Constructing new rhythms from classic breakbeats

Tips & Tricks
5

You need to do it like the oldschool (classical) composers. Imagine the sounds and know which notes and timing they have independently of the real song tempo With just a piano there to get a better idea of how harmonies sound with each other, because imagination is not a full orchestra but just some cloudy ideas, but still… Yes, it can be done, and it needs experience and training, but then you are able to construct beats taking as much time for defining details as you like. For imagining beats it can be useful to translate the drum hits to vocal syllables, i.e. “Bop” for bassdrum, “Tek” for Snare, “Tik” or “Tssss” or “Trrrrrr” for…then find combinations of syllables that are like words, you can get inspirations by listening to the original breaks and traslating them to somekind of rhythmical babble you could rearrange or try to form new words from. Then you need to translate the single actions of the beat in your head to the tracker timeline, dividing it into blocks and then locating each hit can help, obviously proper line highlights are also useful for this.

As for keeping the original groove…in jungle style people would not use single hit samples, but samples with multiple hits in succession in original pace, keeping the original timing a bit. For single note hits, you could for sure extract a groove from the original break, and apply it to a mechanical set pattern, to make it live… I’ve sometimes tried finetuning a groove by hand using note delays, its real tedious work and you’d just want to make very few variations and then apply them to the whole song, but if you have good feeling for the lags and so it can be done. Rendering slice markers to phrase can extract timing of hits from a breakbeat you marked properly.

6

I had been focusing on the timing of drum hits and proper positioning of notes within the pattern editor, however certain times I could not get the notes to sit right due to their length when chopped from the original break. This is often noticeable in the case of samples with multiple hits. Sometimes the hits extend beyond, or fall short of the point at which the next sample is meant to be triggered, even though I have applied beatsync to the sample to make it fit perfectly into one pattern length before chopping for rearrangement. However I think this is simply a result of poor chopping choices, and/or wrong song settings i.e BPM, LPB, and pattern line length, that cause the samples to not match up perfectly. Particularly in the case of sped up breaks, where a new song BPM combinde with a sped up break throws hits out of position from the grid. An example would be say: a hi-hat section chopped from a break that when triggered runs for too long and falls out of time not finishing on a beat but a little off

Would I be right in saying this happens? Is this what needs to be resolved when you say you to use note delays to finetune hits?

I really like the idea of translating hits to syllables and forming words, thats a great tip. I work with typography a lot so sounds, and the arrangement of sentences makes sense to me in terms of structuring and arranging phrases. I never thought to translate that into my music, usually I would just imagine the drum sounds as they were.

Thanks OopIFly :slight_smile:

7

I only have a generic comment … and that’s … “DON’T GIVE UP”.

I know what you’re saying, man. I’ll start chopping something and it just ends up pretty shit, so i start taking things away and it starts to sound like the original. But I play this tug-of-war game until some new idea pops emerges, then I try to layer on that. I go through many layers of utter shite before I come up with anything half way groovy that isn’t a straight rip off.

Keep it up, don’t let the groove police get you down.

8

A comments a comment, no matter how generic.

It’s just nice to be talking to people who are into the same thing as me. I live in the middle of nowhere in the countryside, and no-one around me is into drum and bass, or even know what it is at all. So it’s kinda just me on my own, trying to figure stuff out from old internet posts, manuals, and plain old experimenting, and dial turning.

Closest people into any form electronic music I know of are the techno heads, but they all live miles away in the city.

So the forums are the place for me I reckon :yeah:

Don’t plan on giving up anytime soon. I haven’t even really touched on the topic of melody, and harmony cause I don’t own a synth yet, but I’m saving for a Minilogue. So I got that to look forward to next. Getting the beats stuff down first in the mean time.

Thanks for the encouragement Mars.64

Much appreciated :walkman:

1 Like
9

Well, you’re very welcome!

Two thoughts:

  1. If you’re interested I’m sure there’s someone around here that would be down to screenshare/collab type thing to get the creative juices flowing

  2. Ireland, hey! Where abouts? I don’t know anything about the area, but I do know that Code, Mecca and Ricky Force are out your way (maybe hours away but … Wardance would be worth it! Subtle Audio is one of my favorite labels ever, so I gotta pimp them :)).

10

Here’s a simple way to get flowing drums: program one or more phrases that sound good to you. Then sequence them using 0Sxx to set the line offset. This way you’re not playing with individual hits anymore - you’re using small parts of phrases that already flow well.

