Hygro's Production Tips
Edit: For new readers, there's a wall of text war going on some posts down. Feel free to skip the war, it doesn't finish because I realized it completely detracts from the the thread and decided to stop responding. Thanks!
I decided to start a tips thread because I keep learning or figuring out new tricks at a fast pace and I think some of you might enjoy them.
This thread will have a completely random set of tips. I will be updating it frequently over the years. If I become famous for this I will continue
Questions welcome, both related and unrelated to what I've already discussed. Disagreements and other takes welcome as well.
Before I begin allow me to qualify myself: I have no qualifications. I have no releases. My one submission to a label resulted in a temporary flop . I have like 3000 listens across 3 soundclouds over 3 years. I've been "producing" for "5 years" but only really taking it semi-seriously for about 30 months (2.5 years). I work on music an average of 30 hours a week. Out of my 10,000 hours to reach mastery, I can claim about 4,000 hours. However, that concept also requires that those are real practice hours, and guided by instruction above your level. In that case, consider me closer to about 1000 over the 5 years. Currently I am working on a game soundtrack for an indie game studio.
That said, my production skills have reached a point in which I can make professional enough sounding dance music. I've played my own tracks at clubs and have seen people keep up the dancing. I won 7th place in a remix competition judged by a major label's subsidiary for a track that wasn't close to my talent level. The general opinion was that I had the "cleanest" sounding mix. A friend of mine, a music degree graduate from a top music theory/composition program in the nation likes that I really understand music.
Ok here we go, tip of the day: Just figured this out last Friday while playing around.
1) When you have a draft far enough along that you have most of the sounds melded together--drums, bass, bass compliment, lead, pads, sufficient effects, hook, needed resonance, etc, put a low pass filter on the master track. I recommend -12db, .71 size (that's about a 2 octave span) cut at about 380 Hz. What this does is simulate really foamy and shitty headphones, but it also allows you to really understand the interplay of bass frequencies.
This will inform you two things: does it sound good for the earplug people, but more important, it lets you know what your groove really *feels* like. Transients, the quality and energy of the mids and highs, etc, hi hats, they all can mask or cover the real gut level groove of a song, so muting everything except bass instruments isn't even enough. You really want the low frequencies of all instruments without their high transients to feel this out.
This can help identify a lot of problems, from muddiness and a lack of clarity, to a lack of it-factor, to just a poorly designed beat. It can also show you how strong or awesome your piece is, and that you have a strong foundation for everything else.
Last edited by Hygro; 09-08-2012 at 02:09 PM.
This is a semi-facetious, loads-of-exceptions-to-these-rules thread I started a while back with some tips I may write more about here later. http://www.djforums.com/forums/showt...advice-is-true These are kind of starting points for beginners to take more seriously than they want to, until you learn when and why to break these rules.
http://www.djforums.com/forums/showt...ng-cool-synths is a way to make unique sounds. I have some really good ones from this.
"You must spread some Reputation around before giving it to Hygro again."
Originally Posted by Finnish_Fox
Originally Posted by RDRCK
Haha no worries, but thanks for the holler.
2) Parallel Compression
This one is fun. Sometimes you run into an issue where you need distinct drums with big transients, but you also want the feel of routing all your drums into one fat compressor. Instead compromising halfway, you get to do both! Have your drums largely uncompressed. Route them all to one buss channel. Apply hard compression. Really get that pumping in there. Nil attack, low threshold, high ration, big gain boost. Then mix volume levels to taste. What you get is warm fatness without toooo much added volume, and you keep mad dynamics.
You can also play around with other paralleling. I often use parallel compression on one buss, and parallel distortion on another. You could try overdrive, bitcrush, vocode, whatever who cares. EQ, whatever. Ok, now I'm just talking about buss effects routing but the point is to get a strong sound without a loss in dynamics, these are some ways to do it.
3) How to make your effects work in the right order, aka LEARN YOUR DAMN SIGNAL FLOW!
DJ Producer Tim of Blix Cannon taught me before I ever started producing that the most important thing I could learn was "signal flow". Simple--you send a signal and it flows through a series of modifiers until it comes out as a sound.
Your signal will probably be a midi instruction. That instruction will go to an instrument. That instrument will go to your output. Sounds good, yeah? Wait, we forgot a few steps.
