the mix, but for me it can sometimes make a huge difference, mix polish wise, if you target kick, bass, snare, vocal with a little eq on the master bus. Mix bus compression can have a big effect too, of course, so it's worth thinking about what you actually want to achieve with it. Sometimes a mix is better without it.

Actually, talking about compression, sometimes I will use drum bus compression, other times I will just use individual bus compression and no bus compression. Or just bus compression and no compression on individual channels. I might use parallel compression on the kick and snare, or on the whole kit, or maybe none at all.

- Parallel compression. I don't reach for parallel compression off the bat. I might parallel compress drums, or vocals as required.

- Mix rebalancing. In my template I have a set of 8 VCA faders linked to all drums, lead vocals, backing vocals, guitars, synths, strings, etc. I got the idea from how the SSL 4k is set up, so I can easily rebalance the mix. For example if I think the drums are too low in the mix, I can easily rebalance by eg lowering the volume of everything else, or whatever.

Really it all comes down to what you want to achieve - ie what you are trying to fix what is lacking in the track, and often there are a number of things to try. Sometimes one or the other might work, other times you might find you have to throw the kitchen sink and a lot of heavy clipping at a problem.

Well, I seem to have written a lot. Not sure if there is anything there that helps or not. Perhaps it might just confirms what you already know (which can be useful in itself I find).

Really for me, it's about learning (google ,youtube, etc to find out how the pro's do it), trying out different things and kind of building up your own toolbox of problem solving techniques. Eventually you'll hopefully find that you hit upon things that really work well for you in any given situation.

Of course, it might be worth posting here if there is any specific things, masking wise, that you feel you have an issue with regarding this mix, and perhaps I and others will be able to post possible ideas on how to go about solving it.

Cheers!

Edit: another thing that might help just occurred to me. I might use more than one instance of eq on a channel for different purposes. So I might have an initial eq that notches out some resonance, followed by another eq doing low and high pass, then another for boosts. I might do some more eq after compression. I find that I generally prefer to eq before compression, but compression can also 'eat up' eq moves, so I might add a touch of bass, or upper mids or whatever after the compressor too. Using more than one eq helps me to seperate out the purpose of each in my mind, and also allows me to switch them on and off easily too, to check I'm still wanting the effect of each one.

[mikej]

After all that, I somehow managed to miss your last point regarding the balance changes.

Yeah, you can think about eq as being like a frequency dependant volume control. So boosts and cuts will of course change the volume of an individual track, and therefore change the overall balance. As with any other processing you need to watch out and compensate for any level changes and re-adjust the fader (or output control) accordingly.

A lot of processors do seem to add a db or so of volume by default (because louder sounds better...), which is another thing to look out for too.

Cheers!

[SonicTramp]

Sure, man. I appreciate the philosophy you tried to pass on. I agree with most you wrote.

I care very little of how one shapes one's sounds. One can boost or cut as one pleases. It depends.

I cut more, mostly bc I obey the 11th commandment (watch George Carlin's how the Bible came up with 10 commandment). "cut is to make better, boost is to make different". Some wiser mixers already spent the time figured it out for me, so I just follow. I agree bc I understand the reasons behind the statement. I don't want to make any sound given to me different; I just want to make them sound better together. There's a difference, imo. Besides, economically speaking, cut costs less than boost in term of pain. Though faithfully obeying the "rule", I occasionally boost to get what I want. I find that 95% of my EQ work are cuts. Cuts, in my experience, helps separation easier and less hard on the ears. But like you said you do you, man, so I have a chance to do me. I have no problem with that. Do whatever to get what you want. Just make sure it's the right tool, the right move, and on the right instrument, right?

Though I know the reason why cut is better, I am still unsure of which one comes first? Sure I can boost and cuts but what do I do first? why? And does it even matter?

My current answer is: cut comes first. Bc cuts are to do the following: get rid of annoying frequencies, resonant, mud, shrill, harshness, masking, etc. Boost only gives me what I want initially but does nothing to clean out the sound, not to mention the possibility of potentially masking other instruments. I also think that it matters which one comes first bc it affects the mix as a whole. Boost should be in the "definition enhancing" step, which is a step right after separation, if there's a need for it at all, and it would be more effective after instead of before. Like sanding before wiping and then painting. One cannot wipe before sanding bc it would be a waste of time and effort, if not ruin the whole job, or one would have to go back and... rewipe. So, in short, I cuts only in the separation stage. And of course, rebalancing levels. It seems to me this step will be repeated throughout the mixing process. Always going back and forth rebalance levels.  I'd love to hear your or anybody's thoughts on this. I don't mind being corrected. Thanks.

[A.sch3]

Very interesting takes on the difference between cuts and boosts. I never saw it like that.

For a long time, i had a relative who doesn't know much about mixing, but acted like he was, and he often say that you should never boost, always cut.
My take on that statement was always that it didn't matters because if you gain compensate properly, a boost could be consider as a cut of everything else.

In the end i still think that the intention behind an EQ move is what's matter.

[mikej]

Yes! Exactly. It's the intention that matters. Knowing why you are making a particular mix move.