1 Like
11

Not sure if it has been said, but a simple way to program beats is to figure out a basic kick/snare pattern. Even with chopping you will know where your upbeats/downbeats will go, breaks are often jazzy/syncopated so it helps to know the basics (even if you just sound it out). Just take stock samples for that, program the kick and snare, then just chop breaks over them, add layers, other breaks, then just remove the stock samples or replace them. A lot of the complexity to breaks comes from the high hats, ghost snares, other percussion, or the “flickas” as some people put it, those are important to work with. Don’t treat them as one shots or overly chop your breaks beat by beat. Breaks work by the sound between the hits, the reverb, the ghost patterns slightly off time etc. Don’t mess with it and over complicate your programming. Take those in their entirety and use them to fill things in. If you analyze some breakcore tracks you will hear nothing but high hats from different breaks overlapping in the background and foreground. It’s really neat! (never try to program that haha, the magic is in the breaks original performance and the other things mentioned)

That said, you’d be surprised how common patterns become and how a simple pattern can sound really complex sped up. But remember, velocity, timing, different sounds from hitting the snare and hats in different places, really minute stuff adds to that complexity. Chopped breaks sound really crazy, but when you hear them in their original track, they are often super slow and boring. I like fast music, but you never make music fast, so the advice to slow things down is really key.

Don’t over analyze things, as others have said, slow things down and focus on one part at a time. A composition is written one note at a time, computers (or should I say midi :P) distort that, but really if you think of the full break arrangement at once you will get nowhere. Break it down bar by bar, zoom in, analyze it, program it, go to the next. When you zoom out and hear how it flows, it should be banging. Though really everyone works differently, taking little bits of info from each person helps build your own style, practice is the one thing everyone does. So make sure to do lots of it!

Or you could do this: http://www.dailymotion.com/video/x1zyx5_we-do-not-need-any-beatslicers-insi_creation?GK_FACEBOOK_OG_HTML5=1

ps. this video is gold

pss. https://www.youtube.com/watch?v=o7N4QhlDXJw This song I think really outlines what I’m talking about. Rotator often uses a really aggressive kick and snare on beat to emphasize a steady dance beat, and just layers the flickas all over the beat to make it sound crazy. Each one is from a different break, sped up, distorted, pitched differently etc. Layering is so important it’s stupid. Some people take it to like a collage like level where it’s just too much but yeah, I spent a lot of time just listening to this stuff. (generalization obv ->) Realizing it was all made on samplers just blew my mind but when you understand there isn’t any beat by beat chopping but taking small segments and arranging them it kinda clicks. Idk if I am explaining any of this (v. drunk) but I hope something makes sense

3 Likes
12

mars.64

A collab might be cool, I’ll look into it further down the line once I get a bit less withdrawn about sharing my music :slight_smile:

I’d never heard of Code, Mecca, or Ricky Force before, I didn’t know there was even a DnB/Jungle scene in Ireland at all. Stuff sounds real good. I live in a really small village, in a county called meath. The places those guys live are about an hour to three hours away depending on which city you go to. That’s not too bad, but sadly it’s just not practical for me at the minute. Although I’ve bookmarked them and I’ll keep note of them in the future :walkman:

pat

So far I’ve avoided the 0Sxx command for triggeringat specific point in full break loops, for the simple reason that jumping between the pattern editor and the sample editor seemed to slow and laborious. However I know that it can be done to good effect with practice like everything else. When you say program a phrase, do you mean program a phrase within the phrases section of the sample editor, and then use the 0Sxx command in the pattern editor to trigger the phrase from new positions each time? If so I never thought of doing that, I haven’t really used the phrases section much yet, but if that works that would be amazing. Being able to create a new arrangement and then offset trigger it for new rhythms instead of having to type in new sequences in the pattern editor would be so handy :dribble:

4kb

Nice advice. I’ve often had difficulty with chopping sections of hi-hats effectively. They just never seemed to sound right, they didn’t have that flicka sound that I heard in other tracks. I think possibly this was just poor programming, and using dull hi-hat patterns from poorly recorded breaks. I found that in the last days practice of working at slower tempo I was able to hear the flicka sound better and program it more effectively through multiple one shot samples with slight sound variations, volume, and 0Sxx offsetting. I haven’t tried this yet with chopped hi-hats patterns but I’ll give it a go and see how I far out.

