Making drum loops sound realistic

calimoose

New member
I'm victim to EZdrummer loops. I'm not a drummer and have no access to a real kit, let alone a properly miked one. So sometimes, I just pick loops and accept that they will sound like ass.

A fellow HRer provided some info to help remedy this sickness, and I think it would be useful for everyone to read (should you have similarly weak drum loops on your tracks):

There are four common occurrences that are dead giveaways for fake drums:

1) Drum programmers aren't drummers. Not much you can do about that, but being an actual drummer myself, I hear a lot of fills and beats that sometimes just can't happen with a real human drummer. A drummer has two hands and two feet.Two of those hands are hitting drums, the feet work the kick drums, hi-hat, maybe a cowbell, whatever. You can't have too many things landing on the same beat. How can two toms be hit at the same time as two cymbals? They can't, unless you have an octopus playing the drums. In that same vein, and this is the harder part for programmers, they try to fill too much space with too many drums. Real drummers like to bang around, but a good drummer doesn't cram every available space with fills, beats, cymbal crashes, ghost hits, etc. A good drummer plays the song. Listen to AC/DC. Phil Rudd does nothing behind the kit, but very few bands have the locomotive rock drive that AC/DC does. It's a like a Panzer tank flying down a mountain. Keep it simple and keep it like a real drummer would play it if you can. Naturally some genres like prog have more complicated drums. Then you can go wild, but still keep it within the limits of an actual human with only two hands and two feet.

2) Dynamics. This is always a dead giveaway. Programmed drums can very easily sound like a typewriter. A real drummer, even a good one, doesn't hit every drum the same way every time. I know that for myself the sound of my snare changes just by switching my other hand from the hats to the ride. When I go to the ride and "open up" my upper body, my snare whacks tend to fatten up due to not having my hi-hat hand riding over the top of my snare hand. Simply put, I can hit the snare flatter and harder. So that happens. And we all have a dominant hand. When rolling around the toms, usually a drummer's right or left hand will hit a little harder. Maybe he'll hit harder on the quarter notes and the eights might be a little softer. Like ONE-e-and-uh-TWO-e-and-uh-THREE-e-and-uh...and so forth. Stuff like that goes a very long way towards killing the typewriter sound. And of course, drummers aren't robots. We don't stay exactly on the grid at all times. A good drummer will use the click track as a guide, not a rule.

3) Cymbals. I don't know what anyone can do about this, but sampled cymbals always sound like ass to me. Good drum packs use real drums for samples, and they generally sound good, but I don't know what happens with the cymbals. You're just stuck with generic sounding boring ass cymbal sounds. Like the actual drums, using dynamics goes a long way towards making cymbals not sound all the same. A drummer won't hit the hats the same way every time. He won't hit a crash the same way every time. Vary it up. A good ride cymbal is like 4 cymbals in one. Where you hit it yields different sounds. When you're pinging the bell, mix in a few body hits.

4) The stereo spread. Drum programmers that haven't spent much time with real drums don't seem to understand that a kit is not 70 feet wide, and most drummers tend to set their kits up the same way with just a few little variations. The hats and ride are rarely ever on the same side. Toms usually go from small to big around the kit. Kicks and snares are centered. Have some spread on the toms and cymbals, but going all the way wide is distracting and sounds weird. Having a crash or tom pop in and out of one ear is just awkward. Keep it realistic. Find the toms in your fake overhead tracks, and pan the close mic'd toms in it's corresponding area of the overheads for a natural sound.

And then after all of that, don't over-process what are probably already processed drum samples. Further compression and EQ just makes them worse. Go easy. They probably already sound fine as-is. Real drums don't sound like plastic machine guns. Draw in your dynamics, timing, and use smart panning and drum samples left alone will probably sound much better than trying to murder them with further processing.


The last bit particularly resonated with me, as I often try and process the drums with mixed results. Cheers.

-CaliMoose
 
This a pet subject of mine... I agree with everything our anonymous poster has said (Hi Greg!) above.

And additional observation:

Despite the best efforts of software developers, it's still very easy to sound robotic - every time a real drummer hits a snare drum, it sounds slightly different because they cannot hit it twice in succession at exactly the same velocity, at exactly the same angle, in exactly the same place on the skin. This is a good thing.

