Composition/Production Proccess

[GLJdrew]

When you are creating songs, do you usually take it one track at a time or just do as you hear? I've been geting stuck on some songs and I think it might be helpful to just focus on finishing the actual track before adding spice. So say I have a piano riff I really like, would most of you recommend just finishing the general outline of what the piano will do for the song? I see a lot of pros and cons for this. On one hand it keeps me moving forward, but on the other hand I might forget some ideas that come spur of the moment-which are sometimes the best ideas. What's working for you guys because i'm getting stuck a lot!

Thanks for any advice

[Samaritan Sound]

At a big level, I separate my creative sessions from my technical sessions. For example, a technical session might include gathering all the loops, clips, and samples I plan on using in a song, setting warp markers, creating audio tracks for recorded instruments, etc. That way during a creative session, I don't get bogged down by these sorts of things.
Now, to your question, when I'm first sketching out a song, I only use presets (Ableton and Reason), and don't do any mixing/processing. Once I've got my basic ideas down, then I start playing around with the different instruments. After that is when I start adding any additional bits and playing with special FX. After all that's in there, then I mix the track.

[invol]

At a big level, I separate my creative sessions from my technical sessions. For example, a technical session might include gathering all the loops, clips, and samples I plan on using in a song, setting warp markers, creating audio tracks for recorded instruments, etc. That way during a creative session, I don't get bogged down by these sorts of things.
Now, to your question, when I'm first sketching out a song, I only use presets (Ableton and Reason), and don't do any mixing/processing. Once I've got my basic ideas down, then I start playing around with the different instruments. After that is when I start adding any additional bits and playing with special FX. After all that's in there, then I mix the track.

I basically approach things the same way. I separate my sessions into different workflows - 1) Templates, File Management, etc, 2) Sound Design / Preset Creation 3) Creative Session (though I often do some basic processing and mixing as I work), 4) Final Mix and Mastering.

If I get stuck on a composition, I will leave it for a bit, work on something else, and think about what I want to do with it. Once I have an idea of where to go I get back on it and finish it.

Cheers,
Brian

[Angstrom]

my process is

1: block out the main bit of the song with quite bad playing, poorly programmed clips, and a hastily chosen kit
2: realise I only have 2 mins 30 of song
3: replay and repair those bad parts from step 1 ... but get distracted/inspired into writing new parts
4: song is now 8 minutes long, with several unrelated sections some of which are terrible.
5: remove some bits, fix some instrumentation, try and make a coherent arrangement from all this stuff
6: add some last drum variations and fills, finally fix those bum notes and timing issues which somehow escaped the first 5 steps.
7: mix it, about 6 times

I'm sure that doesn't help at all.

[Samaritan Sound]

One thing I need to start doing is when tossing bits out, dropping them into a folder. Who knows, they may eventually find their rightful place, lol.

[Vios]

I follow these 5 steps RELIGIOUSLY, once I started doing this I cut the amount of time needed to get from scratch to a final track in half:

5 Steps to Writing Better Music Faster

Previously I had serious problems with doing tons of production work when I hadn't fully composed the track yet! Now I'm able to leave things imperfect, knowing that I'll come back later and fix all of the production problems.

[petit nuage]

the video is interesting...


but i would like to have an advice :


when im in the sound design phase... in order to design by exple a kick ..must i use eqs, comps and even short delay or short reverb , to realize the sound i have on my head ? or must i do that in the mixdown session ?

- some engineers prefer to premix / make decisions on the sound before the recording to dont have to make that after -
like equing, compressing even pan or very subtle ambiance reverb to give a dimension to the sound .


and


when im making sound design, is referenced to the main and the leading sound/intrument in the future track by exple a kick ..
so i design my sounds based on the kick : so equalized, compressed etc.. in order to they complement well the kick ...


so is this a bad habit or ?