Your order of effects matters a lot. You have an instrument and it has this cool sound, but you want some EQing, some big hall reverb, some compression, the sausage fattener (don't do that), an echo, a bitcrusher, plate reverb, and side chain compression. Let's ditch the bitcrusher, plate verb, and saussage fattener leaving us with:
2) Synth (in this example, it will be a fast attack, fast decay, short release pluck square wave playing melody).
3-7 in no order yet) EQ, Verb, Compression, Side Chain, Echo
3-7: It matters a lot the order of 3-7. Every effect will work on all the effects before it in the sequence. Do you want to compress your instrument, and then EQ that sound, or do you want to EQ your instrument and compress that sound? For example, if you want to cut the bottom everything below 300 and above 8000, you might choose then to compress the remaining sound and boost it all together. Or maybe you want to compress in the bass because it gives the midrange a nice flavor and then cut off to free those frequencies. Some recommend compressing after cuts, but before boosts (you could even have two EQs, or more).
So lets say you need that reverb. Here's where it gets obvious. If you compress before the reverb, the reverb will play the compressed sounds, which will sound pretty normal. But if you compress after the reverb, you will effectively make the reverb louder and stronger and not tail off gentle. This can be a cool effect, but it's not what traditional reverb is supposed to do.
But you also want echo. Before or after the reverb? If you do before, then you have your pluck, and it echos, and then you get reverb tails on all the little plucks and echos. But if you echo second, you end up echoing the reverb tails. It's a matter of taste, but A) mixing the two can get nasty, fast. B) echoing the reverb gets the sound a lot more washed out and will take up even more space in the mix.
Alright, now lets say you have your EQ cutting, your compression melding, your echo giving some rhythm, your reverb giving some juicy flavor. It's all pretty intense, but it's lose some of it's rhythm, the drum isn't cutting through, and you want more bounce to your mix. So you decide you want to sidechain the sound to the kick. This one will be the most obvious in relation to the flow. Briefly, that means every time the kick hits, the volume of this signal you are working on will drop by an amount for a duration of your choosing. That Bennassi Bros "Feel Alive" pump sound if you are being obvious, or just a subtle bit of breathing room for the kick to be louder.
So you put your side chain at the end of the flow. That's pretty common, I do that a lot. Usually not with pluck instruments but maybe your melody is on the offbeat, who cares. But now your cool hands in the air, spine tingling echo reverb combo is getting completely washed over. So maybe you want to put the side chain earlier. Or maybe not, maybe you want that washed over sound, but you want the kick to slice through it. But lets say you don't. So you put your sidechain earlier, let's say after the echo but before the reverb. This way, the instrument itself gets ducked, but the reverb can still can still wash over. But every echo that happens on the beat gets squashed. Note, if you compress after your sidechain, its like the reverb and it kind of defeats the point of sidechaining. EQing after sidechaining is just weird, I don't see the point. I do see a point in EQing after certain reverb though, but it's generally preferred before it.
Ok, you're good. Something doesn't feel quite right though. You wish your reverb only worked on the main pluck, to put that pluck in the skies over the listeners' heads. But you want that echo more down low bouncing around the ground. To really make it rain, you want them separate. But how, Mr. Hygro, is that possible?
We now go to step 8 and 9 together. Move your echo and reverb to separate busses kind of like I talked about above with parallel compression. These busses are not where the current sound is being routed--leave that alone for now. These are shooting off like their own branches grown from the instrument but independent of each other. This way both your reverb and your echo only effect the pluck, but don't interact with each other. (Bonus points if you sidechain the reverb buss to the echo buss so that the reverb ducks whenever an echo hits. I never do this, but why not?)
Now what? Ok, let's pretend a couple of other things. Let's say your lead melody is actually two instruments. So you have this cool effected plucking synth, but you also wanted a fatter saw wave to hit an octave under it. You really want the two to gel so you are going to route them together to compress them and EQ them together so that they fit nicely.
The lead melody buss will be step 10.
But where do you route your reverb and echo busses? And what about your sidechain, now that that instrument is feeding into a channel with some more compression? You have a few options here. But long story short, whether you route your echo/verb busses to the lead melody buss or not will change the sound. And if you move the sidechain from the instrument to the buss or not will also change the sound.
