Ford Van
Banned
LOL
I was going to save this for an article, but decided after some recent listening to various projects done by home recording engineers that I would "rant" this in this forum.
I get asked the questions "What frequencies should I cut or boost for this or that"? Or, "What should I use compression on"? Or, "How do I get my mix loud"?
While many books could possibly be written covering these subjects, I am going to share some thoughts about mixing right here that I think will serve most well. Believe it or not, this is not going to be that long of a post.
1- From listening to many home recorded songs, I have come to the conclusion that many are monitoring at levels far too low to mix effectively. There may be issues involved with turning it up a bit, so it is not that I am without sympathy for those reasons, but mixing at low volumes means you will have the tendency to mix too much low end into your mix. Equalizing the monitoring system is not going to solve this problem. Turn those puppies up!
I would recommend that you record a 1KHz test tone on your system, and play back that test tone through the same D/A converters you will use to monitor your mixes. With the test tone recorded at -6db digital full scale, use a dB meter and turn up your monitors until that test tone is at about 85db with a C weight on the meter. I am not kidding.
It is a good idea to do one speaker at a time so that you can see if possibly you have a bad side of your amp, a speaker that is different then the other (it can happen), or bad cabling. If there is a difference between the two sides of the amp, suspect your cabling first, either the amp or D/A converters second and third, then your speakers.
2- Your near field monitors are NOT going to have as much low end, and will have a tighter low end then home stereo speakers will. You HAVE to get used to this difference in the sound. A great way to KNOW the low end of your monitors is to listen to a lot of professionally recorded mixes using the same D/A converters and monitors. There is no other way that allows you to be as subjective. If you are recording on a DAW, find a way to digitally transfer professional mixes to your DAW so they are available to play back.
3- Blending instruments is nowhere as hard as you might think! It only requires that you pay attention to what you are doing, and sadly, I don't hear many people doing that.
Let's take for example that the hi hat and the vocal are going to have some similar places in the mix. This is not uncommon in Pop music. The sibilance of the vocal is going to be about the same as the hi hat frequency. Make sure that you don't have one or the other dominating in the mix!
A kick drum and a bass guitar pluck should have a very similar sound too. Both instruments require that you rid them of low midrange frequencies, like around 200Hz. Solo both and check to make sure that they are not killing each other. While a kick drum can get a little click sound for definition with a eq boost around 3 or 4KHz, the bass guitar will get articulation with a boost around 500-1000Hz. But both are going to have a similar 100Hz body to them, and pay very close attention to that. Also, don't get over-kill on the stuff below 100Hz. With most monitors, it is hard to hear anything below that, and you will have the tendency to boost 80Hz a lot to try to make these two instruments sound fuller. Be aware of that because a 3db eq boost on a kick drum at 80Hz is going to eat up a lot of your potential dynamic range in the mix, and it may not sound as good as you think once you start playing the song on home stereo systems
In the midrange, watch for the vocals to have harsh words that jump out at you. One of the problems many face is that they are performers too, and their midrange hearing starts to get shot, and they don't quite hear those 2-4KHz jumps in the vocal that kill your ears through other converters and play back systems. Also, the guitar is going to have a lot of it's top end in this range, and you need some room for that.
With guitars, watch those boost that are below around 250Hz. Anything below that is not worth boosting on a guitar. You once again are just eating up potential dynamic range in the mix with a boost that WILL not sound good on home stereo speakers. If your guitar sounds too bright, try cutting stuff above 5or 6KHz. I roll out a lot of 10KHz high cut filter on guitar, unless they sound very dull, in which case, you need some boost there to give them "air". At no time is a boost below about 250Hz good for your mix with guitars. Don't do it and be happy in the end!
When you are done with your mix, take a good close look at it in a .wav editor. Do you have a lot of spikes in the mixes that are 6db hotter then the rest of the song? Is your average level well below 50% of the available dynamic range? If either is the case, I assure you that your mix is not good by any means. Those 6db jumps mean that you needed some limiting on a certain instrument, or that you have a eq boost on something and it is causing the track to be super sensitive to that frequency. Say with a guitar, those big low end boosts means that when ever the instrument hits a note with that fundamental frequency you get that much more boost at that frequency. Not a good thing. Eq boosts are only really good for deficiencies in the track. I revert back to if you are trying to use eq to make something sound drastically different, then your best bet is to record the part again and get that type of sound recorded. This way you KNOW what you are getting in the sound.
It really comes down to paying attention to instruments that share certain frequency ranges. The click on a kick drum shares the same space as the upper end of guitars, and the bulk of the energy in vocals, but the kick drum is not playing continuously, so it is okay to boost that up a bit to bring it out. But boosting an guitar at that frequency, or even a vocal means a mismatch with the other instruments in that range. We are looking for balance. If you are finding that something is dominating in the mix in a certain frequency range, check to make sure that you don't have boost eq in that range that is contributing to this. If you don't have boost eq on it, then you may need to apply a little CUT eq, but just enough to mellow it out, which is usually around a 3-5db cut. Any more then that and you have a big problem with the way the track was recorded. But another consideration too is how important any instrument is in a certain frequency range. A vocal masking a guitar a little bit is okay because usually, the vocal, drums, and bass are important during vocal passages.