...or must i just to have an orginal tempo and making sound on the fly /from scratch based on anything : just the perfect sound that i have in my head ..and then using fx (even pans, reverb,chorus delay et) to realised/ design it , before recording it ?


and after ,during the mixdown session , i can put eqs+ comps etc to fit the sound created in the mix, correlated to the others elements ...

by exple : in the sound design session : if i want a big reverated sound must i design it like that ,with reverb, and record it or must i design it and put the reverb later on the mixdown phase ?


i would like to have some advices on it

THANKS

[Angstrom]

the video is interesting...
but i would like to have an advice :
when im in the sound design phase... in order to design by exple a kick ..must i use eqs, comps and even short delay or short reverb , to realize the sound i have on my head ? or must i do that in the mixdown session ?

Sound design quite often involves compression, EQ, delays and more, but that's still a different part of the process than mixing.
Think of it like a guitarist going through a certain chain of effects, amp and speaker combo. That's the sound design phase.

A lot of people flatten their sound design to a wave, to make a firm mental distinction between the design stage and the mix stage, although I don't do that because I occasionally find it more useful to pop back into the sound itself to fix something for the mix. I try not to break what was originally good about the sound though, while making it fit better.

[deanc2000]

Here are my steps: 1) Choose a chord progression - To make the process simple, I have been using the same 8 bar progression thru out the song. I only introduce a different progression if I use a bridge.

2) Establish melodies for each section of your song, ie Verse, Pre-chor, Chorus, Bridge. I usually do this by recording myself humming, or singing melodies to the chord progression, and recording takes to a basic drum loop. Ableton does not have a take facility, so I use another DAW. I chose the melodies that stand out from those options.

3) Write lyrics to the melodies: I start with the Chorus as it's the most, if not one of the most, important part of the song. I then go to the Verse, which is the hardest as it involves the most writing, then the other parts of the song.

4) Once you've got the main melodies and lyrics to your song, YOU ARE 80% DONE. The above process can take me days to weeks. I am very particular about lyrics. I then write in the instrumentation, which can be as simple or complex as you want it.

5) Then arrange it, add intro, then lastly mix it.

For what it's worth, I've been very productive working in that way and have written a few good sounding songs.

D.

[petit nuage]

thanks angstrom !



yes i was thinking about your guitarist exple and its very interesting ...
a bit like when you record a sound in town or nature.
this sound already contains in its "core" some volume variations, pan, reverb, filtering by elements around it ...



but when the guitar sound is recorded, its the guitar sound the guitarist find at home and which is good for his ears but without any others instruments .
if this sound dont fit with the bass sound and drums during recording sessions ..must the engineer tweak it a little bit with eqs etc..to make it sit with the bass and drums before recording ?



imo, sound designing is maybe creating a sound with no other elements around than premixing include that sound will be part of the overall sound of a track ...so maybe sound designing a lots sounds and premixing them(eq,comp,early reflections...) in order to they fit well with each others .


so maybe sound designing first, then premixing and after the final mix ?
for me its the problem to have a reference point ..is the main element for the track : the voice ? the bass ? the kick etc..and make my premixes based on this MAIN sound, and recording them modified ...


i believe i ve read somewhere that eddie kramer (sound/mix engineer for hendrix led zep etc) recorded hendrix with all eq, pan, comp, early reflections...etc included, to dont have to do it later .
yes, the editing was on tape at this time so people have to make decisions before recording more than nowadays i think...


maybe the creative sound designing /premixing phase, shouldnt interferate while im mixing ? (making my right and left brain works at the same time may be confusing, loss of brain clarity and counterproductive? )
its a very interesting and vast question....


@deanc2000 interesting method

plus, you are right about singing all the differents parts, even beats etc...
imo, its a natural and instinctive way to brush the main lines of the canvas and to have a reference later when you will build the song/track and you dont know which way to go.
this recording can become a good framework to work into, without creating or experimenting too much (leaving/forgetting the main idea) and then wasting time .