I often get into the collage level with my programming, and I agree that it can be too much. It can sound amazing and get great results for me, although it can become a mess thats easy to get lost in as well ruining the flow of the song. But then again I have a perfectionist personality, so I kinda enjoy the mess and trying to sort it all out :slight_smile:

Thanks for the advice. Hope the hangovers not too bad :badteethslayer:

13

There are many factors to consider and one way could work well in one case but fail in an other. My best advice is to experiment and then experiment some more and some more.

Here’s a little example i cooked up in a hurry:

1 Like
14

When you say program a phrase, do you mean program a phrase within the phrases section of the sample editor, and then use the 0Sxx command in the pattern editor to trigger the phrase from new positions each time? If so I never thought of doing that, I haven’t really used the phrases section much yet, but if that works that would be amazing.

It does work, and it is amazing. Try it!

15

Usually i imagine the process in my head,

then i patch the things, play Renoise experimenting and shaking my head.

It can go away from initial concept [in some ways],

but till i enjoy it, i make the repetitions of this process then [tuning and improving the methods i like].

This works for chopping also. Just play and push the record button sometimes and you will find your way of chopping!

In fact there so many ways to chop in Renoise, i haven’t seen so much in other instruments.

16

TheBellows

Thanks for the example file. Gonna pull it apart and study it. Reference files are always a big help to me. Makes it easier for me to get hands on with the process rather than watching tutorials. Cheers for putting in some acid and bass lines too, drums would have been enough but you went the extra mile. A true gent. I haven’t really touched on much modulation yet beyond simple ADHSR, and LFO on the volume of drum samples for humanization purposes. So seeing all that other stuff going on on the bass and acid lines has opened my head to newer possibilities :slight_smile:

Pat

I’ll give it a shot so. Meant to try it earlier but kinda got side-tracked there today doing some other stuff, with trying to hook up hardware. The wandering brain that I have just won’t settle for long enough. :smiley:

Oise

Yes I usually start with a beat in my head as well. But stray away from it as I go along. Sometimes it’s good, sometimes it’s not so good. It can disrupt the flow of the song, as I get more and more into the patterns. However, I’m still trying to work on my overall song composition, at the minute it’s very loose with no real defined verse - chorus -verse kind of approach, which I personally like, I feel like it keeps things fresh,but that definitely adds to the complexity of the programming.

That’s very true. I’ve been figuring out tons of ways to chop and program beats. I had tried before about a year ago to make some tracks in a DAW, but it just didn’t feel right to me, and I could never get the beats to sit right or sound all that good. Working with the DAW seemed very linear and stagnant or something to me. Needless to say I kinda just gave up then after about 2 months. But since I took up Renoise about 3 months back I’ve found that it just suits me much better. Even though I’m still just learning, I already feel like I can do way more than I ever could have with a DAW. However that’s simply personal preference. What works for me may not work for another. :slight_smile:

17

one little tip - when chopping into individual hits use an alternating loop on the hit tail to give you some extra sustain/release time and more freedom with your sequencing

18

Here’s my approach to creating new breaks from existing ones.It’s time consuming as f**k, it gets really tedious and can drive you nuts,but it allows near-perfect timing when rearranging.
First and foremost, get your hands on Ableton Live, it’s time-stretching algorithms are near flawless for our needs. Ableton offers a 30-day trial, even a demo would be sufficient if you can record the output to an audio file.Set the tempo, import your break to an empty audio track. Before I get down and dirty with it, I chop off any parts of the break that I won’t need which would leave me with a bar or two of a perfectly timed break after all the processing.Then I’d follow Weyheyhey!!!'s advice:

  • [Edit] Make it mono first!

  • Pitch it up (by semitones, not arbitrary)

  • EQ it (aim towards a flat eq profile, you can do this easily with Ozone’s “Matching” EQ feature)

  • Multiband Expansion (reduce rumble from the kickdrum present in hi-hats, snares etc)

  • Multiband Compression (beef up the kickdrum)
    Those two steps are only done when the source break is of crappy quality, most of the breaks I’m currently using have already been pre-processed at some point so the dynamics are on the ball and ready to fire away.