Ok, so everything is routed the way you want it. You reach the output, part 11. On the output you decide you want to color the sound so you put a multi band compressor (multipressor) on everything. Be careful how you set that up because the previous signals are getting affected by this. For example, lets say you cut the lows off everything from the bass and the kick, which you already made perfect to your ears (in the low frequencies, by using tip 1). If you have a multipressor after your already compressed kick and bass, you're going to change that sound and might only want to then compress mids and highs. But if you have a good bass sound, but you have a lead or pads that dip into bass frequencies, maybe you want some light bass frequency compression. Etc.
Know your signal flow, and be very conscious of where you put your effects and busses.
4) Picking a Reverb.
As a Logic user I have a lucky built in reverb setting called "Space Designer" in which different reverbs were modeled out of real places. Exploring this has better helped me understand more normally synthesized verbs.
Reverbs are tricky. They add so much to a mix but can cause a lot of trouble. They can muddy things up, they can take away all kinds of definition, and take up way too much of your overhead. So before putting reverb on things, ask yourself why? What's it supposed to do, emotionally? What kind of space is it supposed to invoke?
For example, most reverb is trying to signify something is further away. Reverb is a great tool for "panning" front to back. The more reverb you add, the more that track is in the back of your mix. But in real life, the standard reverb sound is very bassy. Bass travels further than mids and highs, so the more bass biased a reverb the more its pushed back in the mix. The more reflections your reverb has, the further back it is (there's more walls and things for the sound to bounce off before reaching your ears). And of course room size puts it back more.
So if you wanted some kind of warm pad kind of lulling the listener from afar like a tidal wave, you might want a long reverb thats not too bright. In the kind of music I make, I want my instruments to be more present, and I need to keep the low end clear for the sake of the groove. Because I often want melodies or strings or pads to kind of come from the sky, I'll pick brighter reverbs, or do my own edits. Cut out the bass, boost the highs, etc. This often gives a more in-the-sky or the ceiling feeling, more electrifying. I like longer lasting reverbs for my more melodic sounds so as to give a more harmonic, dreamy feel. To not lose definition on the original instrument, I make sure my reverb is bussed out, so it's not actually on the instrument's signal flow. I often share one reverb with many instruments.
If I need a sound to feel really big and mountainous, I'll go for a deeper reverb but cut more space around it so that it works. I'll use a shorter reverb if I'm just trying to get a sound to gel further back without making the reverb a thing.
Often times reverb is best when its not noticeably while you listen, but it would be noticeable if you took it away. This depends on the music you are making. With upfront sounds, like pop vocals, this especially holds true.
When reverbing drums I take a few approaches. Sometimes a tuned, really struck percussive sound will be great with a longer reverb, but usually I like to use "plate" reverbs. Plate reverbs are very short and add some real color to the sound without being like in a castle. The same principle of EQing applies: the darker it is, the more back the sound will be, the brighter it us, the more up the sound will be (reverb doesn't make things sound more forward).
I use a really bright plate on a lot of my snares. It gives it a kind of epic feel, but to do it right, I need to make sure the sound isn't bleeding too much. This means not having too much other high pitched resonance going on at the same time. Try putting a short, bright, snappy reverb on your snares and keep the sound space around them pretty open. Sounds pretty sharp, eh?
But basically ask yourself this:
1) how much of an effect do you want it to be? Flavor or feature?
2) where do you want the sound to come from? Lower or higher? Closer or farther?
3) what's this reverb attached to and how does it affect the original instrument's purpose? You you want something more fluid or something more present?
Also, you might want reverb to happen in a panned way. For example, you have an instrument on hard left, but you want it still alive on the right. Buss a reverb panned right. Sometimes you want two different reverbs on one instrument, either in the same space or left and right. Be careful having different reverbs left and right--it can feel very freaky and disorienting as your brain is calculating your position in the world as in two very different spaces.
Quick one before I go:
5) When you need a reverb feel but none of them work:
Sometimes you need something to feel reverby but not matter what it just washes out the mix. Try echo/delay instead. Get a bunch of repeated plucks/vocal transients etc instead of a tide. They will cut through but still give you what you seek.
if one of us were interested in sending you a production piece after "completion" would you give it an ear? I'd actually really like your take on something I completed not too long ago, and potentially for future pieces
this is true, how you put your sound through devices can significantly effect the final product. For example - putting a synth through distortion first then reverb second will sound entirely different than going backwards, it's common sense but many of us forget that as we produce.
