subsekt

In my last tracks ive been focusing alot on EQ, and giving each sound its own place in the spectrum. And though im sure ive overdone it and destroyed the sounds, it seems like the way to go, maybe i just need to dial back a little and get better at it. The thing im wondering is if this is common practice in techno production, to do some rather hardcore EQ shapes? Im watching and reading a lot about mixing and how you should EQ as little as possible, but that usually applies to other genres with live instruments, and of course you will mess up the timbre of for example a guitar if you EQ to death. But if im working with a more abstract sound i cant really see the problem if it still sounds good? Ive been experimenting with making sounds rather flat in the spectrum, that usually alters the timbre quite a lot and sometimes take some life out of the sound, but some distortion and/or compression can usually bring some character and life back.

Am i falling in to a bad habbit or am i just stating the obvious (if it sounds good it is good...)?

As long as it sounds good who cares about how the shapes look like. Just be aware that an eq change on one track has effects on everything else. Btw. if you have not read it already.. bob katz mastering audio, I like his EQ philosophy in there and it seems to work

Great! Thanks for the tip, will look it up!

One more thing btw. Ive heard mix engineers talking about focusing on the midrange, from like 400-4k, since thats the range that will translate most equally to all systems, but in techno its common practice to lowpass, or highpass some sounds out of that range. Is there more common practice in mixing that doesnt apply to techno?

EQ for sound design = do what you want.

EQ for mixing = trying to balance elements so they don't clash. If you have 2 or more elements thats overlap heavily you may have to EQ to a point where it destroys the point of 1 or more of the sounds. Sometimes clashes are better dealt with through arrangement (don't play them at the same time) or choosing different sounds entirely.

I think when you mix your own stuff you can avoid a lot of troubles when you choose the right sounds. If your arrangement works it will nearly mix itself.. I would not look at stuff like what works in what genre or whatever.. Its all about what you want to achieve. Imho mixing is mixing no matter what genre.. you can learn from pop mixers (like pensado or whatever) and use stuff he talks about on techno and also the other way round.

Just don't copy settings 1by1 and listen to what is changing when you turn a knob. It will take a while until you will get better and it can be frustrating at times. Specially when you know stuff in theory but still struggle to implement it. Just keep on going and work on stuff.

For experimental shaping of sound sources you can go crazy with EQ

To EQ for every other purpose (mix balance, pronounce sounds, create depth, adjust space, warmth, boxy-ness etc) it's important to not overdo it and do comparisons in context.
Overdoing the EQ creates problems you probably don't grasp initially and then you try to compensate for the problems you created with EQ by applying more EQ (that wasn't needed)
on another source or sound and so on. The most important thing is to always listen in context and to do proper volume matched comparisons.

It should only be "common practice" to lowpass and highpass if it's ever needed which it's not as often as people think.

pimo wrote:Ive heard mix engineers talking about focusing on the midrange, from like 400-4k, since thats the range that will translate most equally to all systems

Techno/dance music isn't soley made to be played on the radio. The very low end is extremely important and needs to be right, as does the very high end. You should concentrate on making the whole mix as good as you can, but remember the genre is designed to be played loud in a club or similar.

Mattias wrote:For experimental shaping of sound sources you can go crazy with EQ

To EQ for every other purpose (mix balance, pronounce sounds, create depth, adjust space, warmth, boxy-ness etc) it's important to not overdo it and do comparisons in context.
Overdoing the EQ creates problems you probably don't grasp initially and then you try to compensate for the problems you created with EQ by applying more EQ (that wasn't needed)
on another source or sound and so on. The most important thing is to always listen in context and to do proper volume matched comparisons.

It should only be "common practice" to lowpass and highpass if it's ever needed which it's not as often as people think.

Would you say i should put more work into programming and adjusting my patches, and selecting my samples rather than sculpting the sound with EQ? Or that overlapping frequencies are not as big of a problem as people think?

And would it be good practice to do radical EQ early in the fx chain?

pimo wrote: Im watching and reading a lot about mixing and how you should EQ as little as possible, but that usually applies to other genres with live instruments, and of course you will mess up the timbre of for example a guitar if you EQ to death. But if im working with a more abstract sound i cant really see the problem if it still sounds good?