These minute differences aren't noticeable so much in a busy mix with loud other stuff going on everywhere, but if you're in something a bit sparser, then robot snare syndrome is easy to suffer from.

A good software package will give you multiple samples, not just multiple velocities, so it's minimised. Cheaper ones may not. So running multiple snares and blending them together to give you a bit of shading of the sound can make your drumming sound more real.

Snares are one thing, but it's actually more noticeable on cymbals, where you often DON'T get multiple samples. It's not so much that the samples, in isolation, sound bad, it's that they don't assemble together well without a bit of effort.

And rides are the worst. Nothing worse than a ride cymbal slamming away in your head in robotic fashion. It sounds absolutely awful. I will always, always, always set up three different ride cymbals - the main one, and then feed in different small amounts of the others, either as a hit, an edge or a bell, on a random basis, to try to break up the monotony. And all this as well as differences in velocity - not just random, but emphasising this or that beat.

If the software allows it (AD does, SSD4 doesn't) you can get by with a single cymbal, and feeding edge or bell in in small amounts in the same way. As I work mainly in SSD4, I use the 3 different cymbals option.

I can quite easily spend 8+ hours working on the drums for a 3 minute song, and I almost never use pre-done loops, unless as a starting point.

It always amazes me that people will spend eons talking about and searching for the best recorded guitar tone, the best take, the best feel etc. with and then just go, "You know, I'll just use Generic Rock Loop A for the entire song because I don't know anything about drumming."

Listen. Learn. Put some effort in. It can be painstaking but it's not rocket surgery.
 
I never use loops or samples. I pefer to write my own. Time consuming yes but way more effecfive.
 
I never use loops or samples. I pefer to write my own. Time consuming yes but way more effecfive.

I definitely don't think it's rocket surgery (lol) but obviously writing your own drums is time-consuming, particularly when accounting the time it takes to learn from scratch. While it's something I'd like to do in the future, at this moment in my life I only have so much time and have to pick and choose my battles.

One day, I will attack the drums head on. One day...
 
Yeah it is time consuming but once you get into a rhythm, like anything else, you become more adept at it. Especially when building a song as you can copy and paste parts and then just slightly edit those RE: velocity, timing, etc, to sound a bit different from the original copy, or just copy them. Not many people would notice.

I'm no master at it by any standards but the practice over the last 2 years or so has given me a bigger confidence in writing my own drums tracks. I find it quite therapeutic and a relaxing way of working on songs and a nice break from playing instruments. I started out by using a sample and editing that to make it my own which is a good way to find your work flow, and a lot of it will come down to the DAW you use and it's midi editing capabilities.

When I first started using a DAW I was using Sonar 5 and hadn't a clue what to do with midi and could never fathom it out in that program, or Sonar 6. Reaper really opened me up to midi, samples, loops, etc because the workflow "works" for me and I found it a very natural thing to get in to.

I also recently started to learn to play the drums as well which as helped me loads. :thumbs up:

EDIT*

Another trick I find quite useful, if you have a midi keyboard, play the fills using the keys. For example, create a a drum track with your VSTi of choice and make the straight beat for the song as you want. Then create a second track with a drum kit and just play the fills in the parts you want them.

Then it's just a case of copy the fill parts into the original midi track and taking out the the bits you don't need, ie, double hits and bits that a real drummer wouldn't be able to do.

It's quite a good way of doing it, if you have the time and patience to give it a go.
 
Last edited:
Pretty much (almost) all the info above is correct and helpful, from everyone that posted .

I'll just say this, to help ram (RAM?) the point home.

Velocity is 90% of it, in my opinion. You CAN have snare hits that sound all the same, and you CAN have ride cymbal hits that sound all the same (In fact, I'd say that with real drums, there isn't much difference from one ride cymbal hit to the other). The difference comes with how a drummer would play their hats and ride.

A simple exercise to illustrate this is to program a very simple beat, with the hi-hat or ride playing 8th notes. First, don't vary the velocities of the ride/hats. For the sake of this example, put all the velocities at 127. Now, just take every second hit (the "and"s), and bring the velocity of those down to about 90. Right there, you already have a way more human sounding track.
 
