I have always had difficulty finalising projects and editing in a DAW has never been a comfortable (or particularly creative) experience for me. I am torn between recording everything tightly synced (which gives me the opportunity to loop sections to a grid and have multitrack tempo relationships repeat easily) and recording something much freer for a period of time (which I’ll then notice has something I hate happening and I’ll find it almost impossible to edit out (can’t just cut a chunk out and butt it up against the other region for instance). I have tried warping tracks in ableton but a) I’m still not very good at it, b) the stuff I’m making has few transients at the moment, and most importantly c) manually trying to get things to loop nicely by ear totally kills any good vibes for me when making music (again, probably because I’m not much good at it).
Sorry for the long post but this is a real productivity killer for me, I have hundreds of unquantized ‘great starts’ in a folder and I find it difficult to know where to go next with them.
So, whether from the start of making a sound or during some involved editing, does anyone care to share their approach, workflow, philosophy, or insight into making ‘finished things’ specifically in relation to timing?
Thanks
Well put.
I do think that as in with writing or any creative project, it takes some grunt work to come to a well-rounded finished project. It’s not all fun and wonder all the time.
We had a thread or two on this last year
This is my favorite topic 
… looking forward to more discussion!
Most of the music I make could effectively be described as working with unquantized chunks of audio. I suppose I have a few strategies to do so, but I really like the challenge of figuring out what I’m able to do with the audio, so a lot of mixing and editing is all part of the process.
For recording fodder, some general rules I try to stick to:
- Record single parts in mono as much as possible.
- Record longer than I need, try out multiple takes, variations, extra bits, etc
- Re-perform/record the part if I want it but can’t edit out a mistake or part I don’t like
I also compose in one bit of software, do major edits in another, and mix in another (unless it’s all recorded audio outside of software, then I tend to arrange and mix in one place). I find the separation helps me think about each task in its own way, and avoid getting stale looking at the same thing for too long.
Also taking time away from it helps, for fresh eyes/ears. I also find separating major tasks helps me make extreme/bold choices when I need to.
Lately I’ve been experimenting with sending hardware instruments into multiple loopers in Ableton live (one at a time) – allows easy looping mechanics coupled with some track separation, as I best see fit. Some loopers get multiple loops, some stay with a single voice. Can also record looper output into a separate track for later use.
3 seems like a very good rule to stick to. Edits like that always kill the feel.
Do you just record the output of the looper or drag and drop the loop? I guess that way you’ve got less to do later, but I do like undos when using a looper
I just send the track that the looper is on to another track and record whatever’s happening. That way i have a ‘backup’ on a separate track, in addition to the built in looper functionality – using undos to backup, as you say.
You can build a rack with two (or more) loopers so that you can midi map between them to choose which you’d like to record to.
Ah nice, I’ve been meaning to get more involved with ableton looper since I moved over from logic. Lots of potential there. Now I just need to get my guitar sounding less rubbish in ableton.
Yeah, sometimes something just isn’t salvageable, easier to just do over. If you’re prepared to re-record from the outset, it seems less daunting to do. Plus it helps improve your skills in the long run.
Nice!
Makes me wish Ableton would realease a “Live Set Player”: let me run a set w/o display, on a Linux box (Raspberry Pi FTW).
i am always hanging out on either side of some kind of perfectionism and a wabi sabi approach.
technically, i make lots of lists and do give myself hard deadlines for things. i like the feeling of working hard and being a little rushed at the end of an album creation process. for me it’s the only way that i’ll finish. also when i say “rushed” i mean pretty chilled out but knowing that we have a mastering date the next week. and philosophically i think it’s very important to be at peace with a final work existing as it naturally came into form after you put in the work that felt right for it. if you like it, someone else is going to like it.
i have fallen into the trap of working something to death and i think in some cases, working on something for longer doesn’t always mean it’s going to make it better. but i think the essence of what you’re actually creating has a lot to do with all that too. the music that i make is very rooted in human nature and flickless and emotions and imperfections. so i try to make things that communicate that as well.
still, i really do believe just setting goals for yourself and giving yourself a reason to meet them (a show booked that you wanna have recorded music for, a physical release deadline, the beginning or ending of a season, the anniversary of something, etc…). this has been my method for my whole musical life and it’s always worked out pretty fine.