My answer to the question - yes for me, after balancing a mix I suppose I generally cut what I don't want first (usually any mud, resonances, rumble etc), then I boost what I want more of. Because I don't want to boost more of what I don't want. I boost more than I cut, because that what works for me. I mean, if you should never boost, surely the eq wouldn't have a boost knob? Towards the end of the mix I will maybe make the odd small dip here and there if something is covering up another track that I want to cut through. I don't find I do a whole lot of this and I find it's generally only small moves of like a db or so that's required. I think you can get a bit too hung up on this and do too much - and your mix ends up sounding weird.

*Edit: Of course I might add some cuts and the end or even the middle of a chain, you know if I add some saturation and it makes things too muddy. I generally eq before compression, but you might want to boost bass, top end, or whatever after it, if the compression changes the freq balance too much. Or if you are trying to add a big top end boost and the compressor is eating it up. Another common question is eq first, or compress first? Depends what you are trying to achieve. When you know, you know.

Really, I guess what I am saying is my view is don't worry too much about rules. Just do what is required to get a great mix out the speakers. Collect all the ideas, tips and techniques that you can. Add those things that work for you to your own personal mix toolbox so you can pull them out the bag when required. Build your own mix philosophy. If always cutting, never boosting works for you, then great!

Some examples of 'rules' that don't seem to work for me:

Never add drum samples because it's 'cheating'. Never boost, always cut. Always mix to -14 LUFS. (I think deep down everyone knows that the correct LUFS value for EDM and Pop-punk is -5 ).

I read somewhere that a top (pop) mix engineer always sends his mixes to the mastering engineer at around -7 LUFS so as not to give the mastering engineer much room to alter the mix too much.

Yes the streaming services might turn a mix down to -14 LUFS, but the mix as delivered likely won't be -14 LUFS, unless it's jazz perhaps...

Cheers!

[SonicTramp]

Very interesting takes on the difference between cuts and boosts. I never saw it like that.

For a long time, i had a relative who doesn't know much about mixing, but acted like he was, and he often say that you should never boost, always cut.
My take on that statement was always that it didn't matters because if you gain compensate properly, a boost could be consider as a cut of everything else.

In the end i still think that the intention behind an EQ move is what's matter.

* Thanks for making a comment. I am glad you join the conversation.

You should ask your relative for an explanation on his statement. Perhaps he'd offer some insights into it that might surprise you. I am a big believer in the Socratic method.

"a boost can be considered as a cut of everything else"...(You almost got me there, but let me put up some resistance)

Only if it's a zero-sum process, imo. But we know that this is not the case, not with sounds. Just bc one boosts, say, 3kHz, doesn't necessarily mean one cuts the same place to everything else even if one pulls fader down to compensate. The boost is still there and adding on to everything else. If boosting 3kHz masks the rest in the same place then you are right. But then a new problem is created at the same time: masking. Your thoughts?

Even if that's true, the question then becomes: is that a good economical move, getting one but sacrifice everything else in the process?

"In the end i still think that the intention behind an EQ move is what's matter"

I agree. I am not talking about choice; I am talking about process, the chicken or the egg, if you will.

[mikej]

Well, a boost is not quite a cut of everything else. If you did a cut instead, you'd have to raise the volume a bit as well if you wanted to match the boosted version.

Strictly speaking about the process - I would think any phase shift would likely be different in a boost than it would be in a cut. If you had an analog emulating eq I suppose any saturation or noise added as part of the modelling might be boosted, or cut too.

You could maybe test this by recording some pink noise and duplicate it on two channels. On one channel boost, on another channel do the exact opposite (and gain match). Then invert the phase on one channel and see what you are left with.

Some eq's also show the phase shift curve so you can observe the difference between a boost and a cut.

Cheers!

[SonicTramp]

Let me offer what I think about EQ first or compressor first. I think EQ first. Because it's the first tool we use to balance level in the beginning. Unless the recording is immaculate, which we all know that's not possible, EQ must be first in the chain bc of its capabilities, far more effective than, say, a compressor or a gate. Efficiency counts. I don't see the reason why anyone should put anything before the EQ. Don't we have to filter out the rumble/shrills in almost everything? Can a compressor or a gate do that effectively, more so than an EQ? Yes, to the former and no to the latter, imo. So yeah, unless someone comes up with a convincing argument I am sticking with it.

Samples as cheats? I used to think like that but not anymore. I was somewhat annoyed bc I am trying to learn and they have something I don't have in their mixes, samples. It made my work that much harder. And the snare, kick, and toms? They matter. So I make a point by mixing with only what I am given. That way if dudes listen to my mix they know they have what I have and to help make their job critiquing my mix easier. So no samples for me for now. I am not here to show dudes how awesome my mix is. I guess we all have diff reasons to do things. I am here to sort out things that I don't quite understand. Tbh, I look for disagreement more than agreement. I learn nothing on the latter. Hope I get it thru clearly.

[mikej]

Yeah - but a vocal might have been recorded through a compressor on the way in if it was too dynamic and overloading things... and I suppose you might want to emulate that by sticking a compressor on first. Just as a hypothetical example.

There is no cheating, and I think you should be here to show how awesome your mix is . I got your point about missing a trick, but you can always ask people how they are getting their sounds. In my experience people have been more than happy to share. I wouldn't worry about the people critiquing your mix as such.

Personally speaking although I think it's good to know the science, I also think you can maybe also get a little too hung up on certain tropes too, and your mix can suffer for that. That was really my point.

Cheers!