If you are using digital eq's, any more then a 6db cut or boost is going to sound awful! The algorithms are just not good enough to provide a smooth sound and color to a track with excessive eq, so large cuts or boosts with digital eq's should be avoided at all costs. With cheaper analog eq's, you have a similar problem. Are you catching on here? Get you tracks close to the way you want them to sound while tracking and avoid heavy eq.
Before mixing a song, it may be a good idea to run a mix where you use no effects or eq and just adjust volumes to get the best mix you can. Burn that to CD and start listening on many different systems. You may be surprised at how good it actually sounds on home stereo systems, and how much more open and unprocessed your mix is. You will certainly still have a few things you would like to fix, but now you KNOW what needs to be fixed, and are not being biased by the better sounding D/A converters and monitors your setup has.
A great way to think about blending instruments is considering the importance they play in the mix. Obviously, a vocal 99% of the time will be very important. But that doesn't mean that the vocal should over power everything else. I call this "oblique" mixing, where there is an underlying theme that stays consistent, and you have parts that build over that but don't interfere too much. If fact, if you consider what oblique means, mixing a song effectively is very similar in the dynamics. You have a very steady bottom with a little color on top. The color grows and shriks, but you still have a solid foundation that is not really effected by it. This means you really have to consider what each instrument REALLY contributes to the song. Also, "moods" of certain parts should be reflected in the dynamics too. A little guitar ditty during a verse doesn't really need to be blasting away does it? But during a build up at the end of a chorus, you may have a lead that comes in for the last measure to embellish the transition. You need to leave some sonic space for this add in, and leave enough headroom for the added volume.
In Pop tunes, the kick and snare drums are very important, and should be very articulate in the mix, along with the vocals. The bass guitar is usually the harmonic foundation, and should occupy the low end of the mix predominately. Guitars tend to sound sort of quiet in the studio, and at lower volumes. Be wary of that. But, solos should occupy the same dynamic and sonic place the vocals did in the mix.
Strive for when a instruments stops playing that your mix doesn't drop in volume much at all, if at all. Viewing RMS metering of some sort will help a lot with this. If you know how to view peak metering for the RMS value (a skill that takes awhile to learn) even better because you can also see just how often you have used up your dynamic range.
One thing I have noticed a lot is people trying to put extreme dynamics in a song that doesn't need extreme dynamics. Drummers are by far the worst about this in recording, followed closely by singers. Consistent volume levels are desired in recording. How the parts fit together, and to a certain degree, the actual notes being played will create natural sounding dynamics in Pop tunes. Also, an instrument stopping in the middle of the song for whatever reason creates a change in "perceived" dynamics, even though your metering may not show anything different.
Another thing I have been hearing a lot is either sparse or heavy reverbs. Don?t be afraid to use a nice reverb across several rhythm instruments. It is the glue that holds them all together. At the same time, big huge long reverbs all over the vocal line don't work for anything other then Whitney Houston tunes, or the Scorpions? What I have noticed about the big reverbs many people use is that they leave in a lot of high-end content in the reverb. Hey guys, try cutting everything above 3 or 4KHz on that reverb. Also, I have seldom EVER used a reverb recently where I didn't have at least 40ms of pre-delay on it. A very short decay, with a hi-cut around 3 or 4KHz and a pre-delay that is appropriate will really fatten up a track without having to resort to eq! Play with it and you will never use those huge 3 sec plates with all that 10KHz content again!!!
Use your ears AND your eyes to gauge your mix. Your ears will tell you when thing sound screwy, but viewing your mix in a .wav editor will certainly "show" you some potential problems. View a professionally recorded song in a .wav editor and you will usually see a very well developed .wav, with very consistent peaks on the kick and snare drums, and you won't see the volume jumps when the vocal comes in, and drop when it goes away. Creatively using eq can help tame dynamics a bit, and gentle compression on instruments will do the same.
Anway. Just a rant so to speak about mixing. If it helps anyone, great! If not, at least you have some more material to curse my name by?
Ed
Back to running a mix without effects and eq.
You need to know WHAT in the mix needs to be fixed before you do anything. I have seen people start eqing stuff that sounded great because that is what they normally do to that instrument. My oh my..... The idea in recording is to get as close to the sound you want while tracking. Many would be surprised to find that they got it right before a bunch of processing was done to it, thus effectively killing the original sound altogether. Also, run that mix in mono. I would say that about 80% of the music I hear is predominately mono in nature, with the effects spead out over the stereo landscape. Mono mixes that sound good will sound good on ANY system! Remember also that by panning stuff hard, if you were to get out of the stereo field, or to mono the mix, those things would disappear. By the same token, if you panned those things hard, went to mono monitoring and adjusted those things to mix right in mono, when you go back to stereo, those things would be way too loud.
