Drum mixing basics ?

[d-a-m-o]

So, I'm getting back to production after a long, long break (almost 10 years ) and since a lot of you are way more qualified than me in this domain there's a couple of Noob questions I want to ask, the main one being : what are the basics of mixing electronic drums ? Are you EQing every drum track before sending them in a drum bus ? When sending various drum tracks in reverb or delay bus how do you keep everything glued together ? What about the level of effect tracks ? what about compression ? Any help would be highly appreciated !

[chim]

Most of the time you'll want to do a drum bus with one track per hit (similar sounds like cymbals and hats can usually be grouped), and process them individually. For your drum hits you need to be careful with trimming too much fat on the EQ, even hi hats and shakers can have lots of goodness in the low mid range, try to maintain balance level wise instead. Any reverb should be room style and mostly aggressively high passed. You can get away with wet FX settings on high frequency content but the meat should be kept dry. But a whole dry drumkit is boring as hell, the trick is to have a good busy wet layer separate from the kick and main snare for rolls and ghost notes. If your track isn't very busy you can do more creative stuff with the meat. If you aggressively compress kicks and snares, try to control the tail as it will often pump up in an ugly way. It's better to have a meaty, short and snappy source sample and then either layer your own highpassed tail separately or customize a room reverb. 

Glue bus compression is more important for acoustic drums but usually means low ratio and slow settings. Lower the threshold until the gain reduction meter or needle sorta dances in sync with the drumtrack. As for individual levels and settings it's something you have to learn manually. EQ and compression is mostly for sound shaping, don't try too hard to make the drums sound fat, it's more important to just manage the levels of the rest of the music.

[Entorwellian]

1 hour ago, d-a-m-o said:

So, I'm getting back to production after a long, long break (almost 10 years ) and since a lot of you are way more qualified than me in this domain there's a couple of Noob questions I want to ask, the main one being : what are the basics of mixing electronic drums ? Are you EQing every drum track before sending them in a drum bus ? When sending various drum tracks in reverb or delay bus how do you keep everything glued together ? What about the level of effect tracks ? what about compression ? Any help would be highly appreciated !

I was beginning to answer this but just the nature of the questions its a huge topic and could go on for pages on each individual type of percussive instrument in your mix. Like how I would mix a sharp kick is completely different from a muted kick, a raspy snare versus a tom snare.

Basically use eq at the beginning to colour your drums, level them, bus them, then compressor and another eq that bus for any major overlapping frequencies that occur with other sounds in the mix. After that it goes to a send where i can do additional mixing with reverb and other effects. For the bus part though, I try to keep it as fx free as possible.

[sweepstakes]

This isn't practical advice at all but I fuckin' love a subtle reverb with some healthy pre-delay on a simple, understated snare, clap, or rim. It can add some odd groove to a bland pattern too. More of a cheap trick that I like.

[Zephyr_Nova]

Regardless of whether I'm recording acoustic or mixing electronic drums, I seem to always end up adding a medium-narrow band boost somewhere between 30-80hz on the kick (wherever the sweet spot happens to be).  Snare usually get a small boost somewhere in the high mids/highs (very broad spectrum of where that might be depending on the snare), and usually a narrow-mid boost somewhere closer to 150hz for some added punch.  Of course, since samples can be anything these things may not be necessary.  They just end up being the case for me more often than not.  I also end up pitching the kick/snare/toms to the key of a song, or enhancing some harmonic frequency that is consonant with the key of the song.  Tuning acoustic drums is a bitch, but sometimes its a matter of getting the drum key out for this same effect.  Ugh.  Anyway, with electronic drums it's so much simpler (just pitch it up for down, find the sweet harmonic frequencies, etc.).  Or conversely there may be a harmonic frequency that's too resonant and needs to be notched down to sit nice with the rest of the track.  I often use a stereo chorus or delay to add extra depth on drums instead of (or in addition to) reverb.  Depends what sort of sonic vibes I'm going for.  Or micing up the room and cranking individual tracks to get a natural ambience on them is another cool trick to play around with.  It's a lot more effort than just slapping a vst on there, but often the results are more interesting/unique.  It's got a way of giving electronic drums a more organic quality.

[chim]

Great tips there Zephyr. There are plenty of ways to think (and record) outside the box. I'd say that tuning drums is difficult for a beginner and it's not a catastrophe if you don't do it. I've only ever bothered with it on 808 style pure kicks. Stereo chorus and delay tricks is a must on dry claps, and can really sweeten hihats, shakers and lots of other percussion. It's easy to overlook the usefulness of sparse effects. 