  • Quantize
    This is what we need Ableton for, although it could probably be done in any other DAW (I’ve done that in Cubase as well, a bit more time-consuming than needed). So we got our break nice and clean, the transients are clearly visible and you can see where every hit starts and ends. Now we remove all the warp markers created automatically by Live which has a tendency to miss the initial attack of the hit sometimes and we replace them with our own, adjusting the timing and placement on the grid - longer hits are usually 1/8, shorter 1/16.While this could make some breaks rigid and too quantized, the most popular ones are usually played with very good timing and the adjustments are rather small and it would be quite hard to notice them (Soul Pride is a c**t though, I’ve wasted countless hours trying to get it right - more on that later). If done properly, you should have now a nice tight break that’s looping perfectly regardless of your LPB in Renoise.Re-arranging any break created that way is pure math and it’s quite easy to keep the flow.Even extracted single hits will still remain in time. Check the file to see how the whole thing looks in Renoise - I couldn’t be bothered :blink: with creating anything out of it since I’m using my work PC for this, but it should give you a general idea and be good starting point.https://www.dropbox.com/s/2bq08zqxz4jgavb/175.xrns?dl=0.

  • Compress (at this stage I just do it generally to “gel” all the breaks and give them roughly the same volume so they don’t really stand out when used all together).

And that’s it, hope it helps mate!

Last but not least, Soul Pride, the biggest c**t of all the breaks ever created by a live drummer. After listening to “Beep Street” about a million times here’s some of my thought on how Squarepusher nailed it (Jenkinson is a bloody genius, it’s still beyond me how he did the things he did with his Yamaha QY700). So, the break itself has two distinctive parts: first 7-8 seconds chugging away with a steady kick drum and a few harder snare hits and rolls ending with a crash cymbal , and the second part roughly around 25 seconds long - full of rolls and fills ending with a double crash. You can clearly hear that first part being used in “Beep Street” up to about 1:08 - the bass sound actually blends in with the kick drum almost all the way through the song so I reckon that most of this has been used as a backbone. The roll just before the crash cymbal seems to be played with different ending point and some additional snares layered on the top of that part, which would of course remain in time as the whole break has been stretched to a particular BPM.A lot of later parts of the break are just sections with different start and end points picked from the original (comparing to the original break, it’s mostly 1/4 and 1/8 of a beat). TBH arrangement like this would be almost impossible to be recreated from single hits and I don’t have a clue how to even start, but this is what I can tell from what I hear :).

19

[…] This is what we need Ableton for […]

No we don’t? You can do all of that in Renoise only. Chopping up a break in its individual hits and re-arranging them is so piss easy in Renoise I could even do it in my sleep. Furthermore you don’t need to time stretch breaks unless you deliberately want to go for that artifacted warp effect sound. I examined your Xrns and there’s honestly nothing special about it, nothing that can’t be edited and sequenced in Renoise.

More complex breaks like Soul Pride have a lot of groove going on … but honestly, all the individual drum hits are in line of 1/8 and 1/16 patterns, I don’t see how anyone could possibly have difficulties to re-arrange them?

To demonstrate my point I attached a small file with a break that - at first glance - has seemingly a lot of groove and complexity to it. But it’s all just 1/4, 1/8 and 1/16 drum hits that just have to be arranged in respective to their length.

20

“although it could probably be done in any other DAW”

I know it could be done in Renoise mate, this is just my approach to creating breaks. I can’t imagine chopping breaks in Ableton though :slight_smile: All I use it for is just to make sure the break stays in time - that’s what the .xrns is for, just to show that the breaks are spot on with the BPM.
That’s some nice choppage man, if you think you could replicate “Beep Street”, it would be awesome to see the results.

21

Here’s my approach to creating new breaks from existing ones.

Sorry for the late reply man, works been keeping me busy.

Anyways, I haven’t gotten my hands on a copy of Ableton yet, hoping to soon.Some real great stuff here to work with. Especially the multiband expansion and compression, still in the process of learning about all that stuff but from what I’ve tried so far it’s beefed up the sound of the breaks real nice. I actually started pitching the breaks up in semitones not long before I read your post, before I’d been pitching them arbitrarily and it was screwing up the timings of the hits in the breaks, so they never matched. So I did the math and figured out how to speed up two different tempoed breaks so they match correctly.

My work flow was basically this:

~ Set the tempo

~ Set pattern length to 192 lines, LPB to 24

~Load a break

~ Choose a pitch in semitonesthat fits with the tempo as close as possible, creating an almost perfect loop,using the computer keyboard, [i.e original file played at C4 – play break one octave up at C5 {12 semitones}]

~ Find the corresponding beatsync value that matches that semitones pitch

~ Hit transpose
~ Deselect beatsync[now the break plays 12 semitones up but is positioned at C4]
~ Apply pitch shift plugin to sample effect chain, down pitching in semitones as close to the original pitch as possible without losing clarity particularly on the hi-hats
~ Auto slice
~ Remove slice markers if too many, move if off position[all slices are now transposed and will change in pitch to correspond to song tempo if editing needs to be done at slower tempos}

~ Find the corresponding beatsync value that matches each slices pitch

~ Round the beatsync value of each slice down to even numbers i.e if a snare has a pitch that matches a beatsync value of 7, round it down to 6. Even numbers add up better mathematical.

