Hey MixMaster, What's The Matter With You?

February 22, 2007 7:10 AM

How can I learn to mix (not DJing, Audio Engineering) my own music in my limited spare time? I've bought books and read what I can on the net, but I find that I'm more of a "hands-on" learner type. Lately I feel that this is what's holding me back with my music-making. I'm in NYC.

Music has been my first love/passion since I can remember. A lot of times it's what keeps me going and at some point I'd like to master the art for my own sake, as much as that's possible (not taking stuido mastering into account), and be able to make my own consumer-ready releases. Relative to that, the past couple of years work has become a major time suck in my life which leaves a lot less time for what's become largely a hobby for me. I'm ok with the "hobby" aspect, but I haven't finished a song in that time (tons of loops/ideas though) and I can't help but feeling that it's largely due to "so-so" mixes. While I subscribe to the theory that a good song is a good song poor mix or no and should reflect that even when composed on, say, a guitar, but I also think it's hard to stay motivated when I know it's only going to end up as a "so-so" mix. I know that mixing comes down to frequencies and cutting them (or boosting as needed) and practical application of compression, EQ, etc, (the theory I mostly get I think), it's the "in practice" part that I get stuck/frustrated with given that simply writing music, beats, etc is time consuming enough on it's own. I have fairly "pro" monitors (DynAudio) and do most of my composing/mixing in a computer sequencer. If I do say so myself, I also have a pretty good ear, it just seems that I get to a point where I'm cutting a frequency " . .is that right? . . is that right? . . .damnit was that . . ." and then I get to where I can't tell if it's better or worse than when I started.

I know there's places like Full Sail and such, but some cursory research suggests those places are largely AE diploma mills and seem a bit overinflated in terms of price and just seem overkill in terms of what I'm after. I'd suck it up and volunteer to be coffee boy or something, but frankly don't have the time for that. I saw that the New School offers classes for this, but so far that doesn't seem to start until Sept. which does my impatience no good. So, what's the best route to go in NYC for someone wanting to learn Audio Engineering in their limited spare time?

posted by teemo to media & arts (24 comments total)

Start reading tape op, a subscription is free, plus they have forums filled with professionals.
posted by drezdn at 7:19 AM on February 22, 2007

Learn to mix by mixing other people's music. Mixing your own is very subjective and if you're looking to develop your skills you'll be better served practicing on material you can approach with more of a neutral ear.

I also second tape op as well as the prosoundweb forums. If you hang around in those places there are often mix competitions you can join - a raw track is uploaded and everyone takes a crack at a mix. Not only do you get to practice but you can also compare your work to others'.
posted by aquafiend at 7:24 AM on February 22, 2007

Thanks, I'll keep those in mind. Preferrably though I'd like to be able to take a class of some kind. Are there better ways to go about this than others?
posted by teemo at 7:44 AM on February 22, 2007

Check in with your local community college, or university, if you've got one handy. Even if there isn't a music production course, you might be able to learn some useful skills and meet other people who you can learn from in sound design type classes for theater -- that's where I learned what little I know about mixing.

I imagine Full Sail and other similar trade-school-for-the-arts places might have that kind of curriculum, but I've never pursued that option and I don't know what level of commitment and investment is required to take a course.
posted by Alterscape at 7:59 AM on February 22, 2007

Hook up with local bands. There's always someone recording who could use an extra engineer for their home studio (usually when one of the guys in the band is the de facto producer). If you're playing out, you should have no trouble finding them.
posted by klangklangston at 8:17 AM on February 22, 2007

Second the looking for local bands.

Bands are always looking for more stuff to record, especially in the days of Myspace!

And don't worry about quality, if you're doing it free people won't sound much, and believe me, you can't do worse than some I've heard!
posted by chrispy108 at 8:24 AM on February 22, 2007

See if you can find time to speak to an engineer. I went into our local studio/music shop the other day for some brushes and enquired about their setup, and ended up having a couple cups of tea and gaining so much advice from the engineer there.
Pure chance perhaps, but I'm sure if you asked an engineer if you could get an hour or two to gain some advice, over a beer even, they'd probably be happy to teach you something.

Otherwise, experience really does seem to be the best way of picking it up.

One piece of advice I can give you that I picked up that seemed to work really well - when applying EQ, find the frequency you want to gain, then bring down a range around it at about an equal amount. It lends clarity to the addition you've made without having to really increase the db of the track, which just brings other issues into the equation.

The other thing would be the quality of the mics and such you use, though you mention samples. But if you use shitty mics you can't engineer something into sounding any better.
posted by opsin at 9:03 AM on February 22, 2007

I second what alterscape said...my local junior college used to offer classes in Studio Recording (and they probably still do, it's just been many, many years since I've attended).