This. It hurts to admit it needs to be done but the work is the better for it, in my view.
I mostly record live takes now, and editing rarely goes further than any combination of the following:
- top and tail
- run through simple FX in ableton
- add a second take as an additional layer
- add layer(s) of found audio
- EQ
- master in Ozone
Like you @HateNames, I find working in a DAW really draining, and not fun. Removing myself from that as much as possible seems to help my productivity and general positivity regarding what I do. Generally the pieces I spend the less time on, editing-wise, is the work I’m happiest with.
As sort of an exercise in discipline (i.e. concentrating more during recording) I’m starting a monthly release series on Bandcamp, for the next 12 months. Drone/noise-based pieces made with synthesis. I’m working with a few simple rules:
- one take
- an extra layer can be added if completely necessary
- FX can be added if necessary
- record on one instrument (modular/kastle/PD Benjolin/etc.)
- duration 20-60 minutes
I’m excited to see how it will go 
I once made an album successfully by looping very rough un-synced live takes on instruments and synchronizing them in Ableton (I’m guessing this is something like what you’re trying to do). And, yea, you’re right, it takes a lot of time in the DAW and it isn’t the most fun, but I’ve found It works better than playing to a click because I can think freely and improvise a little. The trick for me was kind of like “looping in real life before looping in Ableton.” Cause you can’t just splice things together anywhere, more or less it has to be on transients or places that sound the same (you may want to move more toward cross-fading if there aren’t many transients). So I sort of just come up with mostly the part I want, then just play it over and over again into the computer, sometimes changing, sometimes keeping it the same, for like 5-10 minutes. That way I have lots of options for splicing. Then in Ableton I go through and add warp marks at the beginning of every phrase, line those warp marks up to the timing markers, bring the master tempo back to the tempo of the original recording, then get to splicing. I find two or more loops that I like, then splice them together at any point inside them that sounds similar enough, then loop those two loops through the whole track. Then I add more layers. But yea I’ve found this way I can turn any of my random ideas into quantized loops ( though I roll asynchronous these days cause DAWs are boring )
My strategy, which accidentally developed, is:
– Macro improvement is better than micro-perfectionism.
– While working on a song, work quickly. Don’t be too judgemental, just go with the flow.
– Set up the composition/mix/effects/everything before recording. Record a single stereo file, preferably in a single take.
– After recording: unpatch everything. Remove any opportunities to second-guess or to go back and edit – intermediate files or whatever.
– Clean up the stereo file in a WAV editor (I like Sound Forge Pro). Not a full mastering job necessarily, but clean up blank space at start and end, apply fadein/out, tweak dynamics and EQ a little. Overwrite the original recording with these changes.
– Walk away from it and forget it.
(That whole process tends to take about 1-4 hours in a single session, not counting the initial jamming before I committed to making this a recorded song.)
– In a different setting, on a different day, listen to what I’ve done. The most recent songs, some other recent stuff and maybe a little older work for comparison.
– This is the time to judge the work. Figure out what works well and what doesn’t. Use that insight to direct future songs.
– Master and release the best ones.
(For me this process evolved from 13 years of working with 100% software (thus no multitracking), and then a “song a week” project in 2016, during which I acquired some desktop hardware that was pretty unstable and was easier to work with if I could finish in a single session. It’s translated very well to modular though.)
My timing strategy when I’m patching is as basic as ‘1 song in an evening’.
Which is pretty much the antithesis of how I used to work. I used to make music in Ableton, and my problem was actually getting stuff finished (so many options and the mindset of 'I’ll save it and come back to it later). I’d just end up with 100’s of sketches, rather than ‘finished’ tracks.
Having a set time-frame to get something tracked by the end of a ‘session’ is working well for me. If it doesn’t sound good to me by the end of a session then c’est la vie, onto the next endeavor 
If it does sound good at the end of the evening (a rare event 
), then I track it into Ableton (running as a glorified tape machine) via the K-Mix.