Stereo is nice, and allows for some great separation in the mix, but overdone stereo means that your mixes will sound unbalanced unless the person is listening to them in an idea stereo environment. I am not saying don't use heavy panning for some things, but be carefull.
Also, many people don't balance their stereo fields very well. If you have something consistent panned to say 3 oclock in the field, you need something consistent at 9 oclock to offset it, otherwise your mix will be weighted to one side in at least a certain frequency range. This sounds very strange, and often, you will have to make some tough decisions on how to balance things in a stereo field to create a nice balance. Choose wisely friends!
I also am hearing a lot of weird stuff with people mixing sampled drums with recorded guitars and vocals, but also using like a keyboard bass. Weird indeed. Don't go applying a bunch of eq to those sampled drums, it is only going to make matters worse! Leave them flat. Keyboard bass parts are a waste of time unless you are doing Madonna sounding music. Get a real bass to work with and a nice DI box and decent pre-amp to record the bass with.
The NT1 seems to be the mic of choice for many around here. Fine. Just remember to eq out all that hyped up 4-8KHz crap that mic is famous for when mixing. A moderate bandwidth set at around 6.3KHz with a 3-5db cut will take care of the harsh nasties that many don't seem to hear. But beware! You also just lost a lot of vocal clarity, and you will probably need to possibly apply a tad high shelf eq to give the track some air, something the NT1 doesn't seem to have a lot of anyway.
Also, don't be afraid to use about up to 3db of gain reduction with a compressor when recording a vocal track to digital. Use a longer attack time, like around oh, 15-25ms, and a fairly short release time, like around 40-70ms. I would use a very light 1.5:1 ratio, and no more then a 2:1 ratio. Set the threshold to whatever gives you at the most 3db of gain reduction at the loudest part of the vocal track. If the singer (you?) is hitting more then that on the compressor, or you cannot get your vocals to set at around -10db full digital scale somewhat consistently, then your micing technique for singing sucks, or you need to learn to control your voice better (refer to my comments above about applying too much dynamics to song that don't need them.....).
So many of you are using software to record and mix. Great! Now, WHY DON'T YOU GUYS SPEND SOME TIME AND USE THOSE EDITORS THE WAY THEY WERE MEANT TO BE USED?!?!?!?!
I cannot believe that viewing a recorded kick drum on a DAW doesn't show the one kick in the middle of the song that is like 12db louder then the rest! Then when you mix, you have the whole damn mix at like -20db average because if you turn it all up, that one kick drums peaks out. USE YOUR DAMN EDITOR AND CUT AND PASTE ANOTHER KICK DRUM HIT THERE!!! OR AT LEAST HIGHLIGHT THE HIT AND TURN IT DOWN!!! I AM SHOUTING BECAUSE THIS IS HOW MUCH LOUDER YOUR MIXES WILL GET WHEN YOU DO LITTLE STUFF LIKE THAT!!!
Same with vocal parts that have nasties in it. Sheesh!!! It is only going to take MAYBE an hour to go through that track and highlight all the spots that have nasties and apply some eq cut at the appropriate frequency to tame it so that you don't have to apply that eq cut over the whole track!
I don't have ANY digital editing capabilities when I record and mix multi-track and I get a handle on that stuff using manual tweeks on the eq or volume while mixing if need be. Here and there I can fix something very short lived that sounds messed up on a mix during editing while mastering, and that sure beats the hell out of applying an eq over the whole mix to just tame a errant kick drum or snare hit, or vocal nasty. I can also apply a limiter over the mix now and raise my rms level a lot without slamming the limiter too hard, which in digital creates the most nasty sounding thing you have ever heard. Trust me, I just dealt with some crap like that where I was editing wave forms for stuff that didn't have ANY appearent loudness increase but sure in the hell managed to make the limiter do 10db of gain reduction. This sort of thing should have been dealt with during the mix, or edited before mixing.
Also guys, start using mutes! If on a DAW, spend some time on tracks and silence parts that have no music on them. When you do bring the part back in from silence, apply a Expotential Fade In to it, and not a full on silence to full volume thing. I have heard enough of that, and it sounds terrible because the part seems to rise out of nowhere and slam you in the face. If you are using an 8 buss console, and your song has a break in it where maybe all but one instrument stops, assign all the other instruments to a stereo subgroup and mute the subgroup during the break and unmute right before they come back in. This may take a little practice to time just right, but it will sure make that break sound much better by not have amp noises, tom overring, etc....happening behind it. Also, the hiss from the mixer channel will not be present. Sure, one channel of hiss will not make much difference, but 14 channels creates a lot of noise!
With DAW's, there is just NO excuse for having ANY noise happening on a track when there is no music present on that track. Laziness is the only thing holding many people back from much better resolution in their recordings.
Slow Attack, Fast Release.