I learned a hell of a lot in just one semester, and even my old 4-tracks started sounding much better compared to friends that hadn't taken the class.
posted by malocchio at 9:19 AM on February 22, 2007

Maybe it's irrelevant, but a lot of the music I'm making is electronic and/or electronic based. Mic-wise, I have one and am working on the whole "singing" thing, but let's not go there for now. Mainly I'm after trying to get that elusive "pro" sound out of what I'm doing. And again, I've read through a million forums and articles and get the general gist, I guess I'm after having someone around who can go "no, wrong" or "yes" so I have a basis to work from.

I feel like I'm at that point where I know enough to not really know anything at all (e.g., Not even sure what questions I should be asking). Or when I do have questions, forums and the like seem to be a bit hit or miss, particularly when you have one question that begets 5 million annoying follow-ups requiring the patience of a saint.
posted by teemo at 9:20 AM on February 22, 2007

OK, seeing as how it's mostly sample based anyway, what software are you using?
Some things you can get a fairly pro sound out of without trying, and some you seem to have to know exactly what needs doing for that particular piece of software to get anything half decent. And some have elements set up to aid in mastering a decent sound.
posted by opsin at 9:34 AM on February 22, 2007

Also, I'd be happy to listen to something you have down, or throw it track by track through pro tools and see if I can come up with any advice from that.
I couldn't guarantee I'd get anything from that mind.
posted by opsin at 9:36 AM on February 22, 2007

What opsin said. Post some tracks and get feedback.

There are several ways to become better at mixing, and they are practice, practice, practice, practice, and listen to people's comments.

The first 4 things don't do much good unless you start hearing feedback. "the kick is too mushy" or "the hihat really is too shrill" or even "i can't tell what i'm supposed to pay attention to, it's all just THERE" are all useful things to hear, and even non-musicians can give useful feedback.

One thing that is ABSOLUTELY important: you need decent monitors. Even the "high end" computer speakers simply will not do, and headphones will not give you the right perspective on how your mix sounds in a room (they're also important but not as much as monitors).

Fortunately there are a lot of good and reasonably inexpensive monitors out there from companies like Alesis, Yamaha, M-Audio, and (the ones I have) Fostex. I couldn't tell you how much my mixing and production "ears" improved when I got a proper system in place a few years back.
posted by chimaera at 9:53 AM on February 22, 2007

A note about monitors - I was terrified to discover how bad studio monitors are. They are terrible on the low end, and all about clarity on the middle range.

Stereos and personal players all take this into account, so if you don't in the studio, you are going to end up with everything muddy and bass heavy when played back on anything else.
posted by opsin at 10:09 AM on February 22, 2007

From experience, I believe the best way to learn audio engineering (as a musician) is as follows:

1.) Write and record your own music. (Optional: do a bit of mixing yourself).

2.) Take a few songs (completely unmixed) to an experienced engineer for mixing. Make sure the engineer has the same software you intend to use, is friendly, and doesn't mind you looking over her shoulder.

3.) Watch everything.

It will cost you some money, but trust me, it's worth it ;]
posted by gaiamark at 1:01 PM on February 22, 2007

A pro engineer who doesn't care about being watched over the shoulder is a rarity -- or, I guess more accurately, she'd be likely to consider that a tutoring session. If you're straightforward about what you need up front, you'll definitely be able to find someone able/willing in this city -- for your best value, I think you should specifically look for a pro who advertises tutoring, rather than an engineer who's okay being observed but doesn't have any experience with training or any particular desire to explan things.
posted by allterrainbrain at 4:18 PM on February 22, 2007

"Mainly I'm after trying to get that elusive "pro" sound out of what I'm doing."

Well, speaking as someone who's written about music, don't do that. You'll end up sounding like a blah middleman, like most everyone else. You want to instead think about production choices and match those choices to what you want to accomplish with the song. If you want it to sound raw, you might want to do something like toss a mic up and record out of your computer speakers, rather than going direct. You'll have to think about the room that you're in, and how you want your voice to sound. If you just go for a professional slickness, it won't serve your music.

"A note about monitors - I was terrified to discover how bad studio monitors are. They are terrible on the low end, and all about clarity on the middle range.

Stereos and personal players all take this into account, so if you don't in the studio, you are going to end up with everything muddy and bass heavy when played back on anything else."