I definitely don't think it's rocket surgery (lol) but obviously writing your own drums is time-consuming, particularly when accounting the time it takes to learn from scratch. While it's something I'd like to do in the future, at this moment in my life I only have so much time and have to pick and choose my battles.

One day, I will attack the drums head on. One day...
I think maybe a good place to start might be to build a song with the loops from EZD and then dump them into your DAW and edit the MIDI to work better for your piece. As I understand, all of those "grooves" are actually played by actual drummers, and should already incorporate a lot of what your quote was talking about. You might learn some stuff just from poking around in them.

Nobody mentioned timing fluctuations, but that's at least as important as velocity IMO. Nobody ever actually hits exactly on the grid except by accident. A good drummer will deliberately push or pull certain hits against where the beat should be. Sometimes there's an ebb and flow. There is always a small amount of randomness. So, like, maybe select every snare hit and nudge it forward like 10 or 12 ticks, then add a small amount of randomness (or "humanization") and then maybe use tempo changes to gradually speed up a little going into the chorus or something.

Somebody mentioned playing things in via keyboard, and that's not a bad option either. You don't have to play the whole kit at once. Even just take the thing you built out of loops and replace the ride or hat with live input. It's good if you can actually keep a decent beat going, but most quantized algorithms have a "strength" parameter so that you can tighten up a sloppy performance while keeping some of the human feel.

I don't play keyboards, though, nor have I ever practiced playing on those little rubber pads (some people do!) but I finally got myself a Roland GK system that actually works, and I've started playing in all my drums from there. Might be a little pricey if that was your only use for it, but for guitarists who want to add anything other than guitars to their one-man-band home recordings, it can be invaluable.
 
Since this topic really applies to me personally, & I've already gotten some more great tips here, I'll try to help a little....I wanna say first that I'm not a drummer at all, I even bought an e-kit a couple months ago, & while I'm learning to play 'em, it also shows me just how far I have to go to be able to call myself a drummer, & also gives me even more respect for you guys that can play.....

I've pestered the hell out of lots of people (some are here, some in other forums...sorry guys...:o) about this very subject, & I've got a bunch of notes put together from our conversations about it...

It is pretty long, but there's some pretty good info (to me anyway) here, & I've tried to omit things already mentioned in previous posts, so I apologize if some of this has already been mentioned...

Velocities need to resemble how an actual kit would be played, (on a roll, some people's left hand always hits a bit softer than their right, & vice versa, depends on the drummer, but you get the idea)...

Try not to have the velocity on the kicks/snares (everything really) 127, I think 124 sounds remarkably better on most vsti kits.
Everything being exactly on the grid can feel stale in certain situations. Try nudging certain hits forward or backwards depending on what feel your going for...

Don't overdo the humanizing. I know much is made of making vsti drums sound real, but you have to walk that line. A good drummer will be in time and mostly on the grid if playing to a click. It's the little tiny variances in power and time that make him sound human. Big variances make man and machine sound bad. Shoot for tiny variances...



In most cases beat one will be harder than the others, with beat three a little louder than two and four; extrapolate this for 1/8th and 1/16th. This is the first step to check on any regular rhythms on any drum...

Pushing or pulling the HH playing before or after the grid (do this as a last step) and using a humanize plugin on it (or even Reaper's included key-H one) to add a very small timing loosening and velocity variance will make for a more organic groove (1-3% on timing and around 12% to 20% on velocity will add a marked improvement)...

Kick is usually best on the beat, velocity can be more constant than other drums -although pay attention to rapid pre and post beat flicks and doubles as they need velocity reduction or may require the main beat to be adjusted to sound real. Play double kick (and to an extent double-pedal) on a separate instance of the drum or different drum (try at least a tiny tuning variation or comp/EQ setting, too)...

Snares respond well to velocity changes, esp on flams and drags, other ghost notes and fills. Don't feel you have to be anywhere near the hardest velocities unless the style really dictates it -and then you really need a bit of velocity headroom to accent that odd hit...