Vocals swells develope much slower then something like a snare drum does. With a vocal, you have the initial part of the sound, sort of the percussive part, then a growing part of the sound that is mostly the lower end stuff, and happens to be where most of the energy of the vocal is.
When you are peaking meters on a word, it is not the attack of the voice that is peaking, it is the body of the voice, the stuff that happens after maybe 10ms. While compressing to tape, you don't want to kill the attack of the voice, you only want to tame to swell from killing your meters.
Kick and snare drums. Fast attack, slow release. If it starts sounding too boomy, quicken the release, but always you want a fast attack on the compressor.
Bass guitar. Can go either way depending upon how the track sounds and what they are playing. For a Funk bass line with a lot of percussive stuff going on, you don't want to kill the attack, unless it got recorded with too much attack. See what I mean?
Slow attacks leave more transients in the sound. Fast attacks tame transients.
Fast release tends to thin the sound a bit and make it punchier. Slow release sort of bloats the sound and makes it a little more rounder sounding. A long release time is a way to add a little beef to a track with no eq being applied. Usefull for bass, kick drums, snares, thin acoustic guitars.
Think of attack and release as independent functions because they are. Attack controls transients or leaves them alone. Release controls resonance or leave it alone. The ratio says how much you want the difference to be for anything above the threshold. A big difference means a higher ratio. A small difference means a lower ratio. Threshold is basically a way to control the Knee of what you are doing, or to what degree you are doing that. Once again, sort of an oblique thing here. I have found no good ways to describe the interaction, or "color" of the controls on a compressor because each control is interactive with another, sometimes is dramatic ways.
When recording vocals to analog, you would mostly be crazy to use compression, unless it is a cassette based analog format. Wide fast moving tape will provide mostly all the compression you need for a vocal. If you need a little control for a certain part where the singer either belts it too much, or doesn't give enough, a little cut/boost on the preamp will take care of it.
I make a point to listen everywhere I go.
I like warm sounding rooms. I try to emulate those rooms when setting up a reverb. Most music produced in the last 5 years does not use those phoney sounding plate reverbs. They have a place in music, just not a prominate one.
Room reverbs will serve you well. Chamber reverbs will serve you well. Medium to small rooms and chambers for medium to up tempo stuff. Big rooms and chambers for slower stuff.
Three settings on a reverb that will make all the difference.
Pre-delay - This will allow the original signal develope before the onset of reverb. There may be times where you want to wash out the original sound, and will not have very much pre-delay at all, but not very often.
Hi-Cut Filters. This will mellow out the reverb, and make it sound like real rooms you will be in. Seldom do I have this set above 4KHz. You just don't normally need reverb content above 4KHz.
Diffusion - Lower settings creates reverbs that are more distinct. If you want a more subtle reverb that is not very noticable, raise the value. It should seldom be above 20%. It can go as low as 7% to sound cool.
Another setting that will make the reverb develope in interesting ways is the X Bass setting. It may be labeled Hi Filter too, but basically, it is a multiplier for the low end of a reverb. A 1X value is the algorythym as it was coded. If the values run 1-10 like on most Yamaha reverbs, you are on your own to figure out what they wanted to be the original algorythym. I would normally just say DON'T use Yamaha reverbs at all, because they are some of the most garbage can sounding things I have ever heard in my life. I find that .8X works most of the time. Here and there, you may go up to 2X, but usually only for reverbs assigned to very bright sounding instruments.
Spend a lot of time on developing natural sounding reverbs and save those to use later. You will find yourself using a lot of the same 2 or 3 reverbs in most stuff you mix. Really. Don't settle for factory presets. Get into the unit and play around a lot and find reverbs that sound like rooms you have been in before. These are the most desirable ones to use.
Here and there you are going to create special reverbs that are very intentional effects. Use these sparingly over the course of a whole CD of several songs. Overuse numbs the user to the effect.
Whether you assign a channel Pre or Post EQ to the reverbs really depends on what you want the frequency to be accentuated with reverb. You may have some cut eq on a snare drum at around 2KHz, but you may want that to be where your snare reverb to reside. It would make sense in this case to use a Pre EQ aux send to feed the reverb. If however you are boosting like 400Hz to get the snare to have a little body because it was tracked a little thin sounding, a Post EQ aux send may serve you better because a Pre EQ aux send in this case will not have enough low end content to excite the reverb in the way you want. Think it through, and when in doubt, try both ways.
Also, Pre and Post Fader aux sends can be usefull when assigning a track to a reverb send. Sometimes, I have a whole bunch of snare in the overheads, and I am depending upon that overhead track to supply most of the snare sound. Now, it doesn't make a whole lot of sense to feed the overhead mics to a reverb, so how do you get a snare to excite a reverb? Easy, you assign a Pre Fader aux send to the reverb and just keep the fader down on the snare track, but turn the Pre Fader Aux send up on the snare channel. Cool eh?
People getting timid about trying things because they don't know if it will work!
