Ok! So you have all of your tracking done… and you simply don’t know where to begin when it comes to mixing and blending all of this stuff together! I admit, it can be a very overwhelming task when you observe it as a whole. Just dealing with all of these sounds and trying to come up with an end result that sounds listenable seems difficult indeed. The key to successful mixing, though, is much like the key to completing any complex task; it must be done in steps, and with great care.

So where do you start? That’s a fair question indeed if you’re to begin the whole process. Where I begin, though, is making sure everything is in order. Basically do a quick scan through your mix, see that tracks appear to be in stable condition, and nothing is out of place. This shouldn’t take too long, as you’re basically looking for anything that seems odd in your workstation.

To begin your mix, you must understand some key points first. Your mix is somewhat like a three dimensional space in which you place the building blocks of your track. Each individual piece stacks on the other to create a final, whole product. And just as in architectural design, your “building blocks” (the sounds of your mix, if you hadn’t caught on already, heh) can only occupy their own space. This is where we delve into the concepts of frequency management and dynamic control.

A common problem amateurs run across when completing a mix is that it seems too “muddy” and that it is lacking clarity. The reason for this is that the different tracks in your mix are more or less tied to a certain frequency range, and the more they overlap, the more confusion and mess you will get. To work around this, we will use the all mighty, wonderful process known as “equalization”. This is, by far, the greatest and most important tool in any audio engineer’s arsenal, and having a very extensive fundamental knowledge of its procedures will make you that much more of a competent mixer. Each element of your mix, each instrument, will take up its own space within your three dimensional field. Knowing which frequencies are abrasive and which ones conflict with each other is important in achieving clarity.

To begin your mix, I’ll talk about a method known as “zero point reference mixing”. The process involves starting with the dominant element of any given song or soundtrack and setting its average level to read around 0 VU. When working with all digital systems that use peak metering, the engineer uses whatever standard that studio is set at. Most digital systems that work in a 24-bit or greater mix environment use the formula 0 VU=(-14). Simply put, this means that in a song, say the vocal would be EQ’d and compressed to read around 0 VU plus or minus (preferably minus) 3 db. Once this is achieved, blend whatever effects might be necessary for the particular song you’re working on. When the dominant element sits at the proper volume, determine the next musical (non rhythmic) instrument. This could be the guitars, piano, synth patch, anything with chords. The way I would determine what the second element in the mix should be is to ask “if this song were an unplugged version, what would accompany the lead element?” When you have blended the second element into the fold, you now have established a proper reference level you can build upon. Try adding bass, set what the EQ should be taking into account the other items in the mix. Take a look at what the compression should be, if any. Once you have the bass in, you might try seeing how the kick drum works with the bass. Try and understand the relationship between all the elements you have up and make appropriate adjustments to achieve a pleasing balance between the instruments.

Now that we have established a very good way to have excellent volume structure in your mix, we can work on equalization and compression. I’ll start with explaining compression first.

Essentially when you use a compressor on a piece of audio, you are controlling the dynamic range. The dynamic range is the relationship between the softest and loudest points in a sound. A compressor “squashes” this range, to make the softer parts seem more equal to the louder parts, depending on the settings used. This can be a very useful tool, but it can also destroy the mix of its life. Most of the time, it is applied to vocals, and other instruments that have a naturally wide dynamic range. It can also be used on percussive instruments to add a lot of punch, as well. The attack time of a compressor affects how quickly it takes for the unit to begin “compressing”. The ratio is how much compression is being done, the release time is how long the compressor stays working once activated, and the threshold is the lowest point at which the compressor will begin to work. Most compression units also have an output option that will allow you to raise the makeup gain level of the amount being reduced.

Equalization is what you want to do next. This is where the true mettle of an engineer is tested. A basic equalizer will give you these functions: the frequency range, which is fairly self explanatory, your boost and cut, which can either raise or lower the volume of the selected frequencies, and the Bandwidth, or Q, which is how wide of a curve the frequency range is. Now I (and most others) will tell you that subtractive EQ is the safest way to go. That is, finding which frequencies conflict, and which ones sound bad, and cutting. My personal method is to first solo a track, turn the boost all the way up, and slowly turn the frequency knob until you find the sound that sticks out in the most annoying way, and then cut to about -3 db. After that, I would un-solo the track and see how it sounds with the rest of the mix. I do this for just about all of my tracks.