^^this.
It's VERY important to realize that in a "normal band" situation, all instruments will mostly occupy their own part of the frequency spectrum,
give or take a few exceptions (for example a piano, which does have 88 keys, so it covers a massive range).
In our case, we are making sounds ALL over the frequency spectrum, so I think it's almost a must to at least cut a certain minimum just to make sure there aren't too many clashing frequencies.
No doubt it's even better to have frequency clashing sounds play at different moments in your track,
but I have a tendency to put in too many notes (since I come from a typical solo instrument because I play piano),
so I usually solve it by cutting frequencies with EQ.
I'm even quite rigid about it : what isn't part of the pure identity/character of a certain sound, I just cut it.
It opens up my mixes.

I know I should probably just arrange stuff better, but I guess a lot of this comes from my background of playing solo or in a typical band setting,
where you never really have to deal with clashing frequencies as long as you play all your notes in the right key.
So that is what works for me.
No doubt people like Mattias are better positioned to give decent advice about this.

pimo wrote:
Would you say i should put more work into programming and adjusting my patches, and selecting my samples rather than sculpting the sound with EQ? Or that overlapping frequencies are not as big of a problem as people think?

And would it be good practice to do radical EQ early in the fx chain?

I think all of this is different for every track you make.
It's always nice to have the "possible overlap" thing in the back of your mind while selecting your patches/samples,
but I also think that overlapping frequencies don't always have to cause trouble.
A little overlap won't hurt, I think, but exact overlap on exactly the same moments/notes,
now that's most likely gonna give you trouble.

Hades wrote: I'm even quite rigid about it : what isn't part of the pure identity/character of a certain sound, I just cut it.
It opens up my mixes.

I have the same habit… I usually put a HP filter on almost every track (even though I know it’ll mess with my phase). I also had a discussion with someone about it recently, and we ended up doing a little blind test between high passed and non-high passed mix. You can’t tell a difference if one or two tracks are high passed or not, but with all of them together you could definitely hear that the non-high passed version sounded a bit more alive. It’s very subtle, but it’s still something you could hear in a blind test. That said, it doesn’t mean that getting rid of unused frequencies isn’t a good idea… usually it is - just don’t make a habit out of it (which is hardly a good thing).

@OP
Personally, I avoid radical EQing as much as possible. I do work for film, where super weird EQ curves and boosts with 8db+ and beyond are usual… so I’m actually used to it and I know that it can be done and that there’s a place for it, but as long as you have the ability to fix the problem somewhere else, then fix it somewhere else. And there are quiet a few ways to fix stuff pre-eq:

- Sound/Sample choice
- Sample pitch
- Arrangement
- Volume envelopes/automation and fader riding
- Sidechaining

pimo wrote:

Mattias wrote:For experimental shaping of sound sources you can go crazy with EQ

To EQ for every other purpose (mix balance, pronounce sounds, create depth, adjust space, warmth, boxy-ness etc) it's important to not overdo it and do comparisons in context.
Overdoing the EQ creates problems you probably don't grasp initially and then you try to compensate for the problems you created with EQ by applying more EQ (that wasn't needed)
on another source or sound and so on. The most important thing is to always listen in context and to do proper volume matched comparisons.

It should only be "common practice" to lowpass and highpass if it's ever needed which it's not as often as people think.

Would you say i should put more work into programming and adjusting my patches, and selecting my samples rather than sculpting the sound with EQ? Or that overlapping frequencies are not as big of a problem as people think?

And would it be good practice to do radical EQ early in the fx chain?

It's so hard to give a solid "right" answer where everything is entirely program dependent when we create our music.

As stated many times over the year around here "the best EQ is no EQ" so yes, you should spend as much time as possible to tweak your patches and samples.
However this may include using EQ to sculpt the content, either to adept it into the substance of the music or to get rid of unnecessary information that makes the
sound spectrum too overloaded, fuzzy, harsh or any other reason.

Many times you end up using EQ for the reasons above and then use a new instance of EQ post all other processing you made for the actual "mix" duty.

Generally speaking, I claim that you should rarely need more then 3 to tops 5 bands (if even that) to adjust a sound for mix duties (not for sound creation).

One thing I realized recently is that most of the time, if a sound isn't really fitting you just need to tighten up the envelopes instead of going crazy with EQ. Even just reducing the release by a bit can drastically reduce the amount of frequency information your sound has.

