So the front page article of The Independent this morning was full of the terrible bird flu again after some parrot from Suriname died in the DEFRA quarantine this weekend. There probably hasn't been this much made out of a dead parrot since John Cleese walked into a pet store in 1969. It just gets a bit silly after a while.
Frankly the press does a pretty bad job at covering these things. They tell you that it's killed some 60 people in 3 years, which frankly doesn't sound that scary. But then someone else is quoted of saying if it mutates it will kill 50,000 people in Britain, which sounds a lot worse. Still, I don't really know how much worse than regular flu this disease is. How deadly is it? What are we talking here, smallpox or a bad cold?
Not that it really does any good to know how deadly it is. Just yet another version of "death from above" ready to strike at a moment's notice. Worrying about bird flu is probably more likely to kill me than bird flu is. Much better to just go watch a Monty Python DVD.
Monday, October 24, 2005
Sunday, October 23, 2005
Jeremy vs. The Jesus Visa
In one of those bizarre ironies that are too weird to make up, my brother-in-law Jeremy is being forced to leave his job at a christian bookstore over his religious beliefs. I kid you not. Jeremy is a devout christian, a "true-believer" if I've ever met one, and he takes his relationship with God very seriously. It appears, however, that his (soon to be former) employers have somewhat different priorities.
The issue, it seems, is the flogging of credit cards emblazoned with the image of Jesus. His bosses want all their employees to be aggressively pushing this with every transaction. However Jeremy has a strong moral conviction that it's wrong to use Jesus' name in this way to make a quick buck. (Apparently he think that it's wrong to profit from the prophet!) Personally, I don't have quite such a strong view on this, but then again I'm not much of a christian either. I do think the notion of pushing a "Jesus Visa" as incredibly crass, and I would be frankly embarrassed to push such tacky crap. It does make a nice symbol for the Republican right though: the evangelical "God Squad" and money minded corporate America melding together in a fantastic morsel of poor taste.
The issue, it seems, is the flogging of credit cards emblazoned with the image of Jesus. His bosses want all their employees to be aggressively pushing this with every transaction. However Jeremy has a strong moral conviction that it's wrong to use Jesus' name in this way to make a quick buck. (Apparently he think that it's wrong to profit from the prophet!) Personally, I don't have quite such a strong view on this, but then again I'm not much of a christian either. I do think the notion of pushing a "Jesus Visa" as incredibly crass, and I would be frankly embarrassed to push such tacky crap. It does make a nice symbol for the Republican right though: the evangelical "God Squad" and money minded corporate America melding together in a fantastic morsel of poor taste.
No End in Sight?
Sometimes I discover things that just strike me as funny for no particular reason. It's a sort of Daily Show "Moment of Zen" sort of thing. (Not that I'm particularly convinced that this has anything to do with Zen, but that's another discussion for another forum.) The long range weather forecast from this morning is an example. It's been predicting this long rainy spell for much of the last week. Amusingly, it's actually quite nice out at the moment. Perhaps this is all some elaborate plan to steal honey. "Tut Tut...?" (Pooh looks down innocently from his perch on the bookshelf.)
We get to watch The Daily Show now. The new digital channel More 4 has taken to airing it on weekday evenings. This is a bit of a mixed blessing. Much as I think this show is really funny, it also acts as a reminder of how truly screwed up the US is right now, which is kinda depressing. I used to browse the New York Times online fairly often but I've given up now. I don't really like reading American news anymore.
I have started reading an English newspaper recently. The Guardian recently had a high-profile relaunch with a new format, which prompted me to check it out. Ironically, The Guardian's publicity campaign actually resulted in making me a regular reader of The Independent. The writing in The Guardian was fine, but the format was truly horrible. It's frankly rather USA Today-ish: Lots of call-outs from the front page, lots of color, and side bars, and info-graphics, etc. It's like a newspaper created by ADD-afflicted editors from heat magazine. Yuck! The best part about the new Guardian is the shape, which folds very nicely into a bag or briefcase. I just can't read it. It feels like a newspaper that could cause an epileptic seizure. Still, I guess it's colourful, (colorful for those reading in America). Perhaps it's all a response to the weather.
Friday, October 21, 2005
Words, Words, Words... and Documentation
Well, it started out well, but in the end that previous entry was a bit of a mess... Part of the problem is that it takes a while to do all the equations. I have a LaTeX equation editor which will output pdf or tiff files, but then I needed to export them to a jpg, and then copy it over to my machine at work, and then upload it to blogger etc... it was all a bit time consuming. Plus, it is a bit of a nasty topic.