Another form of EQ is called the filter, which essentially removes all frequencies above or below a given point. There are highpass filters and lowpass filters. A highpass will cut all frequencies BELOW a given threshold, and the lowpass will cut all frequencies ABOVE a given threshold. That being said, I tend to highpass nearly every track (except for the bass guitar, kick drum, and maybe a floor tom) at around 100 hz to get rid of all bass content that would otherwise clutter the mix.

Here is something to look at to take into consideration when you apply EQ:

20 Hz and below - impossible to detect, remove as it only adds unnecessary energy to the total sound, thereby most probably holding down the overall volume of the track
60 Hz and below - sub bass (feel only)
80(-100) Hz - feel AND hear bass
100-120 Hz - the "club sound system punch" resides here
200 Hz and below - bottom
250 Hz - notch filter here can add thump to a kick drum
150-400 Hz - boxiness
200 Hz-1.5 KHz - punch, fatness, impact
800 Hz-4 KHz - edge, clarity, harshness, defines timbre
4500 Hz - extremely tiring to the ears, add a slight notch here
5-7 KHz - de-essing is done here
4-9 KHz - brightness, presence, definition, sibilance, high frequency distortion
6-15 KHz - air and presence
9-15 KHz - adding will give sparkle, shimmer, bring out details - cutting will smooth out harshness and darken the mix

This can only help you so much, so I recommend playing around with the equalizer as much as possible. Remember, the key is to take away the frequencies that conflict and sound bad.

Now, after you have taken care of the basic foundation of your mix, you probably want to add some effects, maybe do a bit of automation and put some finishing touches on it. Effects can be very cool if they are used properly, or they can totally mess up your sound when abused! Reverb is often used for ambience and to give a sense of dimension with your tracks. I recommend using a small plate reverb for the vocals, and then a very minor touch of verb on the drums. If you want to add even more on the snare, you can, but you should be very careful of over doing it. You can completely ruin a track with too much reverb.

I won’t go in depth with the other effects for now, as they are mostly for flavor and should be experimented with on your own, but I’ll outline some tips to keep in mind when you do your mixing.

*Be conscious of your panning. For double tracked guitars, a hard left/right pan isn’t always necessary. Kick, bass, vocals, and any leads should be centered. I like to put the overhead drums on 75% left and right, and then pan the toms of the drum kit according to how they are set up in relation to the drummer.

*Make sure you control the dynamics of your vocals, because they can sound quite weak in certain songs if not properly compressed. It isn’t always necessary to use heavy compression though. Use good judgment.

*Your master fader should never clip. If it is be sure to check the gain of each track and adjust the volumes to remove the clipping. Likewise, you don’t want individual tracks to clip either.

*Be conscious of signal flow, just as in recording (except much more important here). If you’re using a complex console set up with a large patchbay and a ton of outboard effects, things can get confusing. Take things slowly one step at a time, trace where your signal is going from the source, follow it to everything else.

*Use automation to your advantage. If there are passages in your song when the guitars should drop out a bit and lead the piano do its thing, you should take that into consideration and automate the levels. When it comes to panning and other effects, especially delays, this is more of a production call.


Ok, this is call I can think of for now. If I feel that I’ve missed anything or I see an error, I’ll make the changes! Thanks!

ALSO!!! Here is a rather large "audio glossary" that I think everyone should check out. If you have any question of what a certain term that you've heard means, this will almost definitely have it.

http://www.modrec.com/glossary/

Thank you for the article, it is accurate and useful. I learned some new things from it. In particular : where to start the mix and some moment about compression.

Perhaps some sort of mention here should be added regarding drum mixing? Here's some stuff, anyway:

I typically find 80hz is a good place to add EQ for a kick drum. Keeps it bassy and still relatively punchy- not good for epic double kick, but good for slower tunes. Sometimes the kick sounds great when it's EQ'd at the same frequency with the bass guitar- over times it's best to have some separation (maybe eq the kick a bit higher). The middle ground isn't really worth considering, imo. To bring out the click in the bass drums, 6khz is good imo.

Effectively, for a drum kit- Compression is a must. The snare in particular has a very high peak volume, and compressing it will keep your levels from going too high. Hard compression on the snare, the kick and the cymbals will keep the drums in line.

