|
Back to Nightlight Press |
Archive for January 2003 http://www.jeffreyatw.com/movies/union.mov I hope they don't take it down. I laughed so hard, except at the very end where they got all moody and liberal and weepy. It's an edit of footage from the State of the Union address to make it look like Bush is saying horrible anti-American things. While I see how they did it, it's kind of scary how good the sound matches up so well (if it was choppy, I wouldn't have thought so much of it as I did). I've spent a lot of time in the "studio" trying to figure out how to engineer a song, and Adam recently asked how I do it. Well, I never really got technical about it, and I think it's time to share the paltry things I've learned. I use Cool Edit Pro, not Cool Edit 2000 or whatever the latest version is now. I'm sure they're similar. This is a product that lets you do a lot of tricks with audio, and lets you multitrack. Overview I try to make use of the stereo field. A typical song I work on will look like this: L--7--------3------5----1---8----2---6-----4--------9--R 1, 2: Main vocals 3, 4: Main guitar/instrument 5: Bass or "support" instrument 6: Drums 7, 8: Backing vocals I'm fond of double-tracking both instruments and voices now. I like a wet voice, or at least one that has backing of some kind. I used to dump a lot of effects and chorusing onto the lead vocal because I didn't like my voice. Now I just don't sing songs with notes I can't hit. To double-track, you sing it twice and both versions on either side of each other. Ditto for guitars. I think it makes it sound more fleshed-out and dynamic. I never cared as much for one instrument, one voice stuff. Voice It's probably important how I think about singing, and how it relates to what I think is nice to do with the voice once it's recorded. I'm not a big fan of really emotive singing, like Steven Tyler screaming and going "uh!" and screeching at the ends of words. Nor do I like a lot of breathing on the track. With that in mind, I will try to sing it really straight, few little quirks and warblings and what-have-you. Once that's done, it's time for noise reduction. You can select the whole waveform and do noise reduction on it (like 2 or 3%), but I find it removes some of the track itself and you get artifacts all over it. It's better to record some silence and use that as the sample for the noise profile. Then, remove from 40 to 60%. It'll still sound clean. I used to remove a lot more, but during Electrologica I was desperate not to have any noise on the track at all. It's not worth it; it starts to cut into the voice. Next I use a very mild compressor to clean up any edges on the tops of the waveform which will prevent me from normalizing it later. It's only to remove any outlying peaks. A non-existent compressor profile looks like a line at 45-degrees. I cut it off at the top 3 or 4%, making a little flat space up there in the upper right corner. That will kill outliers without touching the voice. Listen to the track -- if it sounds "compressed" and the volume is wobbly, undo it and reduce the amount you cut off this time. I prefer to add a little tiny bit of echo instead of reverb. Reverb sounds like a deep, empty hall to me, even if you use very little. An echo with a low delay, low decay and high initial volume will drop off quickly and add a nice presence. Also, I play with what frequencies get echoed; I prefer that only the mid-range gets through each time, but make it less than 100% or you'll end up with a whine running through the whole thing. Instruments Lately I've done only guitar, and my tricks haven't worked that great. In the past I could give each MIDI instrument a different something, but let's focus on guitars. If it's electric sans distortion, I will leave the waveform pretty intact, doing the same noise reduction and compression as I did the voice. But the only thing it may need is an FFT filter, depending on where it was recorded. I work in my bedroom, so I'll add a tiny bit of reverb (which is all right on a rhythm guitar, but makes a voice sound mushy to me). I like to increase the treble a little, but lately I've been leaving it as is. If it's an incidental little distorted guitar, I increase the treble a lot. It needs to stand out. If it's sort of a rhythm distorted guitar (which I don't do because I'm not that precise with it and it ends up sounding washed-out), I will instead increase the bass a little, since you want to hear the voice over it. Common tricks with synth instruments involved flanging or chorusing them. I like a minor flange on bassy synths, because they're so monotonous the listener almost forgets they're there and get annoyed at that low noise. At least, I do. I don't think it's such a great idea anymore. I would flange for the sake of flanging. In multitrack Once those are set up, you can go back to multitrack and start fussing with the volume levels. I spend a long time on this because I'm not very good at it. The problem with recording, I've noticed, is I make the voice much louder than it needs to be, because I want the vocals to be clear. Unfortunately, upon mix down, I don't get one big fat waveform, but a thinner one (the guitars and stuff) with fat segments (voice over it). I can't normalize that, it'll just stay at the volume of the voice. Lately I've learned to drop the voice into the mix more. If it's unclear, I might increase the treble. Treble makes indistinct sounds clearer, but too much treble sounds tinny and hurts the listener. I listen to each track separately, and then together again, with their respective pans and