[Angstrom]

...
but when the guitar sound is recorded, its the guitar sound the guitarist find at home and which is good for his ears but without any others instruments .
if this sound dont fit with the bass sound and drums during recording sessions ..must the engineer tweak it a little bit with eqs etc..to make it sit with the bass and drums before recording ?
...

That's very much down to each guitarist, their experience level ( and egotism ).
It's the same way that inexperienced electronic musicians will choose a giant pad preset that occupies the entire audio spectrum and then try and cram that into a song. Most guitarists with a few years experience and some sense will find a tone that lets them be the most audible in the band, which we would know as a kind of mixing. So there's a process of lonely riff creation, then of picking the right settings that fit with the band.

Where this fits with what we might know as mixing ? I know that Geoff Emerick, when recording the Beatles used to go and personally change the settings on the guitarists' amps himself, to get them to sit better with each other. (book link). However, that was the 60's.

There's a world of guitar recording and methods that producers and engineers use, everything from "a guitarists tone is sacrosanct" to a more pop sensibility of "after the guitarist left I ran the DI'd guitar track through an amp sim" . I don't think that's a good idea at all, because the spirit is lost. What any musician plays tends to be very dependent on the specific tone they can hear as they are playing it - the two things are intricately related.

and now, for no good reason at all - Tony Iommi, http://www.youtube.com/watch?v=GvAiPszutv4

[petit nuage]

héhé thats very interesting and soo true !


yes it depends on the style of music and the arrangement it leads.
so the recording and mixing have made in consequence.

ive read that a lot of sound engineers do that : tweaking drums, guitar and bass 's amps, choose to good place to install the drums according to the room etc...


so its like they first optimize the place the instrument sounds the best, tweak the instrument to obtain the sound which is "relatively" good and serves the song/track ( but sometimes not the musican's ego ), install mics to capture, according to their point of view, the sound the instrument should have in the future overall image/mix ..and maybe just for one song only or a moment of the song ...


its like a kind of general premixing before recording ...with the perspective of the final image/mix, already in the head ..
maybe its important to ask the good questions before premixing and mixing-down.

- what is the main element, hook ...?
- is this the right/best sound for this melody etc... ?

[simmerdown]

i hope i can never answer this definitively

[Tysonviolin]

Im always changing my workflow to keep it fresh. I go through phases and switch it up. Right now what's working is going somewhere crowded (coffeshop bar or park) and fuck around for 4 or 5 hours with ideas. Next day I am usually stoked on some of these ideas and I chisel them a little more.... After this or any other process, I need to lay it down in front of an audience a few times. This helps me dial in the tune... From here, it morphs continually until I record and release it, or drop it.

[Vios]

petit nuage,

How much production I do during the 'sound design' stage comes down to what is "essential" for an instrument. For example if I have a piano part that I know I'm going to make really really reverby and I know that's part of the sound I want for the track, I'll put a reverb in my return and maybe spend 30 seconds max tweaking it to my piano part. I know that later I'll go back and change it until it fits just right once I'm in the 'production' stage of my track. I know I'm going to have to change my reverb to fit with the rest of the track later anyways, so why should I spend more time on it right now than needed?

The same idea esp goes for dubstep. If I have a dubstep bassline I'll spend some time with it applying huge amounts of distortion/ auto filter etc. to get it's sound right.

My thought is if an effect fundamentally changes the sound of an instrument, do it during your initial 'sound design' phase. If it's something small like 2dbs of compression or EQing, do during the 'production' phase.

Also I like these words that I've heard from a music producer, and I hope are relevant:

You can only have 3 things in your track stand out at any given moment. In electronic music they're typically: Drums, Bassline, and Main Synth or Main Vocal. Make those those 3 parts sound great Together. Everything else should be produced to stay out of the way of your main 3, and don't need to sound good individually unless they are themselves a "main 3" for part of the track. I wouldn't call this a "rule", but is a good guideline.