Dealing with clashing frequencies

[TrevCircleStudios]

Yep I agree with Frank (as ever). When I run live sound I put a mono mix through FOH too. If you don't people on the extreme left will miss anything panned extreme right and vice versa.

[macleodgrant]

there's a couple of things i've noticed. for the most part instruments are designed to work quite well together if you are able to replicate their real world sound in the computer - this is obviously different for different types of music.

but as far as drums, guitars (elec or acoustic), vocals & bass, once eq'd correctly you should have space for the all the instruments to sit nicely without needing to resort to automation and it should work in mono too. i attend to this before adding affects or anything like that (i do however ignore the mono thing quite often).

if you can get the track sounding clear with everything being heard clearly (or veiled if that's your thing) and incredible in this state then you have a winner since once you add the effects & automate you're sorted.

Elec Guitars - naturally amps have a limited bandwidth so it's quite safe to low pass everything above 5-4k with a semi-gentle slope - this is particularly important if you are using amp simulation. to increase the air around the guitar, rather dial in some nice early reflections and short verb from a good reverb and increase it's high freq decay - powercore verbs are amazing for this

Vocals - i hate that annoying 2k area on some voices so i take a broad'ish Q and cut around that area quite a bit. while doing this it makes the clarity of the vocal stick out and adds brightness & air above 4k. i high pass quite a bit too until the fundamental with a gentle slope

Acoustic Guitars - i firstly put low pass at about 16k to remove the sizzle and makes the guitar sound more like an acoustic would. i also cut at about 2k so it doesn't clash with the vocal - even though i've cut at the vocal, it's still got a bit of 2k. high pass as much as possible on the acoustic that you can get away with. hollow out at about 600k broad Q

those are the main instruments that cause problems, drums are easier to get in their own space

[TrevCircleStudios]

I absolutely agree with your first couple of paragraphs macleoudgrant. I'd go as far as to say that I usually try and get everything to sit pretty nicely in mono before panning (YMMV) but I gotta pick up a few points here as it could lead some people with less experience to do things which they have absolutely no need to do:

Elec Guitars - naturally amps have a limited bandwidth so it's quite safe to low pass everything above 5-4k with a semi-gentle slope -

Er, no. If there are some frequencies that are annoying or clash, absolutely cut them using the method I've described above. For the most part, however, electric guitars have good and interesting stuff up to about 8k. Unless you are going for a particularly rolled off sound on a particular track I wouldnt roll off much below 8k.

Vocals - i hate that annoying 2k area on some voices so i take a broad'ish Q and cut around that area quite a bit. while doing this it makes the clarity of the vocal stick out and adds brightness & air above 4k. i high pass quite a bit too until the fundamental with a gentle slope

Erm, again that's a bit broad brush. It might work on a one particular vocal in one particular song or for a particular vocalist but making broad cuts of anything as a matter of course is going to leave you with a poorer product on 7 out of 10 occasions. You need to do what works for that vocal in that song.

Acoustic Guitars - i firstly put low pass at about 16k to remove the sizzle and makes the guitar sound more like an acoustic would. i also cut at about 2k so it doesn't clash with the vocal - even though i've cut at the vocal, it's still got a bit of 2k. high pass as much as possible on the acoustic that you can get away with. hollow out at about 600k broad Q

Again, a bit broad brush. Some of it might be worthwhile, some of it only in some songs but, in others, you'll only be doing damage to the end product.

Sorry Mac and this isn't intended as an attack on you (though it may sound like nitpicking or you may think 'that's the way I do it and that's what works for me') but there are a lot of newbs out there who grasp at things like this and see them as gospel. (Like the godawful 'preset' culture (I know one famous producer who recently gave away his 'secret presets' to be used on a well known plugin. Newbs the world over are now thinking, well if these presets work for him they must be great. He confided in me over a beer or seven that he simply rolled some dice for a laugh.))

An engineer told me years ago that there's nothing under 100hz worth keeping aside from on bass and kick tracks. I religiously followed this for years until I was working with a 'name' producer in NYC last year. He asked my on earth I was doing it. I said I dunno I just always did. He told me to listen with my ears instead of doing things blindly. My mixes improved overnight.

So in short my reason for picking you up on this is simply because I think it's our job on resources like this (especially since they make it to such a wide audience) to make sure we don't spread disinformation like that old engineer did.

T

[EdRyan]

The preset culture is awful lol. For basic tasks that can be time-consuming, I can see why they have some benefit. But it's gone too far with sample loops, synths and everything in-between.

Visual cues, like RTA's in eq plugins, can be useful to show roughly where the energy in a sound source is. You can then use this to eq off the areas that you don't seem to need as much, and make space for other elements in a mix.

But as I found, you should never rely on them, or any preset. I've been fiddling with an old remix I did tonight and managed to make it sound a lot better by using a different eq, and setting it up myself. The first mix was ok, but I can't believe how much better it sounds now..

If you have the outputs, try setting up a b-mix as well. That way you can hear the changes you are making, and can hear if you are going the right way

[sureno]

i looked up this thread as i wanted to know what the definition of a clashing frequency is, because some times something may bother me in a mix but not sure what, what should i be looking for or really listening out for? i read what Trev said in regards to boosting a frequency at a sharp Q and sweeping it in the suspected region but what is a tell tale sign that a frequency is clashing, sorry for the noob question but im still learning guys

[FrankGIK]

Originally Posted by sureno

i looked up this thread as i wanted to know what the definition of a clashing frequency is, because some times something may bother me in a mix but not sure what, what should i be looking for or really listening out for? i read what Trev said in regards to boosting a frequency at a sharp Q and sweeping it in the suspected region but what is a tell tale sign that a frequency is clashing, sorry for the noob question but im still learning guys

What we're talking about here is creating separation between mix elements so that everything blends but can be heard distinctly. You'll know when something is fighting with something else because one element will overwhelm another. Some very common examples would be kick drum and bass, guitars and vocals, one guitar and another guitar, etc.

