Hello children.

I am currently dissecting some trance music as a learning tool.
In particular I am listening to tracks from the latest Shpongle album, which I really like, and wondering how some stuff was made. I am fairly new to music production so the questions could sound basic.

I use Logic and Live. I know that Simon uses Logic too…I wish I could watch him while working. So, to begin:

- I am confused about the absence of a true step sequencer in Logic, which from what I’ve understood is essential to produce those trance rythms. After some research, I’ve discovered that I could build a step sequencer in Logic’s environment. Is it worth it, though? I’ve seen this nice step sequencer in Reason’s Thor and love its ability to tweak the pitches with the knobs. That seems to be the thing I need! But in Logic, or Live. Take the lead at 1:54 in this track (I have not posted the track on youtube so don’t bust me, bust him if that’s illegal):

http://youtube.com/watch?v=ezVEfZ66-70&feature=related

Summing it up: how would I produce this rythmic lead with the main tools in Logic. I hear this rythm is gated, and this is no problem. I’m not sure what else has been automated, pitches, filter cutoffs, or other stuff. Also I’m not sure what kind of instrument would make that sound. Would I go with the ES2, which has the vector envelope which is a sort of step sequencer in the end? But what if I want to use a sampled instrument?
I don’t find using the piano roll convenient to create this sort of pattern. What I’d like is something like Reason’s Thor! Too bad I don’t like the sounds produced by that synth much, I reckon the ES2 is fatter. But that’s me!

I use the Alchemy softsynth too, which has a step sequencer. I can automate anything there per step, and use samples. But I still can’t get anything close to the lead on that track. Any suggestion? Again this is for practice only.

I was hoping to do that with Ultrabeat, which has an automatable step sequencer. There, I could use a sample and automate its pitch and cutoff, and I’ve done some groovy basslines in that way, and a couple of leads too. Still, nothing remotely close to that sound. I don’t understand why it doesn’t make me modulate pitches via sidechain - I must use a sample to do that. I did a couple of things I like but still not 100% satisfied.

As said the ES2 has the vector envelope but I can’t get anything properly melodic by modulating the wave’s pitches (via the XY controls, which are anything but precise) step by step, nor the cutoff. To make things straightly clear, I simply would like to know what Simon used to make that rythm. If it’s a sequencer built in the environment or the ES2 or what else…Listening to the pattern again, though, I somehow doubt there’s any “step sequencing” in it…can we say it’s simply a gated pad? If so, I need info on how to step-sequence in Logic anyway.

If I wanted to use Live I suppose I would go by modulating pitches, and other stuff, step by step via the clip editor. That would be much simpler to do there than in Logic on any instrument, I think. I still have to try that but to be honest I really want to use sounds made with the ES2 or the sampler in Logic. They’re fat.

- Second one…

http://youtube.com/watch?v=1Shi9tR3GhM

The whole amazing drum insanity in the first minute or so of the track. How does one start building anything like that?
My ideas…
- Mixing four drum loops up with a sample based synth (Alchemy), morphing between them and modulating stuff…
- Slicing the loops up and using NNXT in Reason or drum rack in Live.
Any other idea?

That’s all I need to know for now.

Thanks in advance!

Hi Ducko,

I can’t say much about Logic (I play FL Studio), you’ll get most useful tips in a/the Logic Forum.

1) Do you mean the lead that kicks in at 0:25 ? Sounds like not only the pitch of the note is tweaked, but the cutoff frequecy of filter is singing along. Maybe it’s even just the filter that is sliding up 1 octave (12 semitones or 1200 cent) from a low G to a high g. Look for a good Filter plugin and spend some time with it.

But better than this technical advice is: just keep playing around trying out new things. When I started out 3 years ago I found that in the beginning it’s better to focus on one thing at a time, before putting a whole song/track together. Then you can start focusing on the music and expression again.

2)
For a basic beat or break, it’s worth it to slice up waves per hand, or to use a tool like Livecut (Smartelectronix) or Slicex (temporary at $79, by Image Line).