I'd also add that you shouldn't neglect the actual drum programming as well, funky claps need to be slightly before your snare. Unless you're going for a really electronic sound, humanizing the MIDI (usually available as a randomization percentage midi tool in your daw) or recording a section of hits yourself can really breathe life into a stale rhythm. I especially like to do it with midi shakers. I could probably write a whole post about shakers alone, I love 'em! I also like to record my own claps in the bathroom and pan them around the place. 

[sweepstakes]

Agree strongly about chorus on hats! When I use that I like to turn the wet down until it's indistinct and just becomes a very subtle stereo boost.

Never thought to record claps in the bathroom - I'll try that (and some snaps while I'm at it) today!

[d-a-m-o]

tuning drums is more important with long decay drum hits right ? 

[sweepstakes]

2 minutes ago, d-a-m-o said:

tuning drums is more important with long decay drum hits right ? 

Exactly, I think that's what chim meant by 808 style.

2 hours ago, chim said:

even hi hats and shakers can have lots of goodness in the low mid range, try to maintain balance level wise instead

Wanted to call this out too - I was really surprised when I first noticed this but you can really take the life out of the track if you take too much low-end out of those.

[chim]

I'd say it's more important with pure tone drums, which usually means electronic bass drums and toms, or stuff like those really old drum machines that were built into home organs. Acoustic drums are usually sorta scrambled in the waveform but may have a very pronounced and pure body sound, in which case it is worth a shot. But I think we're pretty used to not registering drum hits as dissonant most of the time. 

[Entorwellian]

https://splice.com/blog/mastering-101-eq-basics/

 

Here's a beginners 101 guide for eq'ing. I agree with most of what is said in the guide. (there are a couple of others there for compression and limiting too). A big part on eq'ing too is going to be based on what eq plugin you use. Not so much for the gains, but for things like aliasing, whether it uses linear phasing or not, and how it processes short attacks and decays. If you have a lot of money, you can not go wrong with fabfilter.

I've also noticed, on rare occasion, that lower quality eq settings sound better than high. Especially for things with very sharp attacks in the lower end of the frequency spectrum.

 

 

http://www.guitarbuilding.org/wp-content/uploads/2014/06/Instrument-Sound-EQ-Chart.pdf

[d-a-m-o]

also, I guess you must adapt to the nature of your beat right ? I guess you don't use the same mixing techniques for a techno beat or for a super complex and crowded IDM type beat ?

[user]

One thing I tend to do a lot is use subtle to moderate amounts of distortion/overdrive and bit crushing on drums. I usually have something like decapitator on the kick channel, something crushing the bits of my hats (and claps work sometimes as well) and then maybe some additional warming/overdrive on the entire buss. I used to strongly dislike the sound of bitcrushers but started gelling with them more when used somewhat subtly on hats.  
 

Related to glueing everything together; I’ll usually mix into something like ableton’s glue comp that’s on the master bus form the very beginning. The basic settings are always the same; slow attack; fast release, have it do maybe between 1-2 dB of gain reduction and compensate for that. I keep adjusting as things are added to the mix/composition. I started doing this after reading a post by a mastering dude who strongly advised this and it’s definitely paid off making my shitty tracks seem somewhat more cohesive and work together. I’ll share the post if I can find it. 
It’s this post; https://www.blacklistedmastering.co.uk/triangular-compression-theories.php

 makes sense to me and works for me but of course ymmv and also ytmnd 

[Nil]

Random unsolicited advices that usually work :

Decide from the get-go which track, between the kick and the bassline, will go deeper / louder, and mix accordingly. Both will inevitably clash if you don't make a deliberate choice.

Mix the drums without soloing them : the way you'll process them really depends on the whole track. FWIW, I've found I get snares & co "right" when setting their volume, EQ, verb late in the mix, otherwise I usually set them way too loud. Globally I now get better results when treating each drum-hit like another instrument, rather than a part of a sub-group.

EQ each drum hit... only if / when needed. Use different samples or synths if you struggle to make them sit in the mix with the usual processing : most of the time, fixing issues right at the source sounds much better (you can"t polish a turd yada yada yada)

Gates/expanders are the drums best friends : I believe we all tend to have drum hits with long decay... but then it sucks out the punch. So the ADSRs and gates/expanders are key here IMO. They combine damn well with buss compression too.