volumes. Then I start listening to two at a time, etc. The resulting mix down should look consistent across left and right. If it's not, I'll use an envelope to maneuver the volumes into being closer, but generally I have to do back and fix it in the multitracker. Listening to it all together also lets me find where I might be off-key and I can go in with the pitch shift and fix it. You'd be surprised at how many sins just dubbing another voice covers. It sounds good together, but alone each voice would be way off. In the past, I'd do voices, mix down, then realize something was off, but I'd mixed down already, so I dubbed over a quieter, on-key voice. That's bad practice too. More to come. I was walking down to the fourth floor, to use that restroom instead of the heavily-trafficked fifth floor one, when somehow I missed the step. As I fell, I managed to grab the handrail and prevent from falling forward, so I landed on my back instead of rolling the rest of the way down. My back hurts in the middle, in the same place it once hurt when I was little and had run into a sliding glass door. It wasn't really a very bad fall, but I'm paranoid that it could have been one. Figures I'd trip at the damn TOP of the stairs. I have all this Coldplay stuff stuck in my head, and they're kind of depressing, but I like them a lot. I'll give you a couple reasons. - I can identify them in an instant. All their stuff sounds uniform. This appeals to me. It's like a cross between Secret Messages and... some other band with a lot of reverb. - The vocals -- he does this thing where, somewhere in the song, once or twice, there's a high syllable. That's another "cute thing" or identifier I like. I want to find something unique like that that my music would be known for, incorporate it into almost every song. - Their stuff is really, really simplistic. This is a mixed blessing. I can learn it easily, and am impressed by what they can do with four chords, but sometimes they don't do near enough. In an attempt to exorcise a song I like, I cover it. This is one of those times: Green Eyes (3.5 MB) I have two versions: an augmented demo (linked) and a cleaned-up version. I don't like the cleaned-up one. If the song was about robots and metal surfaces, I'd clean it up. But this is a hippie coffeehouse song, so there's lots of noise left in. "I'm not one for beer, although I can drink it on occasion. That being said (and being half-Irish), I firmly believe that Guinness is overrated. To me, the time or two I tried it, it tasted like I imagine drinking a tall glass of dirt tastes. Supposedly it's 'an acquired taste.' Eating a bowl of maple leaves or shredded baseball cards would be 'acquired tastes' too, but that doesn't mean you should try to 'acquire' such a 'taste.'" -- off of fark.com OMG OMG OMG OMG 1) On Malcolm in the Middle, which I don't watch, someone was making a '70s mix tape and asked someone else if they should put ELO on there. After that there's a scene where they're in the car listening to Telephone Line! 2) On the last Simpsons, which I've also stopped watching a long time ago, Homer asks Lisa "Who's your favorite Traveling Wilbury? Is it Jeff Lynne?" That probably means The Simpsons will be all about ELO and Jeff Lynne this year! Will he guest star?? So sayeth/thinketh the ELO newsgroup. Kurt, at end of phone conversation: "Well! Touch your bones. *very long pause* feel 'em up." Kris: "Birthday wishes. Well expressed." I felt really crappy this morning, and I got into work after lunch. I'll take it out of vacation. I don't know if I'm going to end up doing anything tonight as a result. (I didn't really have any plans anyway, unless calling people 20 minutes before we do something is viable.) For some reason, yesterday, I got four or five e-mails about the current "Chex gets ditched" storyline. Why yesterday? I rarely get one piece of mail about CN. I guess a lot of people can identify with it. I can identify with it. This is the most autobiographical CN evar. I really don't like to do things like that, because whenever some artist gets all autobiographical, I can't help but feel uncomfortable and say "why do I care about this?" Especially in the case of rock stars. All I can think is, "you've got an awesome life, and all your songs are about suicide and depression. What's your problem?" So I try to approach it with a good nature, as opposed to turning it into a pity party, as Melissa would say. I guess it's not exactly autobiographical as it only deals with general feelings of being outcast or left out. The situation of mine that led to these feelings doesn't resemble Chex's at all, but the results are similar. That said, I'm having a hard time resolving Chex's scenario. I had an idea, but it would involve Chex learning, and I always avoid that if I can. Hmmmm. I wondered, how often does my brother's name, full first name (Kurtis, not just Kurt) and last, show up on the internet? Only once, from one of the NiGHTLiGHT articles he collaborated with me on. Sorry, old bean, but you're going to have to do better than that if you intend to beat me at all aspects of life. http://www.brokenboy.cjb.net I had a band once. I mean... it was cool. But it was just me and MIDI. Now you can languish in the depths of failure, like I did, or you can be cool like Dmitry, Yelena's brother, who plays the guitar like I do, except he actually plays the guitar instead of pretending to. I was impressed. Speaking of which, my God! MP3.com changed their site! They look more hardcore, and less... furniture store, from back when I was blazing up the progressive rock charts. |