Anyway, during the course of trying to finish the entry, it became apparent that I was going to have to cover that topic in the SupIRCam documentation I suddenly find myself writing at work, and frankly the prospect of having to do it twice is just too exhausting. So, I'm going to abandon the topic at this point, at least as far as this blog is concerned. If you're really hooked, you can follow the continuing adventures in the SupIRCam Users' Guide, which is currently taking shape here. (Or here if you prefer the postscript version).
Writing documentation is a little exhausting. However I've discovered that it's not quite so bad if I write it as I go. At least, I'm much more likely to actually write it this way. Plus, it gives me something to do when I get sick of coding.
Anyway, during the course of trying to finish the entry, it became apparent that I was going to have to cover that topic in the SupIRCam documentation I suddenly find myself writing at work, and frankly the prospect of having to do it twice is just too exhausting. So, I'm going to abandon the topic at this point, at least as far as this blog is concerned. If you're really hooked, you can follow the continuing adventures in the SupIRCam Users' Guide, which is currently taking shape here. (Or here if you prefer the postscript version).
Writing documentation is a little exhausting. However I've discovered that it's not quite so bad if I write it as I go. At least, I'm much more likely to actually write it this way. Plus, it gives me something to do when I get sick of coding.
Tuesday, October 18, 2005
Techie Talk: S/N Calculations in the IR
Much of what I fill my headspace with is stuff of a rather technical nature, and rather than ignore that part of me, I've decided to include it in my blog as well. I'll just have to see how it works. If you're curious about what I do for a living then these 'Techie Talks' might give you some glimpse. (Although not everything I do is as tedious as the work in this entry).
And so...
Observing in the infrared (one of my particular specialties) can be a bit of a pain. The infrared sky is actually quite bright, even at night, and so observing is never a simple matter of just opening the camera shutter for a long time. Instead, you always end up taking lots of very short exposures, and 'nodding' the telescope all over the place. It's a painfully inefficient process and at the end of the night you're left with a great big pile of images on your disk drive that you then have to shuffle together in a variety of ways to 'reduce' your data to a useful form.
Usually, I don't bother trying to keep track of the uncertainties by propagating the 'errors' through the reduction process. Instead, I usually just use the scatter in the reduced images to estimate the noise level in the final data product and leave it at that. However, for the data I'm currently working with, I need to understand the noise characteristics of the data at a much earlier step, in an effort to characterize the new detector I'm working with. There are also some potential advantages to keeping track of the S/N through out the process. In particular, if I'm keeping track of the variance as I reduce the data, then I can combine the various data sets in an 'ideal' manner and optimize the resulting S/N. It will also help to write code that can automatically recognize bad data. Since I'm in the process of writing the data reduction pipeline for this instrument, all this seems like a good idea. So, I decided to look into starting with the "CCD equation" and trying to calculate what the noise levels actually should be.
Now for a nice optical CCD, this is not too bad. The sources of noise are simply Poisson noise from the photon statistics, (both from the source and from the background sky), 'readnoise' from the readout amplifiers, and maybe dark current if you've got an uncooled detector See here for a discussion of noise in optical CCDs..
For my work with SupIRCam, the dark current is pretty negligible, and the source noise is usually not an issue (most of the noise is background), so it should be even simpler, right? The problem, however, is that during the reduction process I'm combining the same images in different ways and subtracting them from one another, so it's vitally important to keep track of what's getting subtracted from what, because some of the noise from individual exposures will get subtracted from itself, and therefore disappear exactly. Thus a naive "propagation" of the errors won't work. Some of these "errors" are correlated (100%) because they are based on exactly the same data. Confused? Well, lets look at my data pipeline in particular.
A typical observation consists of a set of images, all of the same exposure time. Between each image, the telescope is nodded slightly so the target lands on a slightly different portion of the detector. So at the end we have several images of the target and background sky, each of the same integration time, but slightly different positions. Immediately before and after this group of images, we take a "dark frame," and exposure of the same length of time, but with the entrance to the detector closed. (Literally a picture of the dark!). This is important for calibration reasons, to remove the small amount of 'dark current' in the image, and to remove the instrumental 'bias' of the detector.