Sends for verb / delay sounds better than direct inserts most of the time, with the additional benefit to possibly EQ/compress/gate etc... the signal to be processed pre/post FX, without affecting the dry sound.

I used to love drum buss compression, now I'd rather just use gentle  mixbus compression instead. I guess I overdid it ? Anyway, don't fear slow attacks

As time goes by, I believe it's harder to properly mix something as sparse as a techno beat, more than something pretty busy : when you just have a 4/4 kick and a hi-hat, there's not much left in the drum kit (and often, in the whole track) to hide bad mixing decisions. Now, techniques are music genre-blind, whatever suits the vision you have for your tune/mix.

[Braintree]

I use Logic and I tend to separate each percussion instrument into separate tracks. Kick, snare, and all hihats, for example.

The basics are pretty easy. Compress then EQ in your effect chain. Set the EQ for kick to roll off everything below ~30-40Hz (depends on style of music). Set the EQ for snare to roll off anything below ~115-200Hz. Set your hihats to roll off anything below ~250-300Hz. These numbers are just where they tend to sit. You'll probably always need to adjust. However, each one should be separated to perform in a separate part of the audio spectrum. Each one should sound a bit thin on their own, but full when everything is together. Kick and snare in the middle. Pan the rest to taste. Ta-da.

There's a LOT more you can do, but that's generally how it starts.

[sweepstakes]

On 3/9/2020 at 2:11 PM, Nil said:

Gates/expanders are the drums best friends : I believe we all tend to have drum hits with long decay... but then it sucks out the punch. So the ADSRs and gates/expanders are key here IMO. They combine damn well with buss compression too.

If you can draw your own envelopes (like in Renoise), or even just destructively modify the drum samples, you can sometimes get that same punch without a compressor (or dial it in more precisely). Dip quickly after the transient and come back slower, or just layer the slap/thump on a body with a slower attack.

[dingformung]

parallel compression trick: make a send effect track from your entire mix and compress it heavily with fast attack and boost the highs and lows. Then slowly mix this send effect in with your mix so you get a bit more thickness and consistency in the lows and highs without cluttering up the mids. Also, that way you boost the more quiet parts of the track while leaving the peaks and transients intact. Very subtle but nice (relatively obvious) trick.

 

serial compression trick: shave off the thin transient peaks with a fast attack and fast release compressor and then give back some of the transients and punch with a slow attack and fast release compressor. thickens the transients.

 

Oh and DeEss your delays if you think it sounds good.

 

[Nil]

^ Great tricks. The first tip might create some phasing issues with some DAWs, depending on the way they compensate for the latency potentially induced by the compressor plugin.

If using Live, you could instead bus the mix into a Group, then fake a send within the Group using an Audio Effect Rack and 2 chains : a dry one and a processed one. That should do the trick (I use similar techniques all the time, can be super useful to fake a Dry/Wet option when not available, or to run weird / complex FX chains in parallel... with the possibility to easily process the whole)

Won't be an issue at all with hardware though.

2nd tip is great as well, you can even try a clipper or a limiter instead of the first comp.

[Nil]

On 3/12/2020 at 3:02 AM, sweepstakes said:

If you can draw your own envelopes (like in Renoise), or even just destructively modify the drum samples, you can sometimes get that same punch without a compressor (or dial it in more precisely). Dip quickly after the transient and come back slower, or just layer the slap/thump on a body with a slower attack.

Absolutely !

I should have precised that gated, individual drum hits, react extremely nicely to buss compression. 

[Squee]

21 minutes ago, darreichungsform said:

Oh and DeEss your delays if you think it sounds good.

What?

Also, MidSide EQ - pretty important as well.

[dingformung]

24 minutes ago, Squee said:

45 minutes ago, darreichungsform said:

Oh and DeEss your delays if you think it sounds good.

What?

It's a bit of an oddball trick I guess but sometimes delays sound better when they are smoothed out a bit while full on low pass might be too much. DeEssing is nice for smoother transients and a softer sound, not only on vocals but on everything you want to set in the background

[Nil]

Would you then de-ess the input or the output ? I'd spontaneously do the former... I guess ?

[dingformung]

41 minutes ago, Nil said:

^ Great tricks. The first tip might create some phasing issues with some DAWs, depending on the way they compensate for the latency potentially induced by the compressor plugin.

If using Live, you could instead bus the mix into a Group, then fake a send within the Group using an Audio Effect Rack and 2 chains : a dry one and a processed one. That should do the trick (I use similar techniques all the time, can be super useful to fake a Dry/Wet option when not available, or to run weird / complex FX chains in parallel... with the possibility to easily process the whole)

Won't be an issue at all with hardware though.

