This is what happens when I don’t follow camera news every day, apparently.

I’ve been ogling the Nikon D3 (and its smaller sibling, the D700) for a long time. They have lots of neat features, but the big one for me? ISO 25,600. That’s off the charts.

Canon finally did what I hoped they’d do, and matched it with the newly-introduced Canon EOS 5D Mark II. ISO 100 to 3200 “standard,” but you can unlock an extended range of 50-25,600, matching Nikon’s. (And Nikon’s is similar: 25,600 is a mode you can unlock, but that isn’t there standard, since it’s not perfect.) I’m yet to see any sample shots from it, so I don’t know how “usable” it is: my camera will go to ISO 1600, but it’s a little too grainy for my liking, so I consider ISO 800 to be the upper limit, with ISO 1600 as a last resort.

It’s also a full-frame sensor: most digital SLRs have a smaller-than-normal sensor, so they can only “see” the center of the image, which has the effect of cropping the image. (This is a nice “bonus” if you’re working with telephoto lenses, perhaps, but it’s annoying with wide-angle lenses.)

It’s got an absurd 21 megapixel resolution. I keep my 10-megapixel camera set to shoot at 5 megapixels, which is plenty big for me. Perhaps more interestingly, after adding Live View functionality (the ability to view a ‘preview’ of the image on the camera’s LCD, something normally impossible on an SLR), Canon joined Nikon in going for the next step: adding video capability. And this is no “consumer” point-and-shoot in terms of video: as you might hope a $2,800, 21-megapixel camera would provide, the 5D Mark II can record MPEG videos in full 1080p, 1920×1080 resolution, at 30 frames per second. (You may want to pick up some of SanDisk’s news 32GB CF cards before trying this.)

In addition to those major new features, the camera’s been cleaned up around the edges, introducing several features that I think will make a big difference. For some reason, Canon is notorious for putting crappy LCDs on their SLRs. The LCD on my camera is a huge step up from previous ones, and it makes me laugh to think about the postage stamp of an LCD on my old 10D. And yet it’s still not that great of a screen, given that I’m using it to try to judge minor details. (Is the whole subject in focus? Is the focus just right? Is it overexposed slightly?) Nikon’s been whooping Canon here for a long time, and, perhaps most irritating to professionals using Canon SLRs, Canon’s lower-tier point-and-shoot digicams have used nicer LCDs than their SLRs. Not anymore. The 5D Mark II comes with a 3″, 640×480 LCD. They’ve also apparently used some sort of anti-glare coating so that the LCD is finally useful outdoors. And they went a step further, and set the LCD’s brightness to match ambient lighting, so that it will automatically dim in dark rooms and brighten outdoors.

They also introduced Auto ISO. This is a neat feature to me, as it works in all modes except full manual. In general, I don’t care much about ISO. I want it as low as possible (to minimize noise), but I really couldn’t care less whether I shoot at ISO 400 or ISO 200. I care a lot about aperture, as it influences depth of field, and I also care a lot about shutter speed, which determines exactly what the scene will look like. So I usually shoot in aperture priority, where I set the aperture, and the camera chooses the optimal shutter speed. And all the time, I’ll adjust the ISO to suit, but it requires going into a menu to change. In its full-auto modes, the XTi has Auto ISO, but it’s not terribly useful. (ISO 100-400, and no one uses the full-auto modes anyway.) Now, it seems you can let the camera worry about ISO; it sounds like it’ll automatically set it as low as reasonable, abiding by the 1/focal-length rule. (The current ISO setting is also shown in the viewfinder; helpful for anyone who’s ever fired off a burst of shots only to realize that you’d forgotten to change ISO earlier, leaving you with a batch of horribly mis-exposed shots.)

“Lens peripheral illumination correction” was a term I had to look up. It’s an in-camera option to correct for vignetting. Neat!

One final thing I find interesting: the “Live View” mode already lifts the mirror, so they were able to take that a step further and introduce a “quiet” mode. With current SLRs, when you press the shutter, the mirror flips up, and then the shutter opens and closes, then the mirror goes back down. While there’s something satisfying about the tactile feedback, it can be a nuisance, especially if you’re trying to take photos in a really quiet place.

Overall, I’m quite impressed! I’m eager to see some sample shots at higher ISOs, and I’m also eager to be able to afford a $2,800 camera. 😉

I got some decent photos while I was in Amherst yesterday:

I be telling you maties, Pootie-Poot be off his rocker!

I have policyd set up to blacklist any IP that attempts to send mail to any of my “spamtrap” addresses (I have almost 30 e-mail addresses that regularly receive spam, but aren’t used by a single human, so instead of just bouncing the mail as undeliverable, I let them send the mail, blacklist their IP, and then refuse it.) Entries last for 14 days in the database, so that people aren’t penalized for months-old violations. (I actually think 14 days might be too long, but I’m in no rush to lower it.)

I set up a page that lists the 100 most-recently blacklisted IPs and also uses GeoIP to try to determine their country.

But then I got another neat idea. I’d heard about p0f a few times before, but never played around with it. In essence, it’s able to look at a TCP session and (most of the time) determine the OS in use, via passive fingerprinting. Unlike nmap, it doesn’t go out and try a thousand different probes. Just by examining the structure of the packets, it’s able to make its guess.

So I set it up to listen on the server’s NIC, and log its OS guess for incoming connections. And now my list of recently-blacklisted spam IPs shows the OS they were identified as having.

What’s interesting to me is that most of them run old versions of Windows: “Windows 2000 SP4, XP SP1” is especially common. And I have a hunch for why: it seems that botnets are seeing increased use for sending spam. That is, most of the spam I’m receiving is being delivered via old, unpatched Windows machines that have been infected with a virus, which spammers use to pump out spam. (Hence Spamhaus’ PBL, a blacklist of IPs that should be peoples’ home IPs, not mailservers.)

I also set up a p0f script to show you what p0f shows for you, when you connect to that page. It also prints out the various headers your browser sends, so that you can see just what websites can see.

All of this reminded me of OpenBSD’s pf firewall, and its scrub options, to try to “clean up” packets. Thus I show up as:

Windows XP/2000 (RFC1323, w+, no tstamp) [GENERIC] Signature: [8192:115:1:52:M1460,N,W2,N,N,S:.:Windows:?] -> 64.191.108.120:80 (distance 13, link: ethernet/modem)

This is actually still fairly accurate, but I’m actually using Vista. OpenBSD seems to be stripping some of the data out, though, keeping p0f from showing too much.

I’m posting this in the hopes that it’ll help someone who finds this via Google…

ziproxy, the compressing HTTP proxy, is no longer a Debian package, for reasons I haven’t seen explained. I found this out today when I went to install it, hoping to speed up web browsing from my cell phone.

I downloaded the latest beta from SourceForge, but the ./configure wouldn’t complete, spitting out all sorts of errors about not being able to find libgif or libungif… Here are the etch packages you’ll need:

• flex
• libungif4g-dev
• