What comes first?

[VanillaCamilla]

my take on it is to just let the ideas flow, dont get slowed down by small, complex details, just get the track together with the basic idea and feeling. after a few hours of just doing what feels natural the track starts to come into being, yadamean?

[subscan]

for me what comes first is a folder - i then name it whatever it is that project might be - then i start creating samples specific to that project only ( ill go make samples on some early morning walks, just thinking bout the title and what im after - ill think of the setting of my song and try and use instruments i imagine would be playing

Excellent idea!!!

[Robot Criminal]

finding the concept. I always fail badly at that stage. Then I get obsessed with the sounds. Then I obsess some more about what to tweak and where. Some more failure. Ok, now we have something decent goin on lets make it more interesting. Damn.
I've lost track.

[Torque]

I'm guessing by reading some of this that it seems that allot of you don't come from a musical background. I can't honestly say that i know what that's like because i'v been playing with instruments for a long time before i ever decided to make music this way. What i have found though is that allot of people without musical backgrounds end up getting hung up in the composition and arrangement stage of making music. These are much more important than the mixdown stage IMO. My advice to anybody that gets hung up on these parts is to learn to play an instrument. Of course the easiest one to gain access to for all of you should be a keyboard. I recommend getting some sort of program to help you learn to play. For me it was guitar that got me ready and familiar with musical structure. Once you take the time to do this and get familiar with playing other peoples music you will start to understand a little better. Music does have a structure and it's almost like a mathematical code. All this new software and technology is awesome and has some great capabilities but it's impossible to use it to it's full potential without a good grasp on musical structure. I know that it's possible to become very adept at using software and all these machines but IMO it will take you longer to improve without learning to play, as hard as it sounds it's actually a shortcut. Just don't give up if you can't play Bach for the first couple of months, it'll come to you in time and while you learn you will gain a grasp on some other things that will help you.

A good composition is hard to make and takes allot of practice as well as some talent and not everybody gets it right away. If you compose something well the arrangement will almost happen on it's own. I can't really explain why this happens but when you come across a good one you'll see what i mean.

Don't be afraid to sit on a track. I have some that i'v been working on for at least 6 years. Some of them i kept around because i had a couple of good parts that i didn't know exactly how to utilize yet so i decided to let it sit on the shelf for a while till i knew what to do with it. There are some which i knew were good all the way around except i didn't have the ear yet to mix them down correctly and i knew it so i let them chill for a bit till my skills caught up. I even kept my very first composition around and last year decided to do it over using some of the new skills i'v come across. One thing i learned about being around the UR crew is to not put something out until you know you've done it complete justice. I'm not talking about just doing it as good as you can now with the skills you have, i'm talking about doing it better than anybody else could do it and in a way nobody else could ever do it. It's hard to develop patience like that but the payoff is spectacular.

when it comes in the order i do things:

1) figure out what message and picture i'm trying to portray
2)make the parts and the sounds at the same time
3)start to arrange the track, do a rough mix and burn it to cd to listen to in the car
4)come back to it after listening to it a bunch and figure what to cut out of it
5)Repeat 3 and 4 until i'm cool with it
6)Mixdown

[Crimson Chamber Music]

excellent post Torque!

it is good to hear that you keep your tunes for such a long time,
I am always thinking I am an idiot for spending so much time with my stuff, but hearing this I think I am good :-)
I totally agree with the part about playing an instrument, although I would consider a groovebox an instrument by now ,-)
And who needs Bach anyway ;-)

come to think more about the 6 steps you put down, I realise that
my best tunes have been made in adherence to your 6 steps, with special regard to the first one:

figure out what message and picture i'm trying to portray


yes people, Concept :-)

[tone-def]

My approach is similar to mlexicons. I start by collecting and making sounds and putting them in a folder. once i've made this little sample library i start making loops, which are normally 16 bars long. i make about 2 or 3 of these loops which i try to make similar rhythmically (i roughly mix as i go). These 2 or 3 loops then become different sections of the track. if the loops flow into each other nicely i've probably on my way to a decent track. i then repeat the loops and every however many bar phrases i'll change stuff like take the hi hat of whatever sample out or add something else. if you take something out thats driving the track your going to have to add something else with the same or similar rhythm to replace it. that normally takes care of the structure. i then do some fiddly things to make certain parts of the track more interesting like fills, automation on, adding the odd random sample etc... i then mix it down and leave it for a few days listen to it somewhere else, change some stuff. do it again until i think it's finished. hope this makes sense.

i really enjoy remixes and i get them done very quick because i've got a folder of samples that work well together. It was when i did a couple of remixes i found having a folder of samples works for me. probably not the best approach if your a synth guy.

[clubfoot]

this thread addresses probably the biggest hurdle in computer-based production IMO.

Lots of good tips and approaches mentioned above.

My most important points on this subject are:

*a strong collection of sounds and samples is important to start with - these sounds are what the track is all about.

*Don't spend too long working on loops - very hard to finish/further a track this way! Spread your loop out as soon as possible into some kind of basic arrangement, then compose the song on top of this or further tweak the loop as required. I actually force myself to just repeat my base loop for 6 minutes so that I can say I have a song and then work from there. Without doing that, it is far too easy to just noodle around with the same 4-8 bars or whatever for hours and hours.

[ArhiteK]

I start first with drums, bass, some perc and a "theme sound".

Then I come to my crossroads.

If I choose composing/arrangement path, song usualy starts to evolve in some kind of interesting and logical structure.
I spread the structure through time and then I have it laying in front of me. Then I try to layer some sounds, efx, pads, atmos etc. Slowly making it denser and fuller.

But often (way too often) I start refining and evolving my source loop. New sounds, new passages, efx. I virtually build some structure, but structure stuck in time of this one loop. And usually, I loose it here. That becomes "the best" loop in the world, but without any flow in me which can help me build the progression of the song. Too much detail on every sound makes them "incompatible", too stiff and dense for arrangement.

And in some way, my inspiration and idea wear off somehow while growing this loop. This loop breeding process leaves me full of sound, but totaly clueless about, hmmm, message of current track.

Since Logic Pro arrived (2 weeks ago), structure of program pushes me into arrangement/composition mode way earlier.