The best fix is to track them distinctly from the beginning. Make sure they sound a little different, i.e., use a Tele on the bridge pickup on one side and a Les Paul on the other side. If you want low end thud on a kick drum so that it'll pump underneath the bass, make sure you tune the kick right before tracking. One of the biggest mistakes I see all the time is guys trying to EQ in what was never there to begin with. You can't make a 9" metal snare sound like a 12" wood snare by adding 8dB at 250Hz...you'll just get a muddy, pangy metal snare.

Make sense?

Frank

[sureno]

cool thanks for that i can just about get my head round that, because i don't record much in essence i just have to choose the sounds i use from samplers and V.I's etc with more attention to detail, such as frequencies. sometimes i find it hard to get as much separation as i want out of a mix, i try eq'ing parts to no great avail, so i suppose after reading your reply the actual sounds could share parts of the main frequency right?

[FrankGIK]

Originally Posted by sureno

cool thanks for that i can just about get my head round that, because i don't record much in essence i just have to choose the sounds i use from samplers and V.I's etc with more attention to detail, such as frequencies. sometimes i find it hard to get as much separation as i want out of a mix, i try eq'ing parts to no great avail, so i suppose after reading your reply the actual sounds could share parts of the main frequency right?

Absolutely....it's inevitable. There are some parts of the spectrum where there's a whole lot of overlap (100Hz, 150Hz, 200-250Hz, 1Khz, 2.5Khz, 5Khz, 8Khz, 12Khz), and there's nothing you can do about it. The attack of kick drum hitting a new head is going to be very close to the clicky attack of bass strings, for example, so if you need to hear both you're going to have to get creative. A guitar and a vocal overlap between 1-2Khz, so the way we usually deal with it is to pan the guitars away from the voice...if it's a singer/song writer thing where the vocal and the guitar need to be right on top of one another, that's when you need to really be aware during tracking.

If you're dealing with sampled sounds then you have a lot of latitude creatively in terms of what they sound like or where they go, right? Decide that right from the start...what will everything sound like? What's my vision for this mix? If you don't know that, you've got no road map. You'll wander around aimlessly, print mix after mix and never be satisfied because you don't know what "satisfied" sounds like. After you know what it's supposed to sound like, place things in the stereo field where they make sense and tweak them until they sound right.

Try to learn the spectrum by description. For instance, 250Hz is "muddy", 500Hz is "boxy", 2.5Khz is "tinny", 5Khz is "harsh". That way, when you have a thought like, "wow, that sounds harsh" you know precisely what to do...cut some at 5Khz. If it's not *exactly* at 5Khz at least you know where to start, right?

Frank

[sureno]

Frank! thanks dude, iv learnt a lot by this, wow. going to completely re address the way i approach a track now and learn to start associating sounds with frequencies, i know this will take a long time but now at least i got a clearer direction before i would aimlessly mess around with the eq till i found something acceptable now im going to try and predict where the problem frequency lays, think i will just be experimenting for a little to get used to all these types of sounds and just try different things

[TrevCircleStudios]

Frank is spot on again.

now im going to try and predict where the problem frequency lays, think i will just be experimenting for a little to get used to all these types of sounds and just try different things

Absolutely right. Get the core sounds and arrangement right and the song will mix itself (because then there's no need to polish the turd!)

[sureno]

Great im quite excited now, i got a better idea on what to try and do i wonder if it all becomes clear to me now (couldn't resist) is it worth/advisable using spectrum analyzers when choosing sounds to help me see where the main body of frequencies are, i know i want to just be able to listen to sounds and just know what frequencies it consists of but i have a feeling that will take me a long time to acquire precisely

[TrevCircleStudios]

http://www.the-low-end.com/t/goto/bonuslist.php

I'd only say yes if it was for a defined short period. (Can you borrow one for a month and then give it back?) What you need to do is train your ears and the longer you rely on one of those things the harder that will be.

[FrankGIK]

I agree with Trev...only use an SA as a learning tool. It's very easy to let it become a crutch, and before you know it you're trying to mix with your eyes. Don't do that.

Frank

[sureno]

thanks guys, yes im fully aware of that, i would really just use it to put it simply learn what certain frequency characteristics are rather than as a go to tool every time i want to know a frequency band, i eventually want to be able to listen to a bass sound for example and have a rough idea where the main body(s) of frequency lay so then i could may be choose the other sounds around that and hopefully give each sound a rough separation, thats when i use an EQ right? to polish off anything i feel could be clashing unnecessarily, i should start by listening to and EQ'ing basic sounds right? maybe like mention earlier a bass and kick drum? would i be looking for the "presence" of the sound i.e where does that sound seem to be most prominent frequency wise? and choose them like that?

[TrevCircleStudios]

would i be looking for the "presence" of the sound i.e where does that sound seem to be most prominent frequency wise? and choose them like that?

No not necessarily. This is where mixing with your eyes will take you off course. What to emphase/cut is the sound characteristic you want/don't want. Find that by using your ears by sweeping the EQ like I explained earlier til you find the bit you want/don't want. Beware of your eyes - they'll tell you lies (must be a song in this)!