[I suppose this is like quantizing but I’m not sure]

~ EQ

~ Compress
~ Destructively render slices [each slice can now multilayered with new samples if necessary]
~ Create DSP effects chains

~ Create Modulation effects chains

And that’s pretty much what I’d been doing recently. I don’t know if the way I’m using beatsync would constitute as quantizing or not I just figured that because I’m working with 192 Lines at 24 LPB, that beatsync values which divide into 24 would result in perfect sync. So basically 24 = 1/4, 12 = 1/8, 6 = 1/16 and that way sliced phrases would transition smoothly into one and other.

As for Soul Pride, I’ve stayed away from it this far. May take a look at it and see what my sanity is like by the end :PI read your analysis of Beep Street and listened even closer and heard everything you described. He really is a gifted lad. His arrangement of slices are fantastic, I suppose having a trained background in jazz helps.

I’m pretty certain now I know what my issue is when it comes to constructing new rhythms from old breaks. It’s just inexperience basically. Back to practice I guess :drummer:

Heres a quick thing I made about a week or two back. Just some random patterns using mostly single hits chopped out of full breaks, some larger sliced phrases from breaks, and a 303 plugin render down to an instrument. I didn’t finish it or turn it into song, and it’s very kinda janky and unhinged because it’s mainly just noodling around and practicing using lots of different samples together in one track. Couldn’t upload the .xrns file cause it exceeds the upload limit sorry

22

A trick when you need to fill a gap is to reverse the sample, add some silence in the end and apply some reverb to it. Now turn it back around and let the reverse reverb fill the gap. Works sometimes, can’t do it with sliced samples though.

23

Hmm I tried it with a vocal sample but it didn’t work? The reverb just behaved as normal?

1. Load the sample

2. Right click sample waveform – process – reverse

3. Sample effects chain – add reverb

4. Right click sample again – process – reverse

Is this right?

Scratch that figured out how to do it using the render to sample button in the pattern editor.

24

My work flow was basically this:

A different approach for the same results :wink: I got stuck with doing this in Ableton just because I came to Renoise from using Live for about 7 years or so for all my timestretching and generally fucking-things-up needs (main DAW was and is and will be Cubase though, I love it hopelessly :smiley: ).
Multiband compression and expansion are for me hit-and-miss a lot of times tbh, unless I’m working with a break that has not been processed at all, I will probably not even touch it if there’s no glaring issues to be fixed. As a guideline for MB compression I’ve been using this: (picked it up ages ago from DOA):

Example Settings for a Multi-Band Compressor (5-Band)

LOW - tighten up bottom end.
Frequency Range: 0Hz-150Hz
Ratio: 2.5:1
Attack: 20ms
Release: 150ms
Threshold: very low to almost always trigger compression.
Gain: make up gain lost in compression.

LOW MID - tighten up the mix.
Frequency Range: 150Hz-600Hz
Ratio: 3:1
Attack: 20ms
Release: 150ms
Threshold: trigger regularly, but be about 2dB below the point of
rarely triggering.
Gain: make up for compression, or just a little more for warmth.

MID - add punch to the mix.
Frequency Range: 600Hz-1.5Hz
Ratio: 6:1
Attack: 10ms
Release: 150ms
Threshold: set fairly low to almost always trigger compression.
Gain: add 4-6dB or more to make up lost gain and add guts.

MID HI - add presence and increased clarity of individual instruments.
Frequency Range: 1.5KHz-6Hz
Ratio: 3:1
Attack: 10ms
Release: 150ms
Threshold: trigger regularly, but be about 2dB below the point of
rarely triggering.
Gain: add 1-3dB for presence/clarity.

HI - reduce harshness without losing sparkle
Frequency Range: 6KHz-15Hz
Ratio: 2:1
Attack: 10ms
Release: 150ms
Threshold: only trigger when harshness present.
Gain: maybe add 1-2dB to recover sparkle lost in compression.

I like those beats a lot mate, could you maybe upload the .xrns to something like Dropbox?