Quality control

It's faster for me to start over than to try and fix something that doesn't catch me. It has to sound a certain way, if it starts seeming dull at any time then I know from experience I'm better off ditching the whole thing. Maybe not all the raw materials, soundwise, but the whole arrangement gets binned.

After stuff is finished, it gets tougher. But usually it's a relative sense, like what catches me more than other stuff, what perks up my ears the most, top ten, etc.

I'm usually the pickiest person about this out of every group I've been in or worked with.

1. As I'm teaching the song to the band, I change parts and tweak it until I'm happy.
2. Once there's a working version, I let the band make suggestions on things to change if they feel that something could be improved. E.g., the other guitarist might change a few parts or add a bit of singing.
3. If I'm still unhappy with the song and the lyrics are OK, I rewrite some or all of the parts.
4. If we can't get past 3 with a song, and the majority doesn't like it, the song is axed.

Currently, we have one song that's reached 4 and one song that's in stage 3 awaiting the rewriting of the chorus. I might have axed it or totally rewritten it, given my druthers, but the rest of the band protested.

I feel that songwriting is a process of experimentation and refinement, not momentary genius, at least for me. If there are good parts in a song but the song as a whole is a failure, then song should be axed and the good stuff recycled. There's a limit to this kinda refinement, however, else you'll make your bandmates crazy.

For solo stuff: work, slowly chip away at stuff that makes me go "huh?", listen, repeat as necessary

Firstly, I tend not to write songs down when I first come up with them. Or, if I do, I write down just the main verbal hook, and maybe a Gregorian style sketch of how the music should go with them. If a song germ isn't catchy enough for me to remember it, then it's unlikely it will get written.

Then, there's a lag in the demo-ing stage, between the song getting blocked out in Reason, and then properly demoed with vocals and all, in Cubase. I often wait from overnight to a few days to finish a song. Songs which are catchy or have something compelling about them will get finished much more quickly, while those that aren't will languish in the sequenced but still in progress folder.

Finally, once a song is finished, then it goes into the final QC stage of going up on the MySpace, being played to other people (manager, the performers in my previous band, friends) and has the final YesOrNo. By the time it's got through all that previous filtering, there's usually very little debate beyond a very little bit of arrangement. But that was the nature of my Svengali/Songwriter role in that band.

This of course has been very different in other bands which were more collaborative.

listen to things on a micro scale. loop 4 bars/8 bars/whole sections. if it gets boring, it probably is boring.

with band based stuff, i try and play with people who i respect so much i'd never question what they're doing. granted, most of our stuff is improvised. i think of my drummer as some sort of self generating beat machine. i'll play something he'll complement it. of course. same goes for the rest of our players. looking back at older bands it was those horrible "hey mr bass player, why don't you try doing this kind of groove, look watch me play this" moments that really killed everything.

Anyways, there's usually a consensus on what's working and what isn't, but I think I'm a little pickier than the other members, and there have been one or two songs that I took out of the set because I didn't want to play them anymore. It isn't so much about crowd reaction as it is about how I feel when I play the songs. If I feel awkward or unrocking when I play them, they generally fade out of the set pretty quickly.

Yo just reminding y'all that this is mad import

My quality control method is basically 1) write a lot of songs, 2) throw most of them away.