Then to spice it up and make it go really crazy, look for some multiFX plugins like

dBlue Glitch (free!!! but donation for that struggling programmer would be cool),

Livecut from MDSp Smartelectronix (free/donatate)

dFX Buffer Override (free/donatate)

Effectrix (99

Thanks for your reply,

1) No, I actually mean the theme that begins at 1:54 exactly! It’s a completely different thing…and it’s still sort of mysterious to me to figure out what kind of an instrument it is, and how it is done. Also, I am experimenting with different methods to step-by-step modulate anything in Logic, and until now I am finding this fabolous Hyper Edit method: a hidden secret, but worth setting it up. This allows me to modulate any parameter rythmically, exactly like in Reason’s Thor…just not with knobs. Maybe it’s even better to me!
The day I’

1) aaahhh… ok, that’s a vocal sample (singing ‘I Am You’) going through a gate and a filter or phaser or both. A gate turns the output on and off in a rhythmic (trancy) pattern. Cutting up a sample by hand works too, I still do that sometimes for more breakbeat kind of patterns.

2) What those kind of multieffects do, is to chain a bunch of simpler FX (phasers, flangers, gates, filters, delays, reverb, etc.) in different ways. Or like in the example above, to send each bit of a sliced up loop to different mixer channels which each have different Fx on it.

Dub pioneers like Lee Scratch Perry started this by sending one snare hit to a fat reverb, and then the next one, say, to a delay.

It’s certainly good to learn how to set such complex FX chains up and how to slice up loops by hand, but if you like to make lots of breaks some sort of looping device is very handy.

Flangers are sort of multiFX themselves, that emulate doing wierd things with tape recorders, they save a lot of work.

Maybe the NNXT has some break-looping capabilities?

1) a vocal sample!? Dammit, you’re right! Can you believe I couldn’t hear it from “the right distance”...I am only now hearing the words “I Am You”. Now everything changes…

2)  I think I get the concept of slicing and applying effects to slices, and I am pretty sure the NNXT does this sort of things. In theory, one could do it by hand in the sequencer, right?
Can you point me on any tutorial or youtube video on this sort of process? Even in FL, that’s fine. Just to make sure I understand this process of glitch music…
Is Shpongle using just one drum loop here? Can that single drum loop generate so much variation? They sound like many loops playing together, not just one.

—> http://findremix.com/tutorial-beat-slicing-in-logic-pro/

...it’s more like a fractal of loops. You can take recordings of real drums, percussions, guitars, little noises (glitches) (e.g. I like to take bits of scratching the strings of my guitar) ... or patterns and breaks from a sequencer, or some simple loops.

Then you apply some FX and render it (bounce it) and then you take that wave file and use it as a new loop, then you apply some FX and rend….

The art is to make it as complex as possible without getting to messy. Maybe using a noise gate to clean it up again, and EQ-ing with love…

Thank you my friend!

Let’s see what comes out.

Ok, one more question hoping not to abuse you…

http://www.youtube.com/watch?v=LLo8YRkBi2w&feature=youtu.be

Can you dissect this track and tell me what the main elements of it are? What I don’t understand is how RYTHM in it is manipulated. Where does one loop begin, and one finish? Especially the drums.

Thanks again!

No Problem, this world needs more twisted music/ians. I know how mindboggled I was when I first tried to find out ‘hows-it-done’, and I’m still learning how to make real music with it.

Twisted mix of Thom Yorke indeed =)  First a ‘dubstep bass’, e.g. Filter with an LFO that reacts to different pitches with different speeds was added.

And then the whole song/mix went through a loop-scratcher plugin.

Historicly this goes back to scratching vinyls (Herbie Hancock for example).
The digital idea is to draw an envelope for the speed/time

This plugin does it with samples

http://www.youtube.com/watch?v=aja8PKgBN30

.. and this one in real time:

http://www.youtube.com/watch?v=NYuiCVmadig&feature=related

(PS: There are other great plugs, e.g. dBlueGlitch (free!!!), I just happen to know and work with the above mentioned)

There also (expensive) complete software emulations of DJ pults with 2 turntables, but they are more for playing live DJ sets. 