Try to audition snares with sub-100 velocities -a pad is great for this -and playing mostly at lower vel-numbers will help put some real red blood into the performance. You can even the audio level out with compression afterwards (like on an acoustic kit), but the initial velocity differences will affect the timbre of the drum in a realistic way...

Timing-wise, you can do the same sort of things to the snare that has been suggested to the HHs, but think about the emotional pitch of the song at each part and whether the snare needs to hold back or drag everything with it. Use smaller deviations than with the HH in many cases, and remember to deselect any hits with associated ghost notes before putting in any timing randomisation. Velocity randomisation on top of hand-programmed or played syncopation variance works well on snare too...

Put on your favourite recording of a band, listen to just the drums,and then just one drum of the kit and listen for variance. Not just the volume of the hits, but the timbre too. It will help you understand velocity "feel". Load some good played drum MIDI and zoom-in and look at the timing variations -it'll give you n idea as to how far-out good beats can be...

I know some people don't like being told this, but being a drummer helps here, because they know which beats hit harder or softer, and they usually know which ones they can't always get perfectly square. They know when the move from centre to rim and how hard to hit the cymbals (well some do) and how the "feel" goes on tom fills. Just how you'd expect any guitarist to do better with any guitar VSTi, it helps to know the instrument and how the feel is put into it -intentionally and otherwise...

Tempo Changes: Unlike a drum machine, it's impossible for a drummer to play at a perfectly steady tempo; many may seem rock-steady, but their tempo is actually almost constantly varying by tiny amounts such as 0.1 bpm. Slightly more noticeable tempo changes can occur leading into changes & fills when drummers get excited as they anticipate the upcoming change. Tempos may shift during dramatic or mellow parts & songs often slow down as they reach their conclusion. Being able to clock your sequencer or drum machine from a source that allows one to create a "tempo map" is the key to creating subtle & smoothly-varying tempos. Take your time creating the tempo map as it can't be changed later if you've recorded live, unsequenced tracks played to these tempos...


The "Extra-Limbed" Drummer: Don't inadvertently program drumset parts which could only be played by an octopus; just two arms & two legs...This may seem obvious, but it's not hard to make this mistake once one starts adding fills & crashes. Visualize/sketch the layout of your virtual drumset: could a real drummer hit cymbals or drums sequentially that are too far away for his stick to travel to that instrument's physical location in the time available?

Habits of Drummers: Now I'm not talking about things like complaining about how much trouble it is to transport & set up their kit or tuning/paradiddling while others are speaking. LOL ;-) I'm referring to playing habits which are nearly universal. Strong arm/leg hits(right for a right-hander) will be noticeably louder, so figure out which hand your virtual drummer would use for each hit & set individual strike volumes accordingly...

Most every crash cymbal hit is simultaneously accompanied by a loud bass drum hit. Any accented beat will affect all simultaneously struck cymbals/hats/drums making them all louder. 16th-note(& 8th note) hi-hat/snare/tom/double bass drum patterns are rarely, if ever, played perfectly, usually,the second hit of each pair occurs a little early. The first bass drum hit of any quick double tap will usually be louder. Playing the drum samples manually on a keyboard (quantizing off), being sure to use your strong hand where your "drummer" would, will often get you closer to the desired feel which can then be edited to perfection later...

Dynamics: Most drum samples are already compressed to some degree, so don't immediately start squeezing the snot out of anything except for carefully tone-shaping the toms to allow their ring to be more apparent. Anyone who has played with a live drummer knows the incredible dynamic range of a drum kit. Therefore, be sure to program this into your drum part. As mentioned in the previous section there is a big difference between strong & weak-limb strike volumes & I've found that one really has to exaggerate this when programming for it to be noticeable; this is especially true with the aforementioned 16th(& 8th)-note patterns. Rides & hats need a huge variation between the volumes of their accented & unaccented strikes to sound real...

