Detecting Click Tracks
jamie found a blog entry by Paul Lamere, working for audio company Echo Nest, in which he experiments with detecting which songs use a click track. Lamere gives this background: "Sometime in the last 10 or 20 years, rock drumming has changed. Many drummers will now don headphones in the studio (and sometimes even for live performances) and synchronize their playing to an electronic metronome — the click track. ...some say that songs recorded against a click track sound sterile, that the missing tempo deviations added life to a song." Lamere's experiments can't be called "scientific," but he does manage to tease out some interesting conclusions about songs and artists past and present using Echo Nest's developer API.
It's required to make use of drum editing and multitrack syncing. If I were to record garage rock album i would throw everyone in the same room and just play the songs. However to leverage much of the flexibility and power of a digital recording you need a click.
Around here... I wonder if they are using a click track?
On a serious note, I do like the warmth of older music, and my listening tastes tend to meander around the times between 5 + 30 years before I was born. (Child of the 80's).
As much as a tech nut I am, I still believe there are certain area's in life where it should be left at the door.
# cat
Damn, my RAM is full of cats. MEOW!!
the drummer from linkin park spent 8hrs a day for 3 months practiciing to click track before the recording sessions started...and this was for their 2nd album...not the 1st...
what is making things sound sterile is simply crap pop music that is also waaaaay over produced. not being rhythmically correct.
Sometimes there's an obvious speed up or slow down on a song, and in those cases you don't need software to figure out if there's a click track. A quick way to check is to compare the very end of the song and the very beginning. It's similar to acapella singing, sometimes there's a slight change in pitch. If it's not so much that you notice in the middle of the song, then it's not worth worrying about.
There are great albums that used click tracks, and great albums that didn't. Obviously a metronomic sense of tempo is a good asset for a drummer to have, especially if they're looking for session work. But a sense of dynamics and texture is, in my opinion, more important. I'd take an interesting drummer over one that just subdivides everything any day.
Then again, some songs benefit from the drum machine sound. It's all about the vision.
I don't consider a click track on a studio album to be cheating any more than a photographer using a light meter. In a live setting, however, it's a different matter. Not that I've seen anyone actually use a click track live (except for people attempting to sync up with some other prerecorded track and did it out of sheer necessity).
Actually, yes, most musicians need some sort of "click track" if they're playing in any sort of ensemble. It's just that in an orchestra or band setting, they're called conductors. In modern rock/pop bands, they're called drummers.
I do some recording/mixing and have had the privilege of working under a Grammy winning recording engineer (and phenominal musician in his own right).
Great comments here- yes, click-tracks have been around since the 70s (maybe 60s). Tempo throughout a song can change too much without some kind of metronome. It doesn't have to be an actual click track, just something to guide the musician laying down the first tracks. Just because a drummer or other musician listens to a perfect tempo click track doesn't mean the timing will be "sterile". We're still human! However I know some drummers who are scarily close to perfect timing- without metronome.
Most better click track generators have the ability to randomize the timing a few percent (adjustable). One major midi-based recording program that I use (MOTU Digital Performer) calls it "humanize". You can "quantize" a track to get timing, then "humanize" it.
I'm an experienced drummer and I play regularly with, and without, click tracks; I can tell you that the assumption that "feel" or "groove" is only present when a drummer's time varies is not accurate.
... I'm tempted to have a play with it.
There are at least two types of variation that matter in a drummer's performance: the overall sense of time and the moment by moment variations. The ability of a drummer to play a complete number and keep to a set tempo is really important, particularly in this day and age of digital editing. But it is a common feature of "click track performances" for the drummer to sway ahead of and fall behind of the beat (faster and slower). If done correctly this variance in tempo will add significant life to a performance and such a skill takes a lot of practice to perfect.
The subtle qualities of a drummer's performance go far beyond whether or not they stick to a given tempo for the duration of a number; this is just one variable that effects the quality of a performance. Some genres require a rigid sense (metal/electronica) of time whilst others benefit greatly from its absence (fusion/jazz).
Interesting software however
I play keyboards for two different worship bands at my church, and I discovered a pretty amazing trait that our drummer/leader in the morning service has:
He doesn't change tempo unless he wants to.
At all.
To elaborate, as that sounds sketchy unless you know how I learned it:
I'm a pretty rhythmic keyboard player, and one of my favored techniques (especially if I need to fill in empty space from, say, a missing electric guitarist in addition to the other textural stuff I was doing) is to use multi-tap delay and really accurate timing to build rhythms and and evolving chords. It can be a really fun effect.
I don't use it much, though, because even with a tap-tempo delay, which I have in my rig, it's really awkward to stay synced up with the rest of the band. My delay is pretty accurate (built-in effect on the Nord Stage, which is rather high-end. I'm pretty confident it's got sub-millisecond accuracy), and I can stay tight with it, but even decent drummers can have a hard time with that (let's hear it for teachers that make you practice with metronomes, eh?), so I usually have to adjust the tempo a few times throughout a song, and that can make things get ugly fast. A less-than-decent drummer, which is all too common, can't stay consistent enough for me to even try it. Thus, I don't (or didn't, I should say) do this much at all, despite my fondness for it.