I was going to save this for an article, but decided after some recent listening to various projects done by home recording engineers that I would "rant" this in this forum.
I get asked the questions "What frequencies should I cut or boost for this or that"? Or, "What should I use compression on"? Or, "How do I get my mix loud"?
While many books could possibly be written covering these subjects, I am going to share some thoughts about mixing right here that I think will serve most well. Believe it or not, this is not going to be that long of a post.
1- From listening to many home recorded songs, I have come to the conclusion that many are monitoring at levels far too low to mix effectively. There may be issues involved with turning it up a bit, so it is not that I am without sympathy for those reasons, but mixing at low volumes means you will have the tendency to mix too much low end into your mix. Equalizing the monitoring system is not going to solve this problem. Turn those puppies up!
I would recommend that you record a 1KHz test tone on your system, and play back that test tone through the same D/A converters you will use to monitor your mixes. With the test tone recorded at -6db digital full scale, use a dB meter and turn up your monitors until that test tone is at about 85db with a C weight on the meter. I am not kidding.
It is a good idea to do one speaker at a time so that you can see if possibly you have a bad side of your amp, a speaker that is different then the other (it can happen), or bad cabling. If there is a difference between the two sides of the amp, suspect your cabling first, either the amp or D/A converters second and third, then your speakers.
2- Your near field monitors are NOT going to have as much low end, and will have a tighter low end then home stereo speakers will. You HAVE to get used to this difference in the sound. A great way to KNOW the low end of your monitors is to listen to a lot of professionally recorded mixes using the same D/A converters and monitors. There is no other way that allows you to be as subjective. If you are recording on a DAW, find a way to digitally transfer professional mixes to your DAW so they are available to play back.
3- Blending instruments is nowhere as hard as you might think! It only requires that you pay attention to what you are doing, and sadly, I don't hear many people doing that.
Let's take for example that the hi hat and the vocal are going to have some similar places in the mix. This is not uncommon in Pop music. The sibilance of the vocal is going to be about the same as the hi hat frequency. Make sure that you don't have one or the other dominating in the mix!
A kick drum and a bass guitar pluck should have a very similar sound too. Both instruments require that you rid them of low midrange frequencies, like around 200Hz. Solo both and check to make sure that they are not killing each other. While a kick drum can get a little click sound for definition with a eq boost around 3 or 4KHz, the bass guitar will get articulation with a boost around 500-1000Hz. But both are going to have a similar 100Hz body to them, and pay very close attention to that. Also, don't get over-kill on the stuff below 100Hz. With most monitors, it is hard to hear anything below that, and you will have the tendency to boost 80Hz a lot to try to make these two instruments sound fuller. Be aware of that because a 3db eq boost on a kick drum at 80Hz is going to eat up a lot of your potential dynamic range in the mix, and it may not sound as good as you think once you start playing the song on home stereo systems
In the midrange, watch for the vocals to have harsh words that jump out at you. One of the problems many face is that they are performers too, and their midrange hearing starts to get shot, and they don't quite hear those 2-4KHz jumps in the vocal that kill your ears through other converters and play back systems. Also, the guitar is going to have a lot of it's top end in this range, and you need some room for that.
With guitars, watch those boost that are below around 250Hz. Anything below that is not worth boosting on a guitar. You once again are just eating up potential dynamic range in the mix with a boost that WILL not sound good on home stereo speakers. If your guitar sounds too bright, try cutting stuff above 5or 6KHz. I roll out a lot of 10KHz high cut filter on guitar, unless they sound very dull, in which case, you need some boost there to give them "air". At no time is a boost below about 250Hz good for your mix with guitars. Don't do it and be happy in the end!
When you are done with your mix, take a good close look at it in a .wav editor. Do you have a lot of spikes in the mixes that are 6db hotter then the rest of the song? Is your average level well below 50% of the available dynamic range? If either is the case, I assure you that your mix is not good by any means. Those 6db jumps mean that you needed some limiting on a certain instrument, or that you have a eq boost on something and it is causing the track to be super sensitive to that frequency. Say with a guitar, those big low end boosts means that when ever the instrument hits a note with that fundamental frequency you get that much more boost at that frequency. Not a good thing. Eq boosts are only really good for deficiencies in the track. I revert back to if you are trying to use eq to make something sound drastically different, then your best bet is to record the part again and get that type of sound recorded. This way you KNOW what you are getting in the sound.
It really comes down to paying attention to instruments that share certain frequency ranges. The click on a kick drum shares the same space as the upper end of guitars, and the bulk of the energy in vocals, but the kick drum is not playing continuously, so it is okay to boost that up a bit to bring it out. But boosting an guitar at that frequency, or even a vocal means a mismatch with the other instruments in that range. We are looking for balance. If you are finding that something is dominating in the mix in a certain frequency range, check to make sure that you don't have boost eq in that range that is contributing to this. If you don't have boost eq on it, then you may need to apply a little CUT eq, but just enough to mellow it out, which is usually around a 3-5db cut. Any more then that and you have a big problem with the way the track was recorded. But another consideration too is how important any instrument is in a certain frequency range. A vocal masking a guitar a little bit is okay because usually, the vocal, drums, and bass are important during vocal passages.