Ignoring the dark current (which is very small), the noise level of any of these exposures is described by
where S_m is the average level of the background sky counts, g is the 'gain' of the detector (number of photons per count) and r is the 'readnoise' in counts. V is the expected statistical variance for an exposure. Now for a dark image, there is no signal from the sky background, so S_m is 0, and we're just left with
Now the first thing we do is combine the two dark frames to make an average dark. Just using standard error propagation gives
This combined dark is then subtracted from each of the individual exposures. The variance of the resulting dark-subtracted images is
So far, this is not so bad, and in fact this is just how far I needed to go to do the noise calculations that started this investigation. However, it's in the next step that things start to get a bit ugly. More on that later though.... It's bed time now....
.... The next step in the reduction is to build a sky image by combining all the dithered ('nodded') images of the target. However, we can't just take an average, since the target will be contributing to the mean, and we'd end up subtracting the target from itself. The solution is to throw out some of the data before taking the average. For each pixel, we look at that pixel in each of the individual images, and sort them by the number of counts they get. If a star shows up in that pixel for one of the images, it will be brighter than in the other images. So if we throw out the highest pixel value for each image we should be left with a clean sky image. (If you have a really crowded field, you may have a star in the pixel for more than one dither, so you may have to throw out more!).
There's a bit of a catch though: the sky background in the IR varies quite rapidly in time. As a result, the sky will be slightly brighter or dimmer in each of the dithered images. But this means that our trick of just throwing out the highest data won't necessarily get rid of all the stars, especially the fainter ones. The solution is to scale each of the images so that the average of the background is the same for all the images, then throw out the highest values for each of the pixels and average the rest of the data together.
To further complicate things, you also typically want to do a weighted average of the data rather than just a straight average. Since the background level will be changing slightly with each image, so will the S/N. To avoid reducing the higher S/N data by mixing it with lower S/N data, you typically want to weight each of the images by the variance of that image. In background limited images (very often the case with IR work), the sky noise will completely dominate over the readnoise, and you can ignore the readnoise in the variance calculation. In this case, the variance is just the sky counts, and variance weighting is just applying the inverse of the scaling applied to match the sky levels. However, in the general case, the variance weighting will be slightly different than just the scaling factor.
Sigh.... this will take a while.... Gotta go again... more later....
And so...
Observing in the infrared (one of my particular specialties) can be a bit of a pain. The infrared sky is actually quite bright, even at night, and so observing is never a simple matter of just opening the camera shutter for a long time. Instead, you always end up taking lots of very short exposures, and 'nodding' the telescope all over the place. It's a painfully inefficient process and at the end of the night you're left with a great big pile of images on your disk drive that you then have to shuffle together in a variety of ways to 'reduce' your data to a useful form.
Usually, I don't bother trying to keep track of the uncertainties by propagating the 'errors' through the reduction process. Instead, I usually just use the scatter in the reduced images to estimate the noise level in the final data product and leave it at that. However, for the data I'm currently working with, I need to understand the noise characteristics of the data at a much earlier step, in an effort to characterize the new detector I'm working with. There are also some potential advantages to keeping track of the S/N through out the process. In particular, if I'm keeping track of the variance as I reduce the data, then I can combine the various data sets in an 'ideal' manner and optimize the resulting S/N. It will also help to write code that can automatically recognize bad data. Since I'm in the process of writing the data reduction pipeline for this instrument, all this seems like a good idea. So, I decided to look into starting with the "CCD equation" and trying to calculate what the noise levels actually should be.
Now for a nice optical CCD, this is not too bad. The sources of noise are simply Poisson noise from the photon statistics, (both from the source and from the background sky), 'readnoise' from the readout amplifiers, and maybe dark current if you've got an uncooled detector See here for a discussion of noise in optical CCDs..
For my work with SupIRCam, the dark current is pretty negligible, and the source noise is usually not an issue (most of the noise is background), so it should be even simpler, right? The problem, however, is that during the reduction process I'm combining the same images in different ways and subtracting them from one another, so it's vitally important to keep track of what's getting subtracted from what, because some of the noise from individual exposures will get subtracted from itself, and therefore disappear exactly. Thus a naive "propagation" of the errors won't work. Some of these "errors" are correlated (100%) because they are based on exactly the same data. Confused? Well, lets look at my data pipeline in particular.