Ignore this and become more of a gearhead. It's true that your monitors will color your mix, and that you can learn to compensate. But it's not that studio monitors are universally bad, and anyone saying that is unlikely to have an ear worth trusting or an experience broad enough to say. Every monitor has a color, and you want to get as close to a flat frequency response as possible, because when you mix you want your mix to be playable on as many as possible other systems (though what you aim for on mastering changes the sonic character).
posted by klangklangston at 4:45 PM on February 22, 2007

I didn't intend to say studio monitors were bad per se, all I meant is that they tend to colour your music in a way you wouldn't choose for listening, and to learn that sound (your ears will atune to it anyway) and account for it and have something that won't be bass heavy. The NS10's for instance have loads of middle, and if you want to hear any bass in the mix you'd have to add something with more bass response.
posted by opsin at 6:49 PM on February 22, 2007

Thanks for all the great responses so far. And in response to a few of the questions/comments:

I have DynAudio BM5a monitors. Prior to that I had some M-Audio BX8's, but one of the crossovers went bad. So far I love these monitors a lot.

The tracks that I've finished (mind, not in a couple years now) are up at http://tekmonki.com

Agreed about getting good feedback, that seems to be a bit of a problem on the internet. Understandable that most people are just going to say "oh, not really my style" or give you one or two very general tips, hence me preferring to find someone I can sit down with and watch what they're doing and ask questions relevant to the task at hand.

Software-wise I'm using Live. Before that it was Reason. I've tried Cubase and a few others, but really love the workflow in Live. I found Reason even simpler to just knock something out, but the sound output didn't seem great and ReWiring or bouncing a ton of tracks started doing my head in after a while.

As for the "pro" sound, I agree. Though I also think that it can't hurt to be able to emulate whatever the latest and greatest in radio ready sounds is and then get creative from that standpoint. Right now it feels like if I do get something right it's simply because I stumbled upon it. I suppose that's the crux of the issue. I want to be able to go "oh, the bass drum is clashing with the bass synth, try cutting x frequency out" or "that synth is just not working at all, try another" instead of twisting knobs blindly and hoping I get it right . . .
posted by teemo at 5:59 AM on February 23, 2007

OK, as far as you not wanting to just twist knobs blindly to EQ, that's how engineers tend to do it... They may have more intuition from experience, but to learn, especially, you just have to experiment. It's the only way, you really can't learn it, and it's down to how you want it to sound.
If you feel that as you say, the bass drum is clashing with the bass, fiddle until you find what frequency it is you want to draw out in which. As I think I said, you then, ideally, want to pull some gain off in a wider frequency (using the q to adjust the width of the frequencies you're adjusting) around the one you are gaining. It may not always work, but by and large it gives you greater transparency for the sound you are bringing in, without risking maxing out the headroom, or at least leaving you something like your original headroom. So if you had a drawing of it you'd have a valley created, with a peak in the middle of it at the frequency you want. Try it, and then push the valley back to flat and see the difference, and watch the red line.

The other note I'd give about pro sound, is that it's mostly down to compression. If you're not using it, start, and learn how to use it. A 10:1 ratio meaning for every 10 decibels put in, 1 is brought out, when you've hit threshold, and the threshold telling it at what volume to kick in (if you already know, sweet, but so many just fiddle with it and don't actually know the principles of what they're doing). It may be preferable in many ways to avoid it, but since you're working on electronica it's fairly fundamental to getting the kind of sound you're after. And don't worry about compressing tracks, and then compressing the overall thing. That's the biggest, is run a last compression to bring it up to the final volume and make sure you're not maxing it out into the red. If you're having problems getting something to pop through something else you can also think about gating, where it will pull out the bass say, to let the kick through. if you gate it down 6dB or so, you barely notice it, but it can make the world of difference in picking out what you're after. Then again, with electronica rather than live intrsumentation, you should have far more control int he first place over what sounds are overlapping others. Still, it can be worth bearing in mind for the instances it might be just the thing.

As far as giving you advice on the sound, it's hard to tell what of the sound is the encoding, since the first track on the page doesn't sound transparent (I'm guesing at best it's 128k?). You lose a hell of a lot of quality changing it to mp3, and OK, it may be how a lot of people will listen to it, but for my music collection I don't use anything below 192k full stereo, as that is a hell of a lot less noisy than 128k. You can hear the difference. That and I record in 24bit 48KHz, so there's a huge difference between what I would mix and EQ the sound on and an embedded mp3.

I doubt even the best engineer could say that if a bass was clashing with a synth, to pull out x frequency. It would be a case of saying, you need to seperate them slightly, and fiddling with EQ 'til you have them coming through at different frequencies, but again, it's all about the fiddling to see. And as far as saying that a synth doesn't work and to replace it, what you need is a producer not an engineer. That is a personal judgement, one which either you are happy making, or someone else who is straightforwardly co-authoring the music needs to make. It sounds like you just need to gain confidence, which I understand can be a challenge, but it's your call. An engineer is there to record, and mix, and more often than not, not master. It's a finite role in the production of a record, and while some would happily wade in and try and change the music, it's not really part of the job description.