The Sounds: When a drum is played louder, it gets brighter in tone, mainly due to the increased attack velocity. Make sure your drum sounds are programmed to react with a natural-sounding increase in brightness relative to volume & that your velocity curve/scaling is properly-adjusted if you'll be playing parts on a keyboard. If possible, use a softer variation of the sample for lower-velocity hits. I also like to roughly tune my drumkit samples to the song whenever possible as long as the sample doesn't start to sound weird or "chipmunk-ish" as a result. Another often over-looked phenomena is that the pitch of a drum changes a little depending on where, how hard, & how fast it's repeatedly struck. Therefore, if your sound source/module has a "randomize pitch" feature, be sure to use it at the lowest setting that will allow repeated strikes to "chorus" ever-so-slightly preventing stagnant cancellation(make sure this is 'off', though, while tuning samples to the song). Also enable multiple triggering on all drumkit elements so each hit rings into the next to prevent "machine-gunning" on fills.

One element I've never been crazy/patient-enough to try & simulate is the fact that whenever one drum is struck, every other part of the kit vibrates sympathetically to some degree producing additional overtones; I've relied on reverb & simulated "close-miking" to avoid the total nightmare of additional programming that would be required to emulate this subtle effect!

Panning is a matter of taste, but should reflect a real kit in the real world, keep in mind, though, that the lower the pitch of the instrument, the less directional its sound will be: meaning floor toms, for example, will sound more natural panned closer to center than orignially thought...

Mixing is also a matter of taste, just remember that smaller drums/cymbals are always quieter. Despite all of this "sound advice"(ha ha) you will inevitably encounter situations where none of your available drum samples sound right; this is when it's time to use the real thing. I often need to add real mic'd cymbals &/or hi-hats to achieve the desired sound or playing technique.

Once your "drumkit" is recorded it will sound like a bunch of individually recorded bits(which it is) & unnaturally dry. At this point, I start to craft a digital room sound for the kit to serve as "glue" to make it sound like one drumset(bass & snare on their own tracks, but toms & cymbals usually as stereo pairs recorded at maximum width & then narrowed to taste later). I adjust this room reverb keeping in mind that the loudest, sharpest kit elements will have the most room sound(bass, snare, toms). Once I think I've got the perfect sound I turn it waaay down, so low that it's only noticeable when switched on & off as it's not intended to be the kit's reverb, only a simulation of the virtual room in which it was "recorded"; the main verbs will be added later as needed.
 
Velocities need to resemble how an actual kit would be played, (on a roll, some people's left hand always hits a bit softer than their right, & vice versa, depends on the drummer, but you get the idea)...

Try not to have the velocity on the kicks/snares (everything really) 127, I think 124 sounds remarkably better on most vsti kits.

I agree with this wholeheartedly. Only a real asshole hits every drum as hard as he can every time. Real drums when at the velocity that 127 represents are WAY THE FUCK TOO LOUD and generally sound like ass. I think 100 is a reasonable limit for all but the most extreme crecendo type things...or maybe Led Zeppelin.

Dynamics: Most drum samples are already compressed to some degree...
This I disagree with, at least in the current context of ToonTrack's stuff. Some of the channels on some of the kits are sort of "special effects" that are crushed through a compressor, and some of the mixer presets include compression on the mix or a bus or something, but if I honestly believed that the actual individual mics or samples were compressed I wouldn't be using them. They record these things microphone>preamp>done, and that's what we get when we pull them up in SD or EZD. It's a common misconception, and may be true of other packages, but not these.
 
Real drums when at the velocity that 127 represents are WAY THE FUCK TOO LOUD and generally sound like ass.
Then turn them down. It's all relative, isn't i?. Velocity doesn't decide the volume of your drums. It just decides the volume of your hits IN RELATION to your other hits. If you decide that 100 is the hardest velocity you want, then you'll turn down your other hits' velocities relative to that. So, if you decide that 127 is the loudest you want your hits, you'd be turning down you other velocities relative to that, too. You then have a separate volume control for your drums independent of all that.
 
Then turn them down. It's all relative, isn't i?. Velocity doesn't decide the volume of your drums. It just decides the volume of your hits IN RELATION to your other hits. If you decide that 100 is the hardest velocity you want, then you'll turn down your other hits' velocities relative to that. So, if you decide that 127 is the loudest you want your hits, you'd be turning down you other velocities relative to that, too. You then have a separate volume control for your drums independent of all that.
Are we back on this? The velocity that you send to the VSTi changes the way the drum sounds (actually, changes the sample that's played back), and it is completely unrealistic. Just turning it down from that point is not the same as sending it a lower velocity. And with a real kit, there's no volume control on the kit, and turning down the mics won't change the fact that it sounds like ass.
 