A typical observation consists of a set of images, all of the same exposure time. Between each image, the telescope is nodded slightly so the target lands on a slightly different portion of the detector. So at the end we have several images of the target and background sky, each of the same integration time, but slightly different positions. Immediately before and after this group of images, we take a "dark frame," and exposure of the same length of time, but with the entrance to the detector closed. (Literally a picture of the dark!). This is important for calibration reasons, to remove the small amount of 'dark current' in the image, and to remove the instrumental 'bias' of the detector.
Ignoring the dark current (which is very small), the noise level of any of these exposures is described by
where S_m is the average level of the background sky counts, g is the 'gain' of the detector (number of photons per count) and r is the 'readnoise' in counts. V is the expected statistical variance for an exposure. Now for a dark image, there is no signal from the sky background, so S_m is 0, and we're just left with
Now the first thing we do is combine the two dark frames to make an average dark. Just using standard error propagation gives
This combined dark is then subtracted from each of the individual exposures. The variance of the resulting dark-subtracted images is
So far, this is not so bad, and in fact this is just how far I needed to go to do the noise calculations that started this investigation. However, it's in the next step that things start to get a bit ugly. More on that later though.... It's bed time now....
.... The next step in the reduction is to build a sky image by combining all the dithered ('nodded') images of the target. However, we can't just take an average, since the target will be contributing to the mean, and we'd end up subtracting the target from itself. The solution is to throw out some of the data before taking the average. For each pixel, we look at that pixel in each of the individual images, and sort them by the number of counts they get. If a star shows up in that pixel for one of the images, it will be brighter than in the other images. So if we throw out the highest pixel value for each image we should be left with a clean sky image. (If you have a really crowded field, you may have a star in the pixel for more than one dither, so you may have to throw out more!).
There's a bit of a catch though: the sky background in the IR varies quite rapidly in time. As a result, the sky will be slightly brighter or dimmer in each of the dithered images. But this means that our trick of just throwing out the highest data won't necessarily get rid of all the stars, especially the fainter ones. The solution is to scale each of the images so that the average of the background is the same for all the images, then throw out the highest values for each of the pixels and average the rest of the data together.
To further complicate things, you also typically want to do a weighted average of the data rather than just a straight average. Since the background level will be changing slightly with each image, so will the S/N. To avoid reducing the higher S/N data by mixing it with lower S/N data, you typically want to weight each of the images by the variance of that image. In background limited images (very often the case with IR work), the sky noise will completely dominate over the readnoise, and you can ignore the readnoise in the variance calculation. In this case, the variance is just the sky counts, and variance weighting is just applying the inverse of the scaling applied to match the sky levels. However, in the general case, the variance weighting will be slightly different than just the scaling factor.
Sigh.... this will take a while.... Gotta go again... more later....
Wednesday, October 12, 2005
The Unhearable Lightness of R.E.M.
Now that the weather here in England is taking a decided turn for the dreary, it has been time to start digging up my rainy day music. Certain artists and albums just seem to fit the mood of a drippy wet day. The Cure's Disintegration is a pretty obvious example. Other good choices include Roxy Music's Avalon, Depeche Mode's Ultra, The Residents' Demons Dance Alone, or just about anything by Enya, or Sigur Rós. However, somewhat surprisingly, I find that late albums by R.E.M. also work rather well, particularly Up, and their most recent Around the Sun.
I got into R.E.M. about the same time as everyone else on the planet, when Out of Time broke through into the mainstream. I was first introduced to the band through my (then future) wife Sunshine, who had a tape of Green/Out of Time which she'd made from a friend's collection. I had actually heard some of their songs on the radio occasionally, but this was the first time I'd put the name to the sound. I almost certainly would have ended up collecting their back-catalog anyway, but the process was hastened during my second term at Caltech when I started hanging out with massive R.E.M. fan Tara Robertson. (Her 'fan-ness' was massive, not Tara. She's quite the opposite in fact, but anyway...) So inevitably R.E.M. often brings to mind memories of sunny Pasadena, endless late-night homework sessions, wandering through the steam tunnels to turn in said homework, skipping dinner to walk to Poo-Bah's, and bow-tie pasta stuck to the ceiling. (On the other hand, the song "What's the Frequency, Kenneth?" from Monster is indelibly associated with the image of Patrick Spradlin leaping off of Crud roof into Ricketts courtyard during rotation my junior year, presumably a direct association with the music playing at the time.)