But, when I first tried it with Bob (the aforementioned drummer), I was shocked, because it just worked. I tapped in a tempo on his first measure or two, and it stayed tight the whole way through. I really hadn't expected that result - hadn't occurred to me humans could be that accurate.
Naturally, I started trying this in various places where it fit, and so far, I can't remember a single attempt where it didn't stay synced. Granted, I haven't tried it with really dragged out delay times (nothing above about two beats of delay at maybe 100 BPM), but even so...
This is the best of both worlds, because when you need him to be rock-solid, he is, but when the situation calls for it, he can (and does) manipulate tempo intentionally.
I've told him (and others) that playing with him is like having an expressive human metronome, and I mean it. It is amazingly blissful - I can wander out into strange netherworlds of syncopation and/or ethereal tempolessness (yay for pads!) and the foundation never wavers.
I'm sure that at times, he has small amounts of drift, but given that my delay stays tightly synced with him for whole songs at a single tempo, it can't get as large as even a single beat per minute very often.
We haven't tried it yet, but someday I'd like to try him out against some sequenced stuff - I'm pretty sure that if I could handle it (which I don't think I can, yet), he'd be unphased by it, even if it got pretty thick. Live band + sequenced riffs/textures/effects could result in some pretty cool stuff.
So, all that to say:
The guy who wrote TFA is actually just providing a measurement of how consistent the drummers for these bands are. Maybe they used a click track to achieve that consistency, but as a semi-pro living in central PA (not exactly renowned for its music scene), I've found one who doesn't need the click.
"Oh, I like geeks way better than I like humans." - Mari Sarris
From TFA:
Um...really? You can't make a click track gradually change rate over time? Or follow whatever kind of variation you program it to? That's news to me. I thought computers wuz like all smart 'n' stuff.
"A great democracy must be progressive or it will soon cease to be a great democracy." --Theodore Roosevelt
People do indeed have different skills... or are sensitive to different torture, that being a different way to put it. Like many thousands or tens of thousands of folks, I volunteer to run soundboards for various local organizations, but don't claim to be anywhere near pro, but perhaps because of the substandard equipment and layouts I've worked with over the years, I've apparently a rather developed sensitivity to overdrive/clipping, threshold feedback loops, and "tape on the dashboard" effect.
One that gets me regularly is overdriven/clipping distortion. The other nite, someone at work was playing "music" on their cellphone. "It's pretty loud for a cellphone" she said, while it was about all I could do to stop from running away yelling with my fingers in my ears. The poor 1/2 watt or whatever speaker was distorting so much it was worse than fingernails on a chalkboard! Momentarily I had the opportunity to reach for the thing and turn it down perhaps 30 percent or so, without much loss in volume, but a HUGE improvement in quality due to the fact that it wasn't over-driving/clip-distorting any more! MUCH better!
I used to work across from a church, with speakers shaped like bells hung in the "bell" tower. They'd play recorded bells. I guess they finally upgraded to CDs, but before that... Have you ever heard the effect of stretched tape on a bell recording? It was actually funny sometimes, watching people smile and turn to listen to the "bells"... then hear the draaagg and pitch-bend, and realize it was only a (very streeeaaaatttched) recording... or worse yet, not realize it, commenting how nice the bells were, while I and others stood there gritting our teeth.
Sitting in the audience at anything "live" can be most discomforting on occasion too, hearing the threshold telltales that say the system's /this/ close from going into the dreaded feedback squeal, yet being bound by politeness from jumping dozens of rows of chairs and half way across the hall to turn the thing down a notch NOW, then notch the resonating frequency out of the EQ after the immediate threat is passed. I end up just sitting there, ready for the fingers in the ears if the squeal actually does hit, but otherwise outwardly calm and of proper decorum, whatever internal struggle to resist that leap might be going on.
Yet most folks don't notice a thing. What's especially "interesting" is when the guys with the "phat" car stereos or the like ask what I think about their system... yeah, it's loud enough, but the bass is all rattling (apparently to some, this is the mark of "good bass" ) or the tweeters are whining in your ears.
But, like I said, I don't claim to be pro. I do like to think I at least know enough about it to recognize a decent one tho. I've always held that a great audio engineer can often make a bad performance at least tolerable, but one stroke of a fat finger at the sound board can well ruin the performance of the best, and a sound guy that doesn't know what to do to stop the squeal (or rolllingg bbooom), or knows what to do but has such underpowered equipment (and/or poor positioning) he must choose between lack of volume and constantly running at feedback threshold (maybe not even an EQ to notch out)... forget it.
Duncan
"Every nonfree program has a lord, a master,
and if you use the program, he is your master."
R Stallman