If you are using digital eq's, any more then a 6db cut or boost is going to sound awful! The algorithms are just not good enough to provide a smooth sound and color to a track with excessive eq, so large cuts or boosts with digital eq's should be avoided at all costs. With cheaper analog eq's, you have a similar problem. Are you catching on here? Get you tracks close to the way you want them to sound while tracking and avoid heavy eq.
Before mixing a song, it may be a good idea to run a mix where you use no effects or eq and just adjust volumes to get the best mix you can. Burn that to CD and start listening on many different systems. You may be surprised at how good it actually sounds on home stereo systems, and how much more open and unprocessed your mix is. You will certainly still have a few things you would like to fix, but now you KNOW what needs to be fixed, and are not being biased by the better sounding D/A converters and monitors your setup has.
A great way to think about blending instruments is considering the importance they play in the mix. Obviously, a vocal 99% of the time will be very important. But that doesn't mean that the vocal should over power everything else. I call this "oblique" mixing, where there is an underlying theme that stays consistent, and you have parts that build over that but don't interfere too much. If fact, if you consider what oblique means, mixing a song effectively is very similar in the dynamics. You have a very steady bottom with a little color on top. The color grows and shriks, but you still have a solid foundation that is not really effected by it. This means you really have to consider what each instrument REALLY contributes to the song. Also, "moods" of certain parts should be reflected in the dynamics too. A little guitar ditty during a verse doesn't really need to be blasting away does it? But during a build up at the end of a chorus, you may have a lead that comes in for the last measure to embellish the transition. You need to leave some sonic space for this add in, and leave enough headroom for the added volume.
In Pop tunes, the kick and snare drums are very important, and should be very articulate in the mix, along with the vocals. The bass guitar is usually the harmonic foundation, and should occupy the low end of the mix predominately. Guitars tend to sound sort of quiet in the studio, and at lower volumes. Be wary of that. But, solos should occupy the same dynamic and sonic place the vocals did in the mix.
Strive for when a instruments stops playing that your mix doesn't drop in volume much at all, if at all. Viewing RMS metering of some sort will help a lot with this. If you know how to view peak metering for the RMS value (a skill that takes awhile to learn) even better because you can also see just how often you have used up your dynamic range.
One thing I have noticed a lot is people trying to put extreme dynamics in a song that doesn't need extreme dynamics. Drummers are by far the worst about this in recording, followed closely by singers. Consistent volume levels are desired in recording. How the parts fit together, and to a certain degree, the actual notes being played will create natural sounding dynamics in Pop tunes. Also, an instrument stopping in the middle of the song for whatever reason creates a change in "perceived" dynamics, even though your metering may not show anything different.
Another thing I have been hearing a lot is either sparse or heavy reverbs. Don?t be afraid to use a nice reverb across several rhythm instruments. It is the glue that holds them all together. At the same time, big huge long reverbs all over the vocal line don't work for anything other then Whitney Houston tunes, or the Scorpions? What I have noticed about the big reverbs many people use is that they leave in a lot of high-end content in the reverb. Hey guys, try cutting everything above 3 or 4KHz on that reverb. Also, I have seldom EVER used a reverb recently where I didn't have at least 40ms of pre-delay on it. A very short decay, with a hi-cut around 3 or 4KHz and a pre-delay that is appropriate will really fatten up a track without having to resort to eq! Play with it and you will never use those huge 3 sec plates with all that 10KHz content again!!!
Use your ears AND your eyes to gauge your mix. Your ears will tell you when thing sound screwy, but viewing your mix in a .wav editor will certainly "show" you some potential problems. View a professionally recorded song in a .wav editor and you will usually see a very well developed .wav, with very consistent peaks on the kick and snare drums, and you won't see the volume jumps when the vocal comes in, and drop when it goes away. Creatively using eq can help tame dynamics a bit, and gentle compression on instruments will do the same.
Anway. Just a rant so to speak about mixing. If it helps anyone, great! If not, at least you have some more material to curse my name by?
Ed
Back to running a mix without effects and eq.
You need to know WHAT in the mix needs to be fixed before you do anything. I have seen people start eqing stuff that sounded great because that is what they normally do to that instrument. My oh my..... The idea in recording is to get as close to the sound you want while tracking. Many would be surprised to find that they got it right before a bunch of processing was done to it, thus effectively killing the original sound altogether. Also, run that mix in mono. I would say that about 80% of the music I hear is predominately mono in nature, with the effects spead out over the stereo landscape. Mono mixes that sound good will sound good on ANY system! Remember also that by panning stuff hard, if you were to get out of the stereo field, or to mono the mix, those things would disappear. By the same token, if you panned those things hard, went to mono monitoring and adjusted those things to mix right in mono, when you go back to stereo, those things would be way too loud.