For what it's worth, it really doesn't sound far off to me. Needs some EQ in the mid to high frequencies to clean up some of the samples - hihat and snares, and possibly to pull them away from the high synth - but the bass seems fairly solid. Possibly fiddle a little with the bass synth and drums to seperate slightly, and I'd be inclined to pull a little bass out overall (it's approaching muddiness, very slightly, over my monitors when I put the sub on which I use for listening to finished music) but with a very little EQ and the right compression you could have the kick for instance sounding as pro as you could want. You can learn a lot about how to get the compression effects you want from the intertubes, with some careful googling.
It looks to me (I only used a demo of Live through Pro Tools, very briefly) like you may already be using compression on it, so it's probably a case of adjusting what you got.

Well, a bit rambling I know, but there may be something of use in there, and as I say, unless you have 15 years experience, it's all about experimenting to get the sound you want. I always figure, if you like it, someone else will. As for the pro sound, the more radio-like you want it, the more you need to compress it all to within an inch of it's life! Remembering that radio stations run compressors on everything they play to have it all normalised, adding even more to it. As I say, if you don't get anywhere, you can try throwing me each track individually and I can try and mix it in pro tools and let you know what I did, though whenever I worked on electronica I revelled in being able to fiddle with the actual synth or sample settings so much to get the changes in sound I was after.

Otherwise, read everything you can about it, even if something's wrong, if you can learn why it makes a world of difference. And ideally, from the sound of it, you need to pay someone to mix one of your tracks in front of you, as someone mentioned. I watched other people's music being mixed, but that gave me the kind of eyes on experience I needed. To see why and when an engineer makes the changes they make helps, but at the end of the day, it's not all that complicated, and you come away with - fiddle around and see. If you want to be able to make concise changes straight away with full confidence in what you're doing, you need to apprentice for a few years, or go on a really good course, as even engineers are never sure, they just don't mind playing around.
(sorry, more rambling!)
posted by opsin at 8:25 AM on February 23, 2007

Oh, and Live doesn't appear to let you view a proper frequency response graphic, so you need to find something that lets you see what your music is doing even after mixdown (preferable before!), to get it as flat a response as possible. And don't let the transients on drums fool you, you need to take an average of them. They will come in louder than most everything else, but all of that volume is gone within fractions of a second, so the average therefore will be deceptively low.

Seeing the sound helps a phenomenal amount in learning. I even wonder how engineers can work blind with analogue gear, as I am spoilt by seeing the waveform to trim or whatever in pro tools.
posted by opsin at 8:32 AM on February 23, 2007

Stellar response opsin, very much appreciated. I have some thoughts/questions/something on all you said but have to run off to a client, so more on that later. Oh, and yeah, those are low bitrate on the site (128 I think) so they stream better. I try to stay at a min. of 256 or pref. 320 otherwise if I'm doing mp3.

Thanks again.
posted by teemo at 10:25 AM on February 23, 2007

No problem. I'm happy if anything I've picked up can be useful. I'll keep an eye out for updates from you and see if I haven't exhausted my knowledge already!
posted by opsin at 5:26 PM on February 23, 2007

Well, I've since lost my exact train of thought on what I wanted to say, but just re-reading . . .

I suppose it does come down to a confidence issue more than anything. For example, I have a basic handle on compression mainly because I hired a guy for a few hours once upon a time to help me out with the general "mixing" issue above. The main thing I ended up getting out of it was a very basic guideline on compression, ratios, and such. Akin to that, I also suppose that it has to do with feeling like I'm working in a vacuum a lot. I'm sure it's a lot of school of hard knocks in this arena, but it'd be nice to have a network of some type to go "hey, I'm trying to do x, is this the right way to go about it?" instead of fumbling around on one aspect of a track for hours on end (yes, this is definitely part of the fun, but sometimes you just want to feel like you're making progress and not making a giant circle back to square one).

As for viewing frequency response, is there one tool better than another for this? And is it one of those things you want to put on the master bus and view everything as a sum? Or it's better to stick one on each and every track and view that way?
posted by teemo at 12:07 PM on February 25, 2007

I've had a look around for analysers for frequency response, and so far the best I've come across seems to be: Inspector.
It would mean using it as a VST plugin, which I don't know if Live supports. But if so, stick it on the master bus to judge the output of the whole track. I'll try and ask around and see if I can find something hopefully better than that, though I haven't had a chance to try it out yet (it has an RTAS version afterall). That should let you check the flatness of the response fairly well, though I'm still looking for one mimicking the hardware I saw used in the studio the other day...
There appear to be some others around which you can load a finished file into, which might also be of use to account for the encoding, but this ought to be of some help in the meantime.
posted by opsin at 7:03 AM on February 28, 2007