Mattias wrote: Generally speaking, I claim that you should rarely need more then 3 to tops 5 bands (if even that) to adjust a sound for mix duties (not for sound creation).

damn, that's always the maximum I ever do.
and here I was thinking I might have been overdoing it.

oh yeah, to OP,
just because no one has literally said it :
mid/side EQ'ing !!

the most typical use for this is of course with pads to your side, and kick and bass in the middle,
but I have had several tracks where I found myself using that trick with other sounds as well.
Let's say you have 2 different kind of lead sounds with slightly overlapping frequencies ?
put one on the side and the other in the middle using that mid/side trick.

Not sure if there aren't any mastering experts that would say "no no NO !!! Do not ever do that !!!"
But I wouldn't know why not to use that same trick on other sounds, and it has worked for me several times in the past.

dubdub wrote:One thing I realized recently is that most of the time, if a sound isn't really fitting you just need to tighten up the envelopes instead of going crazy with EQ. Even just reducing the release by a bit can drastically reduce the amount of frequency information your sound has.

Easy way out of this stuff is some times adding the transient designer and turning that attack and sustain knob.
(if it's already audio, this is probably the only option unless you load your stuff into a sampler)
Fucking saved my life so many times...

i wish i had the knowledge on eq and all that "Mastering"stuff
still have to finish my book:D
but playing with eq´s can bring your track on the next level.get much more out of it as you thought(sometimes)
but also takes a lot of time.
lovely to read all about in here.

Resin wrote: - Sidechaining

I actually barely use any compression for the normal oumph thing, (I hardly make any oumph music anyway)
but I use compression all the time to sidechain stuff so to make room for all my sounds going on at the same time,
and meanwhile creating more groove (which I suck at most of the times)

It's not an exception for me to have 3 sidechained compressors going on to tame just a single element in my track.
and I'd have like 3 or 4 elements in my track with 3 or 4 sidechained compressors...

Hades wrote:

Mattias wrote: Generally speaking, I claim that you should rarely need more then 3 to tops 5 bands (if even that) to adjust a sound for mix duties (not for sound creation).

damn, that's always the maximum I ever do.
and here I was thinking I might have been overdoing it.

oh yeah, to OP,
just because no one has literally said it :
mid/side EQ'ing !!

the most typical use for this is of course with pads to your side, and kick and bass in the middle,
but I have had several tracks where I found myself using that trick with other sounds as well.
Let's say you have 2 different kind of lead sounds with slightly overlapping frequencies ?
put one on the side and the other in the middle using that mid/side trick.

Not sure if there aren't any mastering experts that would say "no no NO !!! Do not ever do that !!!"
But I wouldn't know why not to use that same trick on other sounds, and it has worked for me several times in the past.

I stay within that range of 3-5 bands aswell for mixing, so its great to hear that im not doing anything strange!

I have recently started testing some mid/side eq, and as you say, usually for pads, and usually just rolling of the bottom from the sides. But i will explore it more. Does anyone have some tip where i can read and understand more about mid/side processing?

pimo wrote:
I stay within that range of 3-5 bands aswell for mixing, so its great to hear that im not doing anything strange!

I have recently started testing some mid/side eq, and as you say, usually for pads, and usually just rolling of the bottom from the sides. But i will explore it more. Does anyone have some tip where i can read and understand more about mid/side processing?

honestly, I have never read any tutorials on this.
I've barely read any tutorials in years, because 80% sucks and as you put in more hours,
you'll seldomly come across decent ones as the years pass.

But it's fairly simple, really.
Let's say you want to use a sound and it's taking up a lot of frequencies.
Turn on the mid/side option in your EQ
Take out pretty much all frequencies in the mid band,
leave just what you want/need in the side band.

That's what you'd typically do with a pad that gets mid/side EQ'd.

Now, what I've learned over the years, is that some times, when you do it this radically,
you might find your final sound to feel really lifeless, as if you over EQ'd something the normal way
(because I find that in general sounds tend to sound lifeless a lot sooner when EQ'd purely to the side with the mid/side thing than if they were EQ'd just the normal way)
So, you might wanna bring back just a tiny bit of the real important frequencies of the sound in the mid bands.
Nothing too much, but just enough to breathe back some life into the sound you just tamed.

And yeah, you can do this with any sound if you want to (afaik).

...If there's stereo info to begin with! ..