However, such memories are rarely summoned by listening to R.E.M.'s more recent efforts, and it's not entirely due to the my different surroundings when these albums came out, though I'm sure that has an effect. I think that part of this disconnect between R.E.M. as I first got to know the band, and their more recent material is that there has been a subtle change in the group's music. This change became clear to me recently when I followed up listening to Around the Sun with Out of Time. "Leaving New York", the first song on Around the Sun, has a 'classic' R.E.M. arrangement full of 'jangle' and overlapping harmonies, but compared to actual classic R.E.M., it sounds a bit leaden and deliberate. It's this sense of heaviness that marks the biggest difference between the band as they are now and the light and bouncy R.E.M. of old.
It's tempting to attribute this to the loss of drummer Bill Berry, except that this transformation to the deliberate actually pre-dates his departure. To my ears it can be heard as far back as Automatic for the People, with "Everybody Hurts", and "Star Me Kitten". The turning point is Monster however, which in retrospect plays rather like the two versions of the band battling for control. It starts out firmly in the 'classic' mode with "What's the Frequency, Kenneth?," but about half way through "Crush With Eyeliner" you can begin to hear the determined, mechanical rhythm settle in. By "King of Comedy" Michael Stipe is truly set into his late period persona, but Peter Buck's firey guitar and a surprisingly crunchy drum track hold off the change. With "I Don't Sleep, I Dream" we get the first good look at the future with a fully formed Michael Stipe midtempo number. Here, the action is all in Stipe's vocal performance and rest of the band fades into the background. "Star 69" then completely reverses the whole processes, burying Stipe under a wave of fuzz in a song that wouldn't sound too out of place on Life's Rich Pageant if you turned the distortion down just slightly. This back and forth tug continues for the rest of Monster, but by the time of the follow up New Adventures in Hi-Fi it's clear which side has won.
Interestingly, I think the new sound works best when they are not trying to sound like R.E.M. Both New Adventures in Hi-Fi, and Up sound quite nice in their own way. There, the melancholia of the songs seem to fit the mood of the music. It's when R.E.M. tried for a classic sound with Reveal that the disconnect galled the most. Reveal is far and away my least favorite R.E.M. album, and despite a number of attempts I've never managed to get inside of it. Around the Sun is a bit looser, and trying less hard to be a classic. As a result, it works quite a bit better than Reveal, at least on rainy London afternoons.
Saturday, October 08, 2005
David Beckham Got Screwed!
Or at least that's what the Match of the Day commentators seem to be saying. Honestly, I couldn't tell you. I know less about football than I do about cricket. I'm pretty sure that they want to kick the ball into the net, and they can't use their hands, but mostly I'm a bit clueless about it.
I do like Beckham's haircut. It's like a weird floppy skunky looking thing, and it makes him easy to see in when the camera is zoomed out and all the players look like tiny cockroaches running around on a furry ping-pong table. The guy with the braided cornrows also looks pretty cool, but he didn't get to play much tonight. (He seemed to be in the doghouse with the creepy old dude running the team.) Most of the players just look like tiny little sweaty blokes, and it's hard to figure out who's who.
I guess cricket had the same problem to some extent, especially with the batsmen who all tended to look like guys in equestrian riding helmets. On the other hand with cricket you had five days to figure out who all the players were. Of course it took all of about 30 seconds to figure out who Shane Warne was: that wicketkeeper guy kept yelling out "Aww... Nice one Shyne" in an outrageous accent every time he bowled!
Gee, if I keep this up I'll know more about British sport than I do about American sport. Of course, it's possible that I've already passed this point.
Newly Arrived for the Future New Arrival
Our child's cot arrived today. (Five minutes after 8 AM! Blimey! The Royal Mail sure gets a right early start round here!) In America we'd call this thing a crib, but here in England a crib is a much smaller affair, usually on rockers (what we'd call a cradle in America). One of the places where American English and British English seem to differ a lot is the phenomena and paraphernalia associated with babies. So, being both first-time parents and American expatriates, we're having to catch up and learn two sets of lingo. Mostly we're managing, though there are still things I don't have names for in either language. Like that rubber sucking thing that is useful for removing grapes from up an infant's nose.
Sunshine's been doing a lot of the work. She's good about reading up on things. Her first instinct when facing a new challenge is to try and find a book about it. She grew up as a big library girl. She's gotten better about not bringing home giant backpacks full of books though. At the very least it's must be better for her spine.