The K-Mix is proving to be a real nice bit of kit and a very elegant solution. It can easily be used as a standalone mixer and/or USB soundcard, and the line inputs will happily take eurorack levels 
since parting with trackers, i’ve always used the DAW in its most simple aspect : a timeline where i can put multiple bits of sound in some relationship with each other. I use very few features: splice, copy/cut/paste, move, reverse, gain and pan automation, compression, eq.
I don’t care much for loops, tempo sync, grid snap or anything like that in my current music (which doesn’t mean it’s not precise to the millisecond sometimes).
The sound themselves are made within other tools.
When i was tinkering with tracker and “loop-based music”, i had this problem very often, to get one, two or three loops that made the beginning of a track, and to spend much time refining them, only to find that these loops were “too perfect” for me to be able to come up with anything that would follow.
With the “DAW timeline as canvas” approach, it is much easier to lay out general plans, and to work on all parts of the piece “simultaneously” rather than sequentially. What i mean here is that while working on a piece, some sessions will be dedicated to adjusting only one specific aspect (for exemple, “how does that resurgent object evolve over its various occurrences, and does it make sense in itself and in its spectral relationship with the surrounding stuff”, or “actually all those different clicky/impulse parts need taming in the highs because when i first did them i was not paying attention to that / had ear fatigue and the 3-6 kHz stands out way to much for someone with an average hearing”).
My process is a bit like operating numerous planing/varnish passes (dubious woodworking analogy), each pass is a general overhaul as well as attendance to one single detail.
The finished state comes from a combination of things: external and self-assigned deadlines, “boredom” of the project, a feeling of “not-much-more-can-be-done-here”; it’s not easy to explain but appears almost clearly when it’s time.
So, relationship to time is two-fold : it’s about how much approximately the piece is supposed to last (which is obviously susceptible to adjustments, sometimes drastic); and about how many hours/days/months i envision it will take to attain a finished state.
I find that i pull much more work when thinking in terms of “how much time do i have to do the next session/pass on this project” rather than trying to think of the final deadline.
After the piece is declared finished, i let it sediment for some weeks, come back to it for one or two last adjustments. Now it’s definitely done.
“… finishing things” 
Yeah, right.
The only way I have gotten anything recorded is to play live and record to stereo. I don’t edit other than to fade in/out and maybe normalize when levels are a little low, and I haven’t (thus far) used any other instrument than the modular. I envy people who have the tenacity to edit and arrange, but I just can’t.
I’ve managed to put out three albums this way. YMMV
Good to know I’m not alone.
more
process talk…
me and sam played for 20 min
(drums and guitar)
recorded at a studio to tape
I made a loop in protools of a few bars
(define a loop that makes sense, doesn’t even have to be in 4/4, if you can play while moving the loop point, you can hear it 'make sense)
copy/paste it for a few minutes
cross-fade the edit points
(if they don’t cross at the zero line
there’ll be glitches, maybe you want 'em)
old protools 7 had a 'duplicate X times with cross fades function
maybe it/others still do
or just hit every edit point
laura did 2 takes of vocals
I edited it into a 'song
crazy cuz a week after this came out,
the 'authorities decided to shoot ('retire) a real mountain lion in downtown santa monica
iphone photos of superba ave in venice
made a video in imovie
the idea was
we can put concrete wherever we want,
and it’s not gonna stay that way forever
reviving this very old thread, and not really sure this is the place but here goes…
so, I’ve written(or half written in some cases) a bunch of new works, now I’m at the point to widdle down, edit and cut the chuff.
looking at how to organize and shape the feel and outcome of this album
my idea is to map it out on some paper/poster board/art type of situation, that way I can monitor types of tracks, feel, length, instrumentation, maybe even dynamics. maybe there is more to find out as well as I slowly move along…
I’m thinking some cue cards I can tape to some plywood in my studio, that way I can arrange/tracklist/delete in an open visual way
I’m looking for maybe some methods some others might be using in their practice
hmu