The weekend so far

Yesterday was our annual June party. It was lightly attended, but there was plenty of good conversation. As usual, the flower_cat held court in the living room, where the conversation is wide-ranging and occasionally raucous, while I stayed mostly in the office where it's quieter and the talk is geekier -- mostly about computers and music.

Sold two more preorders.

I spent Saturday morning and much of the afternoon reconfiguring Dantooine, the 800MHz nini-ITX box that I had been using as a workstation before I built Trantor, as the new print server and Linux client machine, the latter so that guests would have something to browse with. I really need to ditch the old 17" Hitachi CRT -- anybody want a nice boat anchor monitor? Showing its age a bit, but still very functional.

Configuring the printer went well, though it took me a while to figure out that the PPD files in the hpij package were newer than the ones in hplip, which was installed by default. I worked on system admin rather than the audio editing because I wanted something that was easier to interrupt as people arrived at the party.

Managed to squeeze in a short (2-mile) walk at about 4:30; the logistics just didn't work out for a longer walk in the morning.

OK; time for a new backup drive

Just looked at the stats on my backup drive: it's 98% full -- about 133GB out of a 144GB partition. And since it's a Maxtor drive from a known-bad batch, it's time for a new one. This would also be a good time to roll out the updated scripts that I'm using at work. And a little cleanup wouldn't hurt -- multiple temporary copies of some of the track working directories are part of what jumped it from 86% a week ago.

I was wondering whether to copy the old backups to the new drive; that would make sense if the new one were 400GB or bigger, but at 200 I'll have to just archive the old one and start fresh. Good excuse for some more extensive re-organization server-side. And I'll see if 500's go on sale at Fry's this week.

Tracks

Integrated super_star_girl's chorus vocals into "Bugs", and fixed a few assorted mistakes in the stereo map. Fixed cflute's whistle part on "Guilty Pleasures" by swapping in a clip from the first take. (Note to self: cflute has an excellent memory -- wait for her to stop telling me what she got wrong before hitting the stop button.)

Filtered out the sibliants on cflute's vocals on "Demon Lover" (audacity's "acoustic" EQ curve seems about right), and adjusted levels a little.

For those of you keeping track, this leaves 12 tracks that are essentially done, and six that still need work (two from others, the rest just me). I'm hoping I get over my cold sometime soon so I can re-record my vocals on "Someplace in the Net"; spending all day with a scratchy throat is annoying. The ones with just sync problems I may be able to fix by time-shifting -- I've gotten pretty good at that from working on the kids' chorus vocals.

I went to Guitar Showcase after my (3-mile) walk and picked up a shaker that I think will work for "TEOTW"; I'll probably get to that tonight. (Also got a nice set of rosewood claves and a couple of microphone stand quick-connects, but those don't come into this part of the story.)

Wedding planners take note

From this post by technoshaman we get a link to this particularly nice wedding ceremony.

Murder, mastering, and mayhem

From this post by eleccham comes a link to this article in The Times titled "Why music really is getting louder". Basically it's about the current practice of mastering music so damned loud that it clips.

The article has a few inaccuracies: for example, it confuses the kind of dynamic range compression done at mastering time with the kind of data compression done by codecs like MP3. But the fact remains, a lot of popular music these days is compressed to within a fraction of an inch of its life, destroying any dynamic range it might originally have had.

Fortunately, I don't listen to that kind of music, nor do I intend to subject my listeners to it. I do a minimum of compression on my CD tracks, and I do it by hand in Audacity using the Envelope tool. That lets me bring down the notes where I whack a chord too hard, or the verses where I lean to close to the mic, or the spots where I pop a "P", without affecting the overall dynamic range of the music.

Not that I have a whole lot of dynamic range when I'm singing. The main challenge in the final phases of the mixing will be making sure that the songs where it's just me and the guitar don't have me sounding louder than the ones where I have backup. That will require overriding the standard normalization parameters for just those songs during the build, so as not to maximize their volume.

I've also decided not to throw money at Oasis to have my CD professionally mastered. Maybe next time, if I can find an engineer I can trust who's local and who has done folk music before. For the moment I think I can do fairly well in Audacity, which also gives me the opportunity to correct my problems at the track level before they get mixed down.