Stereo is nice, and allows for some great separation in the mix, but overdone stereo means that your mixes will sound unbalanced unless the person is listening to them in an idea stereo environment. I am not saying don't use heavy panning for some things, but be carefull.
Also, many people don't balance their stereo fields very well. If you have something consistent panned to say 3 oclock in the field, you need something consistent at 9 oclock to offset it, otherwise your mix will be weighted to one side in at least a certain frequency range. This sounds very strange, and often, you will have to make some tough decisions on how to balance things in a stereo field to create a nice balance. Choose wisely friends!
I also am hearing a lot of weird stuff with people mixing sampled drums with recorded guitars and vocals, but also using like a keyboard bass. Weird indeed. Don't go applying a bunch of eq to those sampled drums, it is only going to make matters worse! Leave them flat. Keyboard bass parts are a waste of time unless you are doing Madonna sounding music. Get a real bass to work with and a nice DI box and decent pre-amp to record the bass with.
The NT1 seems to be the mic of choice for many around here. Fine. Just remember to eq out all that hyped up 4-8KHz crap that mic is famous for when mixing. A moderate bandwidth set at around 6.3KHz with a 3-5db cut will take care of the harsh nasties that many don't seem to hear. But beware! You also just lost a lot of vocal clarity, and you will probably need to possibly apply a tad high shelf eq to give the track some air, something the NT1 doesn't seem to have a lot of anyway.
Also, don't be afraid to use about up to 3db of gain reduction with a compressor when recording a vocal track to digital. Use a longer attack time, like around oh, 15-25ms, and a fairly short release time, like around 40-70ms. I would use a very light 1.5:1 ratio, and no more then a 2:1 ratio. Set the threshold to whatever gives you at the most 3db of gain reduction at the loudest part of the vocal track. If the singer (you?) is hitting more then that on the compressor, or you cannot get your vocals to set at around -10db full digital scale somewhat consistently, then your micing technique for singing sucks, or you need to learn to control your voice better (refer to my comments above about applying too much dynamics to song that don't need them.....).
So many of you are using software to record and mix. Great! Now, WHY DON'T YOU GUYS SPEND SOME TIME AND USE THOSE EDITORS THE WAY THEY WERE MEANT TO BE USED?!?!?!?!
I cannot believe that viewing a recorded kick drum on a DAW doesn't show the one kick in the middle of the song that is like 12db louder then the rest! Then when you mix, you have the whole damn mix at like -20db average because if you turn it all up, that one kick drums peaks out. USE YOUR DAMN EDITOR AND CUT AND PASTE ANOTHER KICK DRUM HIT THERE!!! OR AT LEAST HIGHLIGHT THE HIT AND TURN IT DOWN!!! I AM SHOUTING BECAUSE THIS IS HOW MUCH LOUDER YOUR MIXES WILL GET WHEN YOU DO LITTLE STUFF LIKE THAT!!!
Same with vocal parts that have nasties in it. Sheesh!!! It is only going to take MAYBE an hour to go through that track and highlight all the spots that have nasties and apply some eq cut at the appropriate frequency to tame it so that you don't have to apply that eq cut over the whole track!
I don't have ANY digital editing capabilities when I record and mix multi-track and I get a handle on that stuff using manual tweeks on the eq or volume while mixing if need be. Here and there I can fix something very short lived that sounds messed up on a mix during editing while mastering, and that sure beats the hell out of applying an eq over the whole mix to just tame a errant kick drum or snare hit, or vocal nasty. I can also apply a limiter over the mix now and raise my rms level a lot without slamming the limiter too hard, which in digital creates the most nasty sounding thing you have ever heard. Trust me, I just dealt with some crap like that where I was editing wave forms for stuff that didn't have ANY appearent loudness increase but sure in the hell managed to make the limiter do 10db of gain reduction. This sort of thing should have been dealt with during the mix, or edited before mixing.
Also guys, start using mutes! If on a DAW, spend some time on tracks and silence parts that have no music on them. When you do bring the part back in from silence, apply a Expotential Fade In to it, and not a full on silence to full volume thing. I have heard enough of that, and it sounds terrible because the part seems to rise out of nowhere and slam you in the face. If you are using an 8 buss console, and your song has a break in it where maybe all but one instrument stops, assign all the other instruments to a stereo subgroup and mute the subgroup during the break and unmute right before they come back in. This may take a little practice to time just right, but it will sure make that break sound much better by not have amp noises, tom overring, etc....happening behind it. Also, the hiss from the mixer channel will not be present. Sure, one channel of hiss will not make much difference, but 14 channels creates a lot of noise!
With DAW's, there is just NO excuse for having ANY noise happening on a track when there is no music present on that track. Laziness is the only thing holding many people back from much better resolution in their recordings.
Slow Attack, Fast Release.
