Thanks, guys.

 Hey Thomas,

 I fully understand where you are coming from in that first post. It was an interesting read to see how some of the others work through a project. Great ideas and different perspectives.

 In the past I tended to get bogged down with the selection process, whether that be sounds,midi drum patterns or trying to properly select a particular channel setting for my track or tracks.

  What has worked well for me is to occasionally take a few hours here and there to go through drum sounds track settings and those tons and tons of synth sounds to make a note of what I like, what I might use and what I will discard for my purposes. I will give myself  tech training time so that whan I decide to record I have a few things I understand pretty well ready to use. The longer you are into recording/playing/mixing the easier this whole process becomes,especially in familiarity with the technology and how it works.

 On the artistic end of things I have a simple setup at home. Even though I have an interface with a lot of input channels on it I simply have a mixer going into two of my favorite channels and into Sonar. Within 5 minutes or less I can have tracks armed and on the ready for anything I throw at it. I have all of my guitars and keys right there ready to go. Sometimes I might be playing something I have rehearsed that I'm playing live the next day and I simply hit record and track my performances. Unfortunately a lot of it is copyrighted material so I tend to mostly track only my original stuff usually.

 Being confirmed ADD I NEED to have it simple in the beginning.

 So I have what I would call a simple hardware template tied into a simple software template. I don't tend to get too far into templates because I don't want my stuff to all have the same qualities. I might have a few track templates that work well with certain instruments I regularly use.I don't tend to make dense mixes usually so there is no need for me to have a 60 track Kontakt mix templated. Just what works well for me.

  Once I have a pretty decent mix in the box then I'll put on my engineer hat and dedicate some time to the mixing end.

 Like others have said if the whole thing starts to stink, I'll simply throw it out or save it so it can take up space on my HDD. ( I almost never go back to an old mix I have discarded).

FOR ME, 
it's a simple two-step deal.


i setup the engineering first, just to CAPTURE (1st step).

i have very simple requirements:
good mic, good pre, good rig sound, good vocal mic levels,  setup monitor and playback levels, then i walk away from the engineering for a while, come back, and simply focus on the CAPTURE (performance)

i'll setup in advance, for multiple guitar parts, bass tracks, vocals, etc....

then once i've set it all up, i just simply hit the RED button, or cue up for punchins, etc.


once i've done all the recording and overdubbing, then i switch to ENGINEER (step 2).


i like doing both.

but i never do both at the same time.

requires a different headspace.
but i can do both, so why not?

The problem is that people don't record music anymore like they used to. In the day, when there was an actual band involved plus an engineer, they could just sit there and play and work on the song and the engineer could tweak mic placement, EQ, comp, etc... The engineer could get it as close to what it should really sound like as was practical (time permitting), and could work on the band to get rid of frequency conflicts, and so forth.

After that, they could then take a break and then come back and record, and it's all there, the art and the engineering. Not that there wouldn't likely be stuff still to do, overdubs over the basic tracks and all that, but much of the sonics could be worked out before record is ever hit.

If you are recording yourself, all that goes out the window. You can't sit there and play four or five instruments at once to see how they are going to fit together. But, it's also, IMO, lazy and ultimately sub-optimal not to figure out what you want to do before you start doing it. Otherwise it's just wanking and the results show in most self-recorded content. Not that once in a while you can't just sort of muddle out something nice from scratch. But that's going to be the huge exception, not the rule.

It seems to me that the only way around this connundrum is to do a sequence of demos of the song before you do the real thing. Even back in the day, when all of the above benefits were there, people often did that. Just get the idea down really roughly with voice and guitar or voice and piano. Make sure it works as a song even in that rough form. 

Then start adding parts on that, without any real concern for performance, just for composition. Abuse it as required to sort of shake it into a useful configuration and work on possible part changes and such. Don't get caught up in engineering, just write the song. Cut and paste like crazy, whatever it takes because it's songwriting and composition, not recording.

Then do another one where you take it up a notch in terms of sonics and performance. Where it's really a rough demo of the song. Live with that a while and come back to it and listen to it after a bit of time away from it, and think of ways it could be improved. Tweak EQ and compression and automation and such to get the feel and blend the way you think it should be. Listen for frequency conflicts and how you could change the parts slightly to avoid them (instead of just fixing it in the mix.) Listen to how this or that part might play around this one other, or this is getting in the way of the vocal, etc...

At that point, you now know how every part should sound, what every part should be, how the parts can interact in nice ways and so forth. At least you know that FAR more than you would have had you just sat down and started screwing around. You can then set up to get the sound you want without having to guess, and you can play the parts and know that they are going to fit. And you can PLAY the dynamics, not put it in artificially at the end with automation. So you can concentrate on both sonics and performance because you've done the preparatory work, and you aren't trying to figure out what you are doing, you are doing what you know you want to do. Importantly, you can also use that rough demo as a backing track to lay down the first couple parts of the final take, since they are always problematic having to be played in a vacuum.