A noise gate is also typically a good idea- essential when recording and mixing a drum kit imo. Effectively, a noise gate cuts all sound that goes underneath it's threshold. Make sure that the attenuation of noise under the threshold is at full, and then move the threshold around until the only thing you hear are the desired element. For an example, take a snare track. The main sound you'll get in the track will be the snare, but you'll also hear the kick drum, the cymbals and whatnot. Turn the threshold up until all you hear is the snare's attack, and then make the release longer until you get the full sound of the snare. This will ultimately make the snare sound much punchier, and the drums in general will sound a lot cleaner and they'll have a lot more clarity.

Finally, remember to do some panning with the drums, pretty obvious really. Of course, whatever panning you do depends on your microphone situation.

..Hope that's useful adivce?

Yes, very excellent additions! I'm glad it helped out a bit, deathfog.

Thanks alot for this post, very helpful.

well written. Those EQ references are a major plus.

Ok, something else that is VERY important that I neglected. The Fletcher-Munson equal-loudness contour curves. These indicate the ear's average sensitivity to different frequencies at various levels. The curves tell us the sound pressure levels that are required for our ears to hear frequencies along the the curve as being equal in level to a 1000-Hz reference level (what's known as a "phon"). Thus, to equal the loudness of a 1kHz tone at 110 db SPL (a level typically created by a trumpet type car horn at about 3 feet away), a 40 Hx tine has to be about 6 dB louder, while a 10kHz tone must be 4 dB greater in order to be perceived as being equally loud. At 50 dB SPL (the noise level present in an average business office), the level of a 40 Hz tone must be 30 dB greater and a 10 Hz tone must be 13 dB greater than a 1 kHz tone to be perceived as having the same volume. Thus, if a piece of music is mixed to sound great at a level of 85 to 95 dB, its bass and treble balance will actually be boosted when turned up (often a good thing). If the same piece were mixed at 110 dB SPL, it would sound both bass and treble shy when played at lower levels.... as no compensation for the ear's response was added to the mix. Over the years, it has generally been found that changes in apparent frequency balance are less apparent when monitoring at levels of 85 dB SPL.

Here is the Fletcher-Munson curve:

Really helpful information, thanks.

Do you have any particular advice about maximising and compression of the whole mix? I find my stuff doesn't so much sound 'muddy' and overpowered as rather sparse and quiet compared to the average level of say, a song on a professional CD or mp3.

Also where you've ever use EZdrummer do you tend to treat it like you would a normal kit?

Oh master bus compression. Well my advice concerning that is to use a "less is more" approach. An "overall" compressor setting for the master bus is typically one that uses a slow attack time, low ratio and a low threshold. You don't want more than 3 dB of gain reduction, or else you will end up squashing much of your dynamic resulting in a lifeless mix.

As far as programmed drums are concerned, yes treat it like a normal kit.

Aye, but do you advise using maximisers to push the whole volume (as high as it can go without clipping)?

No, I don't do any brick wall limiting, that usually turns out to be a disaster which results in a mix with no dynamics.

Before you do any master bus compression, you should make sure your mix sounds perfect as it is, get your volume blend adjusted properly, make sure to apply only the needed amount of compression on each individual track, nothing more. After that's all done if you still want to add some power to your mix, if you don't have it already, go for the overall compression method using settings similar to what I described.

If you want I can post examples of tracks with heavy brick wall limiting, and then one without, only using minimal compression, and you can be the judge.

I take it that a wee bit of make-up gain on the master bus is the way, then...

You should always have maximum gain before clipping on the master bus.

However, what I like to do is set the master fader at -6 dB when working on digital systems to give myself some headroom to work with in my mix, and after everything is complete, I can turn it back up. If all of your tracks are managed properly with excellent dynamic and frequency control, you shouldn't have a problem coming up with a very vivd and hot mix. A bit of overall compression at the end of the chain can also spice things up, but when overdone it sounds terrible.

Regarding master bus compression- It's typically best not to lay it on too thick in the mixing stage. Typically most of the stereo bus compression gets done in the mastering stage, and a mastering engineer won't be able to do all that much with a mix that's already got heaps of compression.

edit: haha, most noob post ever.

Thorgrim is the man. You pretty much completely enlightened me as to the recording/mixing process (I was pretty clueless before.)

Thanks alot man.

Could you (Thorgrim) or someone else make a tutorial using Audacity and such?

Haha, awesome. More info on shit I know nothing about but would be asking about when I got my MBox in a few days... good timing.

Edit: super noob post.

All of this is just fundamental guidelines on how to mix your audio... the DAW host should not really matter, as long as it has the same functions. You can most definitely control channel faders in Audacity, as well as use EQ and compressors, and various other effects.