Anyway, the arrival of the cot and other baby stuff is triggering another shuffling of the furniture, both literally and metaphorically. Life in our flat is beginning to resemble a 15 puzzle, especially when we're both at home and the laundry rack is out, etc.
As the impending arrival of the boychild gets closer, and more tangible effects materialize, it's definitely beginning to seem a lot more real. This, in turn, is precipitating the 15 puzzle effect in my mind as well. I find myself moving chunks around and reshuffling my thoughts to make room for something new, and presumably quite large as far as my internal brain furniture goes. This is probably a good thing. I think I'm a bit of a mind slob. It seems rather messy and cluttered and disorganized in there. At best sort of a college bachelor pad, but maybe even as bad as a garage or storage shed. Perhaps a nest. A big nest. Like Big Bird's nest. Yes, that's it, I have a Big Bird brain!
Perhaps I should think these metaphors through before committing to them...
The Voyeur of Utter Destruction (As Beauty)
Every once in a while I come across a piece of art, music, a movie, a band, etc., which I think is fantastic, but gets panned by the public and the critics alike. In some cases, my appreciation for the Bruce Willis movie Hudson Hawk comes to mind, I'm not really sure I can justify my stance. (I've only ever met one other person who admits to liking this film, and I won't sully his reputation here.) However sometimes I'm absolutely convinced that I'm right, and that the world must just be peopled with aesthetic morons.
Case in point: David Bowie's fantastic 1995 album Outside. This was actually one of the first David Bowie albums I owned. I think I had his greatest hits collection at the time, and maybe Black Tie, White Noise which I had picked up for cheap somewhere. The latter album hadn't really appealed to me. (I didn't really care for R&B much at the time. I've since grown a fair appreciation for that album as well.) I'm not really sure why I picked it up. I suspect the presence of Brian Eno on the album may have been more of a draw than anything.
Walking back to my dorm room from Tower Records I slipped the disk into my CD Walkman and found myself completely blown away. By the time I reached my room, I'd decided to drop whatever else I was supposed to be doing (probably homework, seeing as classes had started). Instead I sat in my room, dimmed the lights, cranked the stereo up and just soaked it in.
It was amazing. It had all the energy and intensity of the industrial music I'd found myself drifting into (NIN, Ministry, KMFDM, etc), but it had a much more musical quality than those bands tended to. Plus, the production was gorgeous, and there was something about it that just seemed to have that extra spark.
I was hooked. I told all my friends about it, loaned them the album, played it on SKURV Radio. Everyone seemed to come back with sorta lukewarm responses. "Not really my thing," or "Interesting". The critical response wasn't much better. Many ended up outright panning the album. Even my man Stephen Thomas Erlewine of All Music Guide, (someone who I normally agree with quite a lot), gave it a mediocre review, calling it "deeply flawed", and giving it the backhanded complement of being "his best album since Let's Dance." Not exactly saying much. Anyway, needless to say, the album didn't do too well, and the two planned follow-ups (it was supposed to be a trilogy) have never appeared. Bowie went on to make other quite decent albums, and Eno went on being Eno. I went and enjoyed their back catalog, and sadly shook my head at the lost opportunity.
I was reminded of this album recently when Channel 4 showed David Lynch's film Lost Highway. The opening image is of a highway frantically racing past the camera placed inches above the pavement to the strains of Bowie's "I'm Deranged." Having thus whet my appetite just before going to bed, I woke up with a distinct hankering and promptly dumped our Bowie collection onto my iPod before walking to work the next morning.
Hearing Outside again in its entirety for the first time in a couple of years I'm astonished at how well it has held up. Arguably it might even sit better in the modern music scene than it did at the time. In 1995 it was seen as David Bowie's take on industrial music. A notion that was helped in part by the fact that "The Heart's Filthy Lesson" closed the movie Seven which also featured the Nine Inch Nails übersingle "Closer". Bowie's subsequent tour with Reznor and Co. probably reinforced this idea. However, at it's heart, Outside is not really industrial, but rather much more of a prog-rock opus. Heard today, the album brings to mind not NIN and Ministry, but rather post-"Kid A" Radiohead. While there are some well defined songs, much of the album is devoted to more abstract forms. Between the fleshed out songs, the album is peppered with "Seques" and character sketches, and filled with mood pieces that can perhaps be best described as tone poems. Certainly the frenetic energy of the album (and the paranoid cyberpunk subject matter) is grounded in the ethos of the mid-90's, but the music is actually quite a bit more exotic.
