Yeah i will echo what Anthony noted and fill in a bit of color: Volume, EQ and Panning are the key. Firstly, not everything can be the loudest. If everything is equal, nothing stands out. This has to do with the writing and composition, but after that, things should sit relative not only to eachother but also with regard to where they sit in time within the song. Dont ever be afraid of automating volumes, such as turning up the strings during their solo or shining moment, and then having them decrease a bit in other sections. Keep in mind that volume is also how distance on the soundstage is percieved. If i have something alot lower, its going to seem further away intuitively, and so the reverse is true as well. Therfore, dont be afraid to turn it down. When you are concentrating on a particular sound during mixing, it may also seem louder to you. Dont be afraid to then turn it down a little more, to send it to the back. Likewise, reverb will send it to the back of the mix, and when things are desired to seem further away, you also want to take alot of the highs out of that reverb, since the ear does this naturally to reverberations that come from further away.Shay wrote:Hi all
I have a newbie question.
What is the best way to prevent a string sound from dominating the song?
I want the bass, drum and lead sounds to stay in front and the string far in the background.
It somehow blends with them and they become less distinct and clear.
Panning also plays a big part not only in things having their own sonic space and not overlapping too much, but spacially in terms of where things sound natural. That is, reality often plays a part in the way human ears percieve things, because we are used to space in natural life, so sometimes bowing to convention in this manner can help solve some problems. For example, a guitar players amp, flute player, bass player and keyboard amp cannot (comfortably) occupy the same space on a stage, and this wouldnt often sound great eiether, therefore it's a good idea to pan things where they would sit naturally in space. From this standpoint, even if you will end up creatively bucking convention and panning wildly later in the mix, i pays to at least start with a natural picture of the sounstage in your head, spacially; front to back, and left to right.
EQing is also essential, because many instruments overlap eachothers most prominent frequencies. Viola mite impede guitar, guitar may overlap an organ, electric piano or synth part, concert bass is going to collide with timpani, and so on and so forth. While some of this is good and adds harmonic coloration and a certain something that can only pop out when multiple instruments play together, too much will mask eachother or cause unpleasant phasing. Therefore, this is a game of sacrifice. I might really like the lower stuff that the string parts are giving off, but if they are runing the low mids of a bass track or vice versa, the point is lost anyway. Maybe i do this through composition, removing or changing a bass part to let that shine through. Otherwise, i'll have to eq that lower end of the strings so it is not as prominent.
Likewise, if certain string parts are killing each other, then the eqing must also speak to the 'important' part of what that instrument resonates at and how you want it percieved. Often many many instruments are playing in a stereo recording, we cannot discern alot of each instruments frequencies clearly enough for them to 'matter', in a sense. If the important part of my 808 kickdrum is 60hz in a particular record, then i can't really also make the bass track thump at mainly 60hz, or they will just muddy eachother up. Be willing to sacrifice some of the sonic richness of each of two or more instruments that are overlapping too much.
If we're talking about violin, you may find that in your particular mix you could roll off (ie: basically get rid of) everything below 200hz. Sometimes i roll higher end strings, and even electric guitars, off at 300hz. Sure, they lose some harmonics. But then again, if i can barely hear them, they are not important. If it's Viola, you may want to keep some more down there, Cello needs a little more lows etc. But likewise, if its the mids or highs that are offending, say, overlapping with flutes, guitars, the choppy grit that may be important on an electric bass...whatever the case may be, you've got to pick your poison.
One of the most important lessons our teachers ever gave was just how important reductive or subtractive mixing really is. This is applicable to not only EQ, but also volume or gain. This simply means that whenever you mix, be it volume faders or in manipulating EQ frequencies on a given instrument track, you should start by taking away what you dislike, long before ever adding anything. The reasoning is simple: because head room is limited. Headroom is the total amount of gain available to you before distortion, and so when all our instruments are playing in the mix, we want this near the top, but not over (at least on a digital meter). Additive mixing, or adding pleasing frequencies and volumes, fills this up quickly, due to the ear candy principle. Lets say there is an offending frequency in that string part you mentioned, and the pleasing part is oh say, right in the middle, around between 300 - 500hz, or something like that. If you boost that, you will get more of what you want. However, now youve increased the volume of that track overall, and used up a little headroom overall. Later, you find pleasing frequencies on other tracks. You start boosting things like 3k on the snare drum, 250 on the kick for the beater, 1k on the flute to give it more presence, and so on, until guess what? Youve got to turn your whole mix down to keep from peaking. But now, youve gone and lowered a bunch of other frequencies, by turning things down, which you may have not wanted to do. The simple way around all of this is, target things in each track which are clearly displeasing, or very succintly in the case of your strings example, overlapping other instruments. Check out some tutorials on notch filtering, tightening and widening the 'Q' (target freqency) of a given eq band, and you will have some of the tools you need to target offending frequencies, for whatever reason you wish to reduce them or even take them out completely. Often only subtle reductions are needed, but dont be shy about drastic cuts or rolling all the highs (research: bass rolloff and bandpass EQing) if you truly need to. Make a big spike in an EQ band with a tight Q, and sweep around (dont listen do loud, you can blow your speakers and/or eardrums!)...you will find that offending frequency and learn alot in the process.
After lowering the volume faders of instruments you want lower, and reducing frequencies of certain instruments you want lower, you then go back and can do subtle boosts. And since you've reduced some unwanted frequencies, the power of small raises in other frequencies will be increased, so you save headroom on both sides of that coin.
I hope some of this is helpful, and please excuse my typical longwindedness; i get excited about this stuff! Plus, i drank a ton of coffee...
By the way, type of song or recording is this string part in?