That, to me, is the answer, but few self-recorders probably ever do it, me included most of the time. It's so much easier to just sit down and wank really, because time is limited. Whereas sitting there for days with an acoustic and working out chord changes and melodies, and then doing multiple demos is a lot of actual work.
 
But, if you look at the truely great albums, it's doubtful that one single song on any of them was made by just sitting down and screwing around. They will almost by definition have been slaved over, with multiple rough demos (at least of the sections) having been done, and with hours and hours of work on composition and tone before any recording was done.   And you'll end up with a far better, far more organic, or varied, interesting, emotional, and rich song in the end if you do, IMO.
 

There is no right or wrong way in terms of how a great piece of music is created and recorded. I think some of that also depends on the quality of the musicians you are also dealing with.

Sometimes I plan out (all) the ideas and concepts well before the engineering part but there is also another approach that I use and that is using improvisation. There is a very fine line between improvisation and actual composition. I might only have a single thread of an idea and at that point sit down and record it but then improvise as to what the rest of the arrangement is going to be. Often the ideas that flow after that are very good and completed arrangements can be completed in no time. In fact it could be argued that it is a very good way to compose. Don't have any pre conceptions as to what the next ideas are going to be but simply let them flow as a result of each preceding idea.

Look at how Sting worked around the time he was working with the likes of Kenny Kirkland etc.. I have seen a great doco on that. He would come in with only a small thread of an idea and that band was so good they came up with the rest of the song and all the parts in minutes in fact. Even Sting himself said he was amazed at how well they could do that.

Duke Ellington often wrote his most famous tunes in the taxi on the way to the session and he scribbled stuff down on a napkin on the way to the session. Got there, arranged it in minutes and recorded the most amazing hits and standards we still have today. No planning involved at all. Genius you see. The first thing that Mozart wrote down is what we have today yet Beethoven agonised over his music, just the opposite. James Horner does not think about the music either, he just writes the first thing he thinks of down but gets standing ovations in the recording sessions.

How often have you heard about the most amazing songs and pieces in history were written and arranged and recorded in minutes. It is often the way. You have got to be a great improviser though and many people are not. In the mid 90's I recorded a whole album of world music with these amazing world musicians. We assembled and had actually no idea of what we were going to play. people would just start and things happened. You would swear blind that some of the music was extremely well planned and took days or weeks to write yet it was all done instantly on the spur of the moment. 

Yeh, if you have a band, then nothing I just said above is relevant. My whole post is about the self-recorder, who doesn't have a band and therefore they can't sit there and work out the parts interactively. If you have musicians standing there you can quickly make adjustments and find nicer parts. If you are a self recorder, you have to start putting down parts long before you can hear it as a whole, and that makes all the difference. Working out the song in a sequence of incremental steps is the self-recorder's version of having a band.

BTW, I have that documentary. It's called Bring on the Night. It's quite good. I like the versions of the songs (which became the Dream of the Blue Turtles album) better than the actual album, because I'd watched the documentary many times before I heard the album. As with Lanois' Black Dub stuff, I came to prefer the sparser, less produced versions.

droddey


Yeh, if you have a band, then nothing I just said above is relevant. My whole post is about the self-recorder, who doesn't have a band and therefore they can't sit there and work out the parts interactively. If you have musicians standing there you can quickly make adjustments and find nicer parts. If you are a self recorder, you have to start putting down parts long before you can hear it as a whole, and that makes all the difference. Working out the song in a sequence of incremental steps is the self-recorder's version of having a band.

BTW, I have that documentary. It's called Bring on the Night. It's quite good. I like the versions of the songs (which became the Dream of the Blue Turtles album) better than the actual album, because I'd watched the documentary many times before I heard the album. As with Lanois' Black Dub stuff, I came to prefer the sparser, less produced versions.

Hi Dean-


When I started this thread, the scenario you describe is exactly what I had in mind. I'll go into my studio with an idea. To work out additional parts, I record the first one. Then, I get sidetracked by trying to make that first take sound a certain way before resuming to adding more parts. 

It's pretty useless to do that. Everyone knows that it's not what it sounds like, it's what it sounds like in the mix. What you are doing is the worst case scenario of solo processing, in that you aren't even waiting until the mix to solo. Unless you are very experienced and know for a fact that you want a particular sound, which might not sound particularly good at all by itself, then it's really a waste of time to work like that. If you have at least the basic backing tracks in a rough form, you can work out the EQ and compression and dynamics and use those as a guide for a subsequent effort.

You can always do an iterative demo if that's more comfortable, where you never really start over again but just cycle around, re-recording parts and improving them as you go, but always with the overall song behind you for inspiration and to help you get the natural dynamics (and to make the instruments serve the vocals.)

If you teach yourself to listen as you play, you can do a LOT of mixing with your fingers, by finding the parts and tones that blend with the other instruments. It's hard if you can't isolate yourself from the source enough, which is a common issue for us self-recorders in small spaces. Sometimes you have to extrapolate from what you are hearing in the air to what you know is being captured.