Take, for example, the album's first single "The Heart's Filthy Lesson." After a brief pseudo-instrumental and an opening song which sound rather like an orchestral warmup, followed by an overture, this third track sounds like a mission statement for the album. The layered production, samples and driving beat sound rather industrial, but then in the middle the song breaks open in a completely unexpected way with a free-jazz piano solo. In fact, the song (and indeed the whole album) has a loose, organic feel, even occasionally a bit of swing, which is entirely alien to industrial music.
This slide into the unusual is continued in the next track, the haunting tone-poem "A Small Plot of Land." Again, Mike Garson's free-form piano fantasia is the focus at the start, driven by a driving rhythm from Sterling Campbell. The noir-ish soundscape creeps in setting a suitably dark urban scene where the drums echo through deserted streets and piano tinkles onto the pavement like broken glass. Into this dark night, Bowie's voice provides not so much a melody, but rather a howling mantra, like a Greek chorus delivering a funeral dirge.
The narrative (what there is of it) is very much a product of its time. Outside is a concept album, but unlike such 70's works as Tommy, or The Lamb Lies Down on Broadway, the text steeped in the post-modern, non-linear zeitgeist of the mid-90's. Rather than a simple linear plot, we're presented with interior and exterior monologues from a collection of characters, all somehow associated with the brutal art-murder of a 14 year old girl. This metaphor of violence as art is embodied by the characters of the Minotaur, the artist and murderer, and Nathan Alder, the private-eye/art critic investigating (reviewing?) the case. The metaphor is surprisingly apt, as art critics, like detectives, are both engaged in trying to get inside the head of their subjects through their creations.
The effect is not unlike David Lynch's Twin Peaks, but an even more like that ever-so-mid-90's fad, the CD-ROM computer game. Like such classics of that genre, Myst, and the Resident's Freak Show, and Bad Day on the Midway, the plot is presented indirectly. We're presented with the commentary of those surrounding the event, and it's up to us to put the pieces together and imagine the excluded central event. In this respect, perhaps it's better that the trilogy was never completed. It leaves the story nicely ambiguous, like a Twin Peaks where Laura Palmer's killer is never revealed (the way David Lynch originally conceived it).
So, all in all, Outside is certainly not an easy listen. On the other hand, anyone expecting a simple collection of pop-songs from a Bowie/Eno collaboration must not have been paying attention to the content of their trio of albums in the 70's. If you're up for trying something different, I highly recommend it. It must be easy to find cheaply somewhere.
Thursday, October 06, 2005
Tinned Meat Product
Amazing. Still less than a day old (google hasn't even found it yet) and my blog has already managed to attract a spam comment. I guess it's time to turn on the word verification option. That way if someone wants to spam me they at least need to do it by hand.
Found a third-party editor... "ecto"... seems pretty decent. A lot better than the painful web interface at least.
Found a third-party editor... "ecto"... seems pretty decent. A lot better than the painful web interface at least.
Identity
Less than 24 hours existence in the blogosphere and I've already started refining my thoughts on this phenomenon. I suspect that there will be a fair amount of modification to this blog as it starts out. Already the limits of this web-based entry system are beginning to show. I suspect I'll graduate to a desktop client at some point. Meanwhile, the content continues to slowly fall into focus. Anyway, on to the topic of the moment...
So when I started this blog last night, I set the description field to "Musings of an Astronomer." This wasn't a very thoughtful description, but I hadn't put much thought into it. However, the more I ponder this the less satisfactory it seems. The basic problem is that while I am an Astronomer, it's not a particularly accurate reflection of my own sense of self. "Astronomer" is not really who I am, it's just something I do. Now I do this a fair amount of the time, as it is my job, so from an external perspective this might seem a reasonable description of me. On the other hand, I also spend rather a lot of my time sleeping, so this blog could just as validly be described as "Musings of a Sleeper." Again, neither seem accurate descriptions of my own internal sense of self.