Bottom line: start playing with dBlueGlitch (i.e. the relooper part of it) =D

PPS: With your Alchemy Synth you can make very organic sounds with grainular synthesis. (Like at the beginning of the track Esc-a-ton)

Thanks as always man, you’re a good mentor.
For starters your Citrinitas track would be beautiful, would it take too much effort to re-master it? I think it’s worth it.

Dubstep bass, got it! DBlue Glitch looks beautiful, but it’s for Windows only, while I’m on Logic on Mac…but don’t worry about it, I’ll find the equivalent (the Reaktor suite would be one).

Granular synthesis: haven’t experimented with it yet. It’s the one with the grains…what was your sample of start in your Esc-a-ton track? I presume it begins from a sample. In Alchemy you paint grains over the sample like you do on MetaSynth…to be honest I haven’t found a use for it yet! Only Aphex Twin sounding stuff (and Aphex Twin is awesome, yes, but we’re in 2010 and we need to move ahead!). Did you paint grains or is some other kind of granular manipulation involved?

Audio rendering question. Here in the forum I heard Ott in person saying he works in Reason and then renders to audio for work in the Nuendo (fat Cubase) sequencer…
But isn’t this a terrible terrible thing? haha…I explain…
By rendering to audio Ott will not be able to tweak anything at midi level anymore, and why is this not a pity?
In my little experience, I end up tweaking stuff at midi level even at advanced mixing stage. Is it because I’m stupid?

Say I decide to add a part to my mix using the same chords of a midi track I did in Reason and now it’s just audio in the sequencer…I would have to go back to Reason and produce that part. That’s ok. But isn’t it simply convenient to keep the midi as midi? I can tweak the raw sounds, edit the midi notes, I can tweak the cutoff filters and the tunings and the waves used in my synths, I can tweak the modulations…and when I render to audio I can’t do that anymore! I would suffer in rendering to audio unless extremely necessary!

When I hear electronic music I hear deep modulations being used throughout the track, which makes me think the performer preserved the midi format in most cases. In the case of Ott he uses a lot of audio loops I guess but definitely he doesn’t render his ever-evolving pads in to audio from Reason - I suppose he added them in his sequencer.
I’m sorry if this sounds stupid! I hope you get my point.

Maybe it’s a question of personal workflow and style, or maybe there’s something I still need to get. In facts, I have worked more with synths until now than samples.

In the case of drums, with all those short samples used since the beginning, I am ok in rendering and re-rendering to audio. But in case of a synth melody, lead of bass, I would like to keep it as midi because maybe in the final parts of the arrangement I want to automate the filter resonance or an LFO rate of the synth to change the basic sound, not the sound of the entire audio loop as in the case I use a filter OVER it. And what if I need to duplicate a midi part and play it with another intrument? I need the midi for that.

Does this make sense at all? I hope so and I hope you have an answer to this terrible question. I thought Ott would be jealous of his midi and would have liked to keep it. But maybe that’s because his sound is sample-based. Maybe when your main lead sounds and basses are synthesized keeping the midi makes sense because as far as I know you’d have more control over the singular sounds - unless I don’t know something special about audio manipulation.

It starts with a sample of Terence McKenna saying ‘Escathon’. I was inspired by Shpongles ‘My Head Feels Like A Frisbee’ and glad when I found out what creates such rubberband kind of pitchbends. Another way is to combine a Micro-Delay with tweaked Delay-time with a flanger.

Painting Grains is a very new approach to grainular. In ‘the old days’ you only had knobs for grain size, space between the grains and time delay. I almost bought Alchemy for that reason.

The context of Ott rendering waves was ‘new album complete’, and he’s just preparing the stems for the mastering engineer.

When creating something like in the example ‘Shpongolese Spoken Here’ (the glitch-paradise-percussion part) it does make sense to render inbetween, because such wild FX-combinations will eventually grind even the most state of the art CPU to hiss a white flag…

Then you delete some FX and add new ones.