Originally Posted by Hygro
Of course! I'd be happy to. PM a private soundcloud link or post it publicly on the production forum and let me know to check it out.
Originally Posted by franksij
Even when I'm going through my recent pieces I'll still see signal flow mistakes. They are like typos--very easy to make.
Originally Posted by Celestial
The following three are about creativity and all get at the same things.
6) Always Say Yes To Adventure.
This was my life motto in 2009 and part of 2010. Crazy times . While in music this doesn't mean ending up trying to find a coffee shop in a Rocky Mountain city hoping to get wifi not knowing why exactly you are there, it does mean that if you have an inspiration you go with it. If you have a sound you are enjoying, record it. Make it. Take it somewhere. Some of my best works have come out of this, but in fact rarely do I end up using the actual sound I'm making. You might think you "said yes" to going out to some underground warehouse that night with the adventure being dancing with a bunch of ex-raving Burners with sweat dripping from the ceiling, but actually it leads to the next day's road trip with a hot, shaved-headed pierced nose girl in search of the perfect place meditate behind a waterfall. This never happened, or course, nor do I owe her brother $400 for the tow truck that pulled the car out of a muddy ditch 60 feet from the road.
Anywayzz, go with the flow and Follow the Energy (my current, far healthier version of the motto). This leads to
7) Don't make genre music
This will contradict a later tip I will give. Nevertheless, don't make genre music. The 90s Electronica wave has a thing or 10 to teach the 2000s trance-electro-progressive-dubstep meets 2010s EDM crowd. They didn't make genre music, they made music. Fatboy Slim and The Prodigy didn't say "we're making Big Beat to top the Big Beat scene". No, they made the best electronic music they could. Same effectively with Daft Punk, Chemical Brothers, Faithless, etc. Their albums are full of weird tracks. They outsold, out performed, and "out-artisted" today's genre DJ-Producers by a hemisphere.
If you are trying to make progressive house, you will invariably sound like everyone else, only worse. You will avoid all kinds of creative ideas because they aren't "prog" enough. You will shy away from awesome basslines, brilliant tempos, beautiful melodies, and undanceable leads simply because you are trying to conform to the music you are most excited to make. You don't choose the genre, the genre chooses you. Your best bet is to follow the energy, say yes to adventure, let your heart and ear guide you.
I write really advanced, interesting melodies compared to a lot of dance music producers (you won't often hear them in my tunes given what I choose to release). I do this because I got pissed off and decided to stop trying to make house. I ended up making shitty 80s music (which leads to my next tip). More on this later. Fast forward, I'm currently working on a video game soundtrack and was banging on a broken piano in a moldy, smoke stained Berkeley Co Op, tapping an old Neo Geo arcade's buttons and hitting the rims of some tossed-in-the-corner drums. My friend was playing guitar. We were jamming and talking about it later, and just talking to him and listening to his work made me understand chord progressions in a whole new way. My melodies took another leap. My basslines improved. This was after I started this thread.... This could not have happened if I was trying to make synthy house music by myself. (you bet I'll have one on collaborating)
8) Make shitty music and impose limits.
Pissed off that my house music sounded like a bloated whale trying to beatbox, I decided to deliberately make shitty music trying to be shitty. What's shittier than 80s synthesizers? Nothing. They are so bad I can't believe they became fashionable last decade. Which of course was the whole point. They're awesome.
But they are so bad, to make music sound good they had to make great music. The synths were really only to catch attention and be novel. 80s pop music, even lofi productions by musical novices, kicks the ass of most modern dance music. This is because modern dance music can hold a moog saw wave on one note and add so much awesome effects, side chaining, and harmonics it can become a hit. They didn't have that luxury so they had to combine the feelings of simplicity with catchy hooks, advanced melodies and chord progressions, and really intricate basslines. Ok, joking about the last part, most of their basslines where tapping the same 8th note over and over again the whole song. Furthermore, they were recorded onto 4 track recorders half the time. That meant no separate compression for your kicks and snares, and the synth and bass were stuck on one track, recorded together with no room for error.
So try this if you are new: limit yourself to no more than, say, 12 channels of instruments, make your instruments awful and bland and tinny, and then see if you can make the best song you've ever made. I'll bet you'll surprise yourself, and have more fun than usual.
(Sometimes I like to host production battles with limitations, but rarely is anyone game.)