Vocals swells develope much slower then something like a snare drum does. With a vocal, you have the initial part of the sound, sort of the percussive part, then a growing part of the sound that is mostly the lower end stuff, and happens to be where most of the energy of the vocal is.
When you are peaking meters on a word, it is not the attack of the voice that is peaking, it is the body of the voice, the stuff that happens after maybe 10ms. While compressing to tape, you don't want to kill the attack of the voice, you only want to tame to swell from killing your meters.
Kick and snare drums. Fast attack, slow release. If it starts sounding too boomy, quicken the release, but always you want a fast attack on the compressor.
Bass guitar. Can go either way depending upon how the track sounds and what they are playing. For a Funk bass line with a lot of percussive stuff going on, you don't want to kill the attack, unless it got recorded with too much attack. See what I mean?
Slow attacks leave more transients in the sound. Fast attacks tame transients.
Fast release tends to thin the sound a bit and make it punchier. Slow release sort of bloats the sound and makes it a little more rounder sounding. A long release time is a way to add a little beef to a track with no eq being applied. Usefull for bass, kick drums, snares, thin acoustic guitars.
Think of attack and release as independent functions because they are. Attack controls transients or leaves them alone. Release controls resonance or leave it alone. The ratio says how much you want the difference to be for anything above the threshold. A big difference means a higher ratio. A small difference means a lower ratio. Threshold is basically a way to control the Knee of what you are doing, or to what degree you are doing that. Once again, sort of an oblique thing here. I have found no good ways to describe the interaction, or "color" of the controls on a compressor because each control is interactive with another, sometimes is dramatic ways.
When recording vocals to analog, you would mostly be crazy to use compression, unless it is a cassette based analog format. Wide fast moving tape will provide mostly all the compression you need for a vocal. If you need a little control for a certain part where the singer either belts it too much, or doesn't give enough, a little cut/boost on the preamp will take care of it.
I make a point to listen everywhere I go.
I like warm sounding rooms. I try to emulate those rooms when setting up a reverb. Most music produced in the last 5 years does not use those phoney sounding plate reverbs. They have a place in music, just not a prominate one.
Room reverbs will serve you well. Chamber reverbs will serve you well. Medium to small rooms and chambers for medium to up tempo stuff. Big rooms and chambers for slower stuff.
Three settings on a reverb that will make all the difference.
Pre-delay - This will allow the original signal develope before the onset of reverb. There may be times where you want to wash out the original sound, and will not have very much pre-delay at all, but not very often.
Hi-Cut Filters. This will mellow out the reverb, and make it sound like real rooms you will be in. Seldom do I have this set above 4KHz. You just don't normally need reverb content above 4KHz.
Diffusion - Lower settings creates reverbs that are more distinct. If you want a more subtle reverb that is not very noticable, raise the value. It should seldom be above 20%. It can go as low as 7% to sound cool.
Another setting that will make the reverb develope in interesting ways is the X Bass setting. It may be labeled Hi Filter too, but basically, it is a multiplier for the low end of a reverb. A 1X value is the algorythym as it was coded. If the values run 1-10 like on most Yamaha reverbs, you are on your own to figure out what they wanted to be the original algorythym. I would normally just say DON'T use Yamaha reverbs at all, because they are some of the most garbage can sounding things I have ever heard in my life. I find that .8X works most of the time. Here and there, you may go up to 2X, but usually only for reverbs assigned to very bright sounding instruments.
Spend a lot of time on developing natural sounding reverbs and save those to use later. You will find yourself using a lot of the same 2 or 3 reverbs in most stuff you mix. Really. Don't settle for factory presets. Get into the unit and play around a lot and find reverbs that sound like rooms you have been in before. These are the most desirable ones to use.
Here and there you are going to create special reverbs that are very intentional effects. Use these sparingly over the course of a whole CD of several songs. Overuse numbs the user to the effect.
Whether you assign a channel Pre or Post EQ to the reverbs really depends on what you want the frequency to be accentuated with reverb. You may have some cut eq on a snare drum at around 2KHz, but you may want that to be where your snare reverb to reside. It would make sense in this case to use a Pre EQ aux send to feed the reverb. If however you are boosting like 400Hz to get the snare to have a little body because it was tracked a little thin sounding, a Post EQ aux send may serve you better because a Pre EQ aux send in this case will not have enough low end content to excite the reverb in the way you want. Think it through, and when in doubt, try both ways.
Also, Pre and Post Fader aux sends can be usefull when assigning a track to a reverb send. Sometimes, I have a whole bunch of snare in the overheads, and I am depending upon that overhead track to supply most of the snare sound. Now, it doesn't make a whole lot of sense to feed the overhead mics to a reverb, so how do you get a snare to excite a reverb? Easy, you assign a Pre Fader aux send to the reverb and just keep the fader down on the snare track, but turn the Pre Fader Aux send up on the snare channel. Cool eh?
People getting timid about trying things because they don't know if it will work!