I suspect that Ken Wilber would say my selfsystem has evolved beyond the role-based conventional worldview stage. Hence the current description: "Ponderings of a selfsystem", which I guess is somewhat more accurate by virtue of being less defined. Still it seems lacking. I'm not really a fan of the word "Ponderings." This process seems a bit more deliberate than that. I like Descartes' use of the word "Mediation," (ie, "Meditations on First philosophy"). On the other hand "Meditations of a Selfsystem" sounds entirely too portentous and Philosophical with a capital P. Yuck. Perhaps I could borrow from Hamlet: "Words, Words, Words." Still, that wouldn't be completely accurate either. The thing about existing on the web is that it's not limited to just words.
Hmmm... I suspect this to go on for a while....
So when I started this blog last night, I set the description field to "Musings of an Astronomer." This wasn't a very thoughtful description, but I hadn't put much thought into it. However, the more I ponder this the less satisfactory it seems. The basic problem is that while I am an Astronomer, it's not a particularly accurate reflection of my own sense of self. "Astronomer" is not really who I am, it's just something I do. Now I do this a fair amount of the time, as it is my job, so from an external perspective this might seem a reasonable description of me. On the other hand, I also spend rather a lot of my time sleeping, so this blog could just as validly be described as "Musings of a Sleeper." Again, neither seem accurate descriptions of my own internal sense of self.
I suspect that Ken Wilber would say my selfsystem has evolved beyond the role-based conventional worldview stage. Hence the current description: "Ponderings of a selfsystem", which I guess is somewhat more accurate by virtue of being less defined. Still it seems lacking. I'm not really a fan of the word "Ponderings." This process seems a bit more deliberate than that. I like Descartes' use of the word "Mediation," (ie, "Meditations on First philosophy"). On the other hand "Meditations of a Selfsystem" sounds entirely too portentous and Philosophical with a capital P. Yuck. Perhaps I could borrow from Hamlet: "Words, Words, Words." Still, that wouldn't be completely accurate either. The thing about existing on the web is that it's not limited to just words.
Hmmm... I suspect this to go on for a while....
Wednesday, October 05, 2005
First fumbling steps...
Famous first words are a daunting prospect. Better by far to not even try. Just hook up the hose and see what flows out. I guess one could be concerned about pollution, but I can't imagine that these quiet little thoughts will cause too much of a stir in the vast ocean of babble out there. In any case, I can't really be bothered...
So here it is, my first feeble excursion into the blogosphere. It's amazing how fast I've managed to fall behind the techno trendy. I must admit I'd never heard of the term blog until Robert Quimby showed me an early version of his clever little GRBlog. Stupid me, I thought it was supposed to be pronouced "GRB log." Mind you I used to read Andrew Tong's ponderings way back when the web was new, but that just makes me sound like even more of a dinosaur. Really, I'm not that much of a clueless old fogey. I just got distracted a bit. Grad school will do that to you.
Hmm... I can see already that this web interface will leave a bit to be desired. Still, I suspect that someone out there has developed some useful tools. Someone almost always has. This I guess is the advantage to being behind the curve: you don't have to build all the tools. Which is fine... I'm building enough tools already.
The picture is nice... I was considering calling this "Groggy Blog" to match the photo. It's probably been taken already though. Toevening is a nicely obscure word that nobody uses... My only concern is that there is another someone with as much claim to the name. Still, I suspect she won't mind my using it too much... Besides, I'd be willing to share.
Boy, I wonder what this blog is going to be about.
So here it is, my first feeble excursion into the blogosphere. It's amazing how fast I've managed to fall behind the techno trendy. I must admit I'd never heard of the term blog until Robert Quimby showed me an early version of his clever little GRBlog. Stupid me, I thought it was supposed to be pronouced "GRB log." Mind you I used to read Andrew Tong's ponderings way back when the web was new, but that just makes me sound like even more of a dinosaur. Really, I'm not that much of a clueless old fogey. I just got distracted a bit. Grad school will do that to you.
Hmm... I can see already that this web interface will leave a bit to be desired. Still, I suspect that someone out there has developed some useful tools. Someone almost always has. This I guess is the advantage to being behind the curve: you don't have to build all the tools. Which is fine... I'm building enough tools already.
The picture is nice... I was considering calling this "Groggy Blog" to match the photo. It's probably been taken already though. Toevening is a nicely obscure word that nobody uses... My only concern is that there is another someone with as much claim to the name. Still, I suspect she won't mind my using it too much... Besides, I'd be willing to share.
Boy, I wonder what this blog is going to be about.
Subscribe to:
Posts (Atom)
