pouët.net

Bit of discussion topic for the musicians...

I always seem to hit a roadblock when my songs are about 2:30-4 minutes long. Usually I have all kinds of great ideas at the start of a track but then I run out of steam; maybe I just throw everything in too early? I've pushed pieces longer a few times but I always feel like I'm just repeating patterns and not offering anything new.

Longer, drawn out builds & breakdowns are definitely something I want to do better. My usual styles (trance, dnb, house, techno) have LOTS of room for repetitive themes but you have to be careful to BUILD on them or the track just gets boring. The structure is pointless if it's not supporting movement.

How do you guys make longer tracks that still manage to flow & hold the listener's interest later on?

at 75% do the tune-up! ;)

then again, for demo sake, 2:30-4 minutes are spot on to be able to have ~10 parts that are interesting enough.

Paulstretch.

don't care about unknown listener, care about making a track you would listen to again multiple times. if you think the length is right, then it's right. if it's getting boring then move on to tell the next part of the story. if there is no more story to tell then it's over.

it's usually better to tease or hint at patterns first and have them fill out later if you're going for a slow and brooding buildup in dance, but what ps said: if it feels right/wrong ... it is, even to the first listeners so polish it

What I really liked when I was stuck arranging or starting songs was Oblique Strategies (online version), a system of cues that can set you off somewhere if you're stuck, it's quite abstract but it's fun

It is as easy as:
Put your first 2-4 patterns in again at the end, reuse samples to make another song after that!

listen to "real" tunes in your usual styles and learn from the ones that you don't find boring. why are they not boring? what's the structure of the song like? what kind of sections are there and how long are they? in similar sections, which elements are there at the beginning of the tune and which elements get brought in during the tune? what are the dynamics like, what is quiet and what is loud and when? how are the dynamics achieved in terms of the characteristics of the sounds playing? etc etc.

ymmv, but i think imitation (even subconscious imitation) is the key to success. up to a point, of course.

What reed said. The devil really is in the detail. It could be something really subtle like the addition of shaker, little edits and so on. Stuff to keep the listener on their toes and engaged.

So the answer is: needs more cowbell!

Sometimes, yes!

everything is better with more cowbell.

My tracks are usually around 8-15 minutes. Because i don't arrange them by moving around patterns or blocks. I build up the foundation of clips and tracks, hit record and jam the thing. Wild Pitch style. The result always represents my mutual flow according to what i hear.

Why? I find arranging tracks the "normal" too theoretical. Every piece of music has its own DNA and depending on what you're working time becomes a relative measure. However, when starting to arrange from scratch we look at those absolute numbers. Minutes, seconds. And start to rush the shit down, totally neglecting the flow.

That doesn't work for me. I need to feel the track from beginning to the end. Especially with the techno & dubtechno tracks i produce automation of details is what keeps those interesting. And i could never draw them into the DAW. I need to record everything in one go to have a layout. Maybe i edit a few things then, or overdub the automation with more things.

That's also how i worked with i.e. Calcifer. Captured the flow of the material once and moved on from there. Then it's all about editing.

most of my tracks i ended up on a similar process to the one rp described, a few sessions to add elements to a pool, just fiddle around jamming to find a sequence and flow that makes sense (sometimes noting down the sequence in a piece of paper or text editor), then take a time out and redo it while recording the process. it's also a nice way to ensure you have something you can use on a liveset. can be a bitch for fixing small quirks later on though.

in my experience this process usually gives you tracks that are a longer then the traditional tracked / sequenced approach.

i fully agree that everything is better with more cowbell. and a small reverb.

As LB once said, writing demo music is writing radio pop - you end up around 3 minutes because you get used to the idea that you don't need a longer intro, you don't have enough content, etc etc.

Once you start listening to more "non-pop" electronic music, you'll get used to the longer track lengths.

make several short tunes and stitch them together? what are you trying to achieve really. why.

When I write music I find it hard to write from start to finish. The better songs are these where I have an intro AND an end (between 1 or 2 minutes). Sometimes there is another part I drop in the center. Then I start to compose the transitions and these often make the song long.

Lately I saw a tv report about music composition which basicaly said that the listener's ears were "deaf" when a sound or a sequence was repeated. In other terms the sounds become "blank" to the listener's. Thus if you bring a slight change to the sequence it becomes different and it draws attention from the listener. This technique repeated to different sequences or just a hihat or cowbel sequence offers a lot of posibilities to the composer not to bore his audience. As sais these are basics but combined to other tricks it can lead to great things...

what's the thing with long tunes? does it make you a more legitimate musician?

It prolongs your listening enjoyment.
That said, if you can't make it longer, don't try to force it to be longer.

The main purpose of longer is for djing, this is why dance tunes have always rolled in at 5-10 minutes. Long the long intros and outtros give you ample space to blend he tunes together and the good amount of time in the middle gives you a chance to pick out what your gonna play next. Almost Every time I play demoscene soundtrack in a set I usually have to loop up the intros and outros so I can get a decent blend. Frankly it's anoyin! !!!

Interesting! Can you give an example, which song?

prime example is space cut, you can't actually play the intro as it's at a different speed to the main part of the track, so I repeat the building vox sample and stutter it to give it a bit of rhythm. The last section stays at full pelt until everything drops out so that usually needs looping up and filtering out while the next track builds up into the mix.

the magical 16 bar are your friends . :)
I always have some trouble when someone needs a tune @2:30 ... 2.30 ! that's just an intro with a drop! ONE dirty cutted DROP! -.- and lb is quite right. the coders have a lot of work to fill your borring bars with some flybystuff :P
if you do not get along with 16 bars, try it with paulstretch and a cowbell. that works!

@rpfr : you rebel!
@yzi: djs like your tracks and play them (sometimes), because they are easier to mix from a to b over a longer range (16 bars). but a good track even feels like 6 minutes. ;)

example:
https://www.youtube.com/watch?v=KIGXGkzjAoc
... pinkelplatte

I guess I'm the complete opposite of rp with regard to work flow: First thing I do before an arrangement is create a blank track and plan out a story with empty midi clips, colored and named to represent energy levels, orchestration, song parts, or particular set pieces. At this point what I usually have ready to deploy into that ghost arrangement is a number of highly developed but isolated parts (loops), which I then use as building block to fill in my track outline, creating transitions and interactions between the parts along the way, but completely asynchronously. I worry that if I were to record arrangements as live jams it would end up being too boring, or meandering, or paced wrongly.