2nd tip is great as well, you can even try a clipper or a limiter instead of the first comp.

Yes was concerned about phasing issues too at first but at least with Cubase 10 Pro this isn't an issue. It compensates the delay perfectly, you can even use look ahead.

1 minute ago, Nil said:

Would you then de-ess the input or the output ? I'd spontaneously do the former... I guess ?

I would only DeEss the wet part and not DeEss the dry part

[d-a-m-o]

thanks for all the tips everyone, keep them coming !

@Nil I was watching your TydalCycles video and noticed that everything is super clean/organised  in your ableton session , do you got a basic project template or tips to share regarding gain staging / level monitoring and stuff like that ? 

[Nil]

Thanks mate !

I try to keep it all as clear and readable as possible, my template keeps on evolving , and I just try make it as simple as possible. I have an almost hidden group, handling all the MIDI routing from Tidal to the 9 MIDI tracks (the 1st one being actually a Drum Rack filled with a couple of Hive 2 instances, handling real-time drum synthesis).

Then each MIDI track has an empty Instrument Rack (I love those and use them extensively), that I fill using synths (sorted via the Collection/Favorite thingie on the top left corner of Live's browser). I gain stage super early in the process, using Klanghelm VUMT Deluxe in VU mode (with 0 VU @ -24dB), as soon as I'm satisfied with a synth patch actually. Whenever using any sort of effect, I'll use TBProAudio AB_LM as I tweak (so that I'm not fooled by volume discrepancies), and eventually will fine tune with another VUMT instance. It can sound too quiet, but it's a non-issue actually, when working with any modern DAW. All I need is to export the final mixdown in 32bit FP, and it's easy to do ultra-transparent volume boost when mastering (aside from any sort of dynamic processing I mean. Most recent ADC/DACs are super clean quiet as well, so run them at moderate level when recording too : I've read countless times Motown's Bob Ohlsson and Kush Audio's Gregory Scott saying that often their sweet spots is ~ -16dB Fs (yes, for peaks), and I tend to trust them. As long as you record at 24bit, you should be alright. I barely work with recordings.

I also often have a peak at MeterPlugs Dynameter (also when mastering), which is super useful to check dynamics.

I have Airwindows Monitoring to dither in real-time, and I can hear its action now (truncation is a real bitch) : Live outputs a 32bit  FP audio signal, my sound card (SPL Crimson) a 24bit one : it has to be dithered in between. If I had outboard, I'd also dither the signal meant to be sent via this outboard. I also use Airwindows PurestConsole, it has its way to enhance depth and clarity that I love. The channel instances are best used post-fader, so I've faked such routing by using Instrument Rack's volume output / pan, and PC is inserted last in chain. Works perfectly. PC Buss is first on my 2buss.

Generally I try to keep all processing very minimal, and spend 80% of time tweaking melodies, synths, arrangement and faders rather than EQing. When everything sounds great raw, then I switch on Crave EQ on each instrument, EQing notes rather than Hertz. I barely ever compress, aside from the 2 buss.

I'm a send/return convertee as well, for all reverb and delay duties (and fuck I love reverb).

Now I've been monitoring super quietly for almost 2 decades (we could easily have a chat sitting next to my monitors as I produce, without the need to talk louder), and I believe it helps getting the overall "big picture" with more accuracy. The Amphion are tremendously useful monitors, even when whispering.

Now I'm also learning to organize my Tidal code (for further live-tweaking once the studio versions are done), and that's also a huge challenge. I've also set a lot of very utilitarian functions there to let the creativity flow when needed.

Sorry for the long answer, could go on and on.

TL; DR: I tend to work on my template when I don't feel inspired to write / jam. Idea is simply to make workflow as pleasant and immediate as possible, so that I don't need to think twice on the "how". I try to hide everything I don't need. And I also rely on a very, very limited set of tools (that crucial for me). I have too many options already, and am stupid enough to buy new plugins whenever I hear something sounding better than what I already have / whenever it can speed up workflow.

Gain-staging is a rewarding discipline to implement, as it makes mixing easier (like, a lot). Can take a little while to get used to it, but you'll get used to it super fast and it'll become a reflex rather than a chore in a heartbeat.

Monitor at reasonable level, and stick to it 90% of time : we all need the predicability and consistency.

As always, I hope it helps !