I'm aware, I'm just oblivious how to use some of the effects and editing and the like, but the original tutorial is pretty informative.

Also one piece of advice. Bussing.

You can create bus tracks to send multiple tracks to, this is good for when you want to add the same reverb (or anything that adds to the sound) to a number of tracks. It saves more memory than putting inserts into every track. Where you're doing things that change the sound (like eq) it's best to do stuff on the track itself.

I'd definitely find this helpful.

Could you extrapolate the idea of compression a little bit?...for the song I'm working on now, I did some clean vocals, first time trying it, and I'm having some threshold issues...

any additional advice you could give on compressing/limiting would be appreciated.

Well you want to keep an eye on your gain reduction meter. Typically I'd go for 4-8 dB of gain reduction, depending on what suits your song.

Typically I'd set a very fast attack time, maybe 1-5 ms and a long release time so you are having good compression going on all the time. Make sure to adjust your make up gain accordingly.

When you set your threshold, look at your gain reduction meter as I said, and just keep lowering it until the desired amount.

Hello there! I am new here and I have no idea of mixing and such stuff. Can you suggest me a mixing guide for amateurs or something?

What do you think this thread is?

Yeah but it mentions stuff I don't even know what they are. SOmething for beginners.

Post questions you have in here and they can be answered.

That is a beginners guide. If there are any terms you don't know, go to the link provided at the bottom of the first post and look them up:

http://www.modrec.com/glossary/

Ok, thanks a lot!

I hate the damn auto-locking feature here... Opened again, for the second time.

Dunno if this is the best place for this, but it seemed like the most appropriate -

Just wanted to suggest to everyone that before you decide on a final mixdown make sure you have listened to it on as many types of stereos/speakers as possible, it's a really easy yet important step that can prevent big problems.

That's a bit excessive. What I recommend instead is to have a reference CD during each mix. Bring an album that you are incredibly intimate with, that you know like the back of your hand. You can test it out on whatever monitors you use and hear all of the nuances that would otherwise affect your perception of the mix. It's much more efficient.

I think you misunderstand what I mean. I just mean that when you think you have a final mixdown try and listen to it on as many things as possible before you press your cd or demo or whatever because things can sound drastically different than whatever you mixed it on, even if it's supposed to be 'true' sound or whatever.

I know from experience when I thought something was mixed down and listened to it on a different stereo one day and the bass was overbearing and not due to any setting on the stereo. Turned out the bass needed further limiting which you couldn't tell at all on other stereos.

I absolutely did not misunderstood what you meant. In fact I know exactly what you're talking about which is why I suggested my alternative, which is much better. If you have a reference CD which you know what sounds like on a range of speaker systems, the need to test your own mix on those systems is nullified. Trust me, I've been doing this for some time now.

Ok maybe your way is better. I was just making a suggestion for others that would have helped me. Maybe explain how your way is better or is supposed to work because I don't think I fully understand. I mean let's say I were to have brought a reference CD for instance Slayer's "Show No Mercy" which I have been listening to religiously for close to 12 years. I don't see how listening to that on the system I was mixing on would have helped me with my recording. It still would have had the problem with the bass on certain systems which I wouldn't have found just judging by Show No Mercy.

You can notice discrepancies between the mixing monitors and whichever other systems you are used to playing your reference album. If your mixing monitors sound a bit bass heavy compared to the sound that you would hear in your car or home system, you can accomodate accordingly. This is a standard AE technique, really. It's used all over the world in studios.

That's a good method to achieve a final mix, but I wouldn't have caught the problem with the bass had I done that alone. The monitors were not bass heavy. When I played the recording on one of my home stereos that I don't always use it sounded ridiculously bass heavy. Furthermore the reference cd I used as an example I know for a fact sounds perfectly normal on this system, not bass heavy at all. So it was something in the recording that wouldn't have been resolved had I just used your method alone.

I mean really the things we are talking about are two different things, your talking more about a method to achieve a good final mix and I'm talking about a last check of that final mix. It can't hurt to be sure about these things, it's just a suggestion anyways I'm not forcing anyone to do this.

Well if you have balanced monitors it could have easily been observed with just the reference CD. I've done it many times this way.

Here are two write ups I did for EZDrummer and Reaper:

http://metal-archives.com/board/viewtopic.php?p=1032153

http://metal-archives.com/board/viewtopic.php?p=1045556