Are we back on this?
Back on this? Your post that I'm responding to is 33 minutes old? Are you senile, or just on your period?

The velocity that you send to the VSTi changes the way the drum sounds (actually, changes the sample that's played back), and it is completely unrealistic. Just turning it down from that point is not the same as sending it a lower velocity. And with a real kit, there's no volume control on the kit, and turning down the mics won't change the fact that it sounds like ass.
OK fine. But that's not what your main point was in your post 33 minutes ago. You said it's "WAY THE FUCK TOO LOUD". That's what I'm responding to. It's not WAY THE FUCK TOO LOUD if you turn it down. The sound issue is another story, but you didn't put that IN CAPITAL LETTERS TO EMPHASIZE IT'S IMPORTANCE.

Fucking hilarious....:laughings:
 
...and just to point out more stupidity, a real kit doesn't "sound like ass" when you hit it hard. Real drums sound BETTER when you hit them hard. Anyone that says otherwise is either a non-drummer or talking out of his ass. In this case, it seems to be both. Stick to talking about drum machines. :)
 
FWIW, the stuff I posted was a copy/paste thing I have saved on a text file so I can pull it up anytime for refrence...I didn't write any of that, but there's a lot that's helpful IMHO...that's the whole reason I have it saved....:)

I agree with this wholeheartedly. Only a real asshole hits every drum as hard as he can every time. Real drums when at the velocity that 127 represents are WAY THE FUCK TOO LOUD and generally sound like ass. I think 100 is a reasonable limit for all but the most extreme crecendo type things...or maybe Led Zeppelin.
Different drum vsti's are different in the way they handle the velocity thing, IE, Slate Drums do sound much different than Toontrack stuff when the velocity is past a certain point...I usually make about 120 the highest I go with my velocity, usually...I will have some things at 127, but it's to make that hit stand out more if I do...Depends on the type of music too...


This I disagree with, at least in the current context of ToonTrack's stuff. Some of the channels on some of the kits are sort of "special effects" that are crushed through a compressor, and some of the mixer presets include compression on the mix or a bus or something, but if I honestly believed that the actual individual mics or samples were compressed I wouldn't be using them. They record these things microphone>preamp>done, and that's what we get when we pull them up in SD or EZD. It's a common misconception, and may be true of other packages, but not these.
You're right about Toontrack's stuff for the most part, but other vsti's (Steven Slate 3.5 comes to mind) are pretty processed right away, I don't know the differences in how they were recorded though...



Then turn them down. It's all relative, isn't it?. Velocity doesn't decide the volume of your drums. It just decides the volume of your hits IN RELATION to your other hits. If you decide that 100 is the hardest velocity you want, then you'll turn down your other hits' velocities relative to that. So, if you decide that 127 is the loudest you want your hits, you'd be turning down you other velocities relative to that, too. You then have a separate volume control for your drums independent of all that.
+1 dude...The velocity does change the volume of the vsti's hits, but it also changes the timbre of the sound too...Most, if not all of the drum vsti's I've used have a setting to allow the velocity to affect the volume of the hits, it's usually found somewhere in the humanize/random function...But yeah, you've got total control over the volume, velocity, everything in these drum vsti's...
 
I'm assuming the point attempting to be made here is that snare and other samples in some software, at 127, aren't as "usable" - something I tend to agree with in my playing around with AD and SSD4 in the context of the music I'm creating. I generally max the snare at about 100-110ish. Kick around 60, most everything else in that range for a normal "hit", with the VSTi and DAW channels set to unity.

But it is really context dependent - the type of music the drums are part of informs the relative usable volumes along the 0 - 127 scale - and just because I don't like the 127 level snare doesn't mean it's crap. Horses for courses...
 
My point was that the velocity changes the sample that is played back, and that usually the is not the best sounding sample for that particular drum in most context in the same way that actually hitting an actual drum as hard as you fucking can is not usually the best sound for anything. The thing about too loud was part of my point that nobody in their right mind would actually hit the drums as hard as they fucking can every time.

You can do whatever the fuck you want. Read the thread or don't. Read my posts or don't. Make stupid comments rather than asking for clarification or not. Set all your velocities to 127. Hit your drums as hard as possible every time. Whatever.
 
You can do whatever the fuck you want. Read the thread or don't. Read my posts or don't. Make stupid comments rather than asking for clarification or not. Set all your velocities to 127. Hit your drums as hard as possible every time. Whatever.
I know all that. I don't need your worthless permission. As far as making stupid comments, your posts in this thread should be in the stupid comments Hall Of Fame. And your attempts at back-tracking are comical at best.
 
With modern drum apps....this isn't supposed to be too hard.... :D

I don't know about other drum apps....but take Superior Drummer....there are included "humanize" options, which by default are usually on for all the included samples, and it is recommended that you leave them all on.
Why....because the program is doing ALL THE WORK FOR YOU to randomize velocities and timing of hits. :)
If your drum app can't do that for you, in most MIDI editors where you drop the grooves/loops/beats there is probably a tool that lets you select a bunch of beats or all the beats, and apply a randomization to their velocities...etc. You can set the high and low points, or you can pick a center point and tell it how much over/under the range for the randomization.

So really...it's NOT THAT HARD to NOT end up with a bunch of same-velocity, machine-gun hits or some rigid, perfectly quantized groove. Hey, you wanna move each beat's timing and velocity by hand...that's cool, there's nothing wrong with that, but unless you are using some antiquated or very basic drum app...you don't have to kill yourself to have a groove that sounds "real". ;)

Also....there is no absolute velocity "sound" based purely on its MIDI value. IOW 127 doesn't sound the same for every snare sample in your library. So...sometimes I've had my Snare and Kick hits all riding between 123-127....and sometimes I like them at 116-121...etc...etc. It's all relative to where you set all your other velocities and what samples you are using.

I agree with what's been said about the Hats....they are key to making the kit sound "real", even if subtle. In the MIDI Editor I keep at least 6-8 Hat variations....based on how open/closed they are, where the stick is hitting them, and then how hard. Even though I am using just one type of Hat for the kit, there are several variations of how it's played.

The other key to making it "real"...are the fills/accents. If you listen to a really tight drummer playing Rock or Pop....the Snare and Kick aren't all over the place, they are tight and solid constantly. So there's no need to go too nuts with making extreme variations in those hits....but when you break to do a fill....that's where the difference comes...and not just the fill, but how you go into it and out of it from the basic Snare/Kick/Hat groove. It's not about making that sloppy...it's about the feel of the transitions and the actual fill....and of course, a drummer has 2 hands, and 2 feet...he can't hit the Hat, the Snare, the Ride and a crash all at the same time while also doing fills. There is a natural way to transition in/out of them, and there is also a velocity ramp up/down...etc..etc.

Getting the right groove feel is what counts...and not so much your choice of Snare or cymbal or other samples.
The drum apps these days all have great samples, that were well recorded (better than most people can pull off in a home studio)...so you don't need to kill yourself CHANGING those samples with tons of processing and EQ....etc. Some people seem to think that it's absolutely necessary to always process the crap out of drum samples, that they need to do that to make them more "real", that a "raw" drum sample is never good enough... but I find that in most cases that just screws up the beautifully recorded drum samples...and then it sounds really fake.
Also...if your really must slather on the processing...at least wait until you have a mix going, and not on just the basic drum tracks without any mix context....but less is more when working with sampled drums, IMO.
The pros that recorded them already made sure they sound really good...and the modern drum apps do a great job of making the whole "humanize" thing easy for you...so you really just need to focus on the groove and pay attention to the feel. Listen to the drums and try to think of them being played....and not programmed.
 
...and just to point out more stupidity, a real kit doesn't "sound like ass" when you hit it hard. Real drums sound BETTER when you hit them hard. Anyone that says otherwise is either a non-drummer or talking out of his ass. In this case, it seems to be both. Stick to talking about drum machines. :)

This ^^^^^

The non-drummers need to drink a nice, big, ice-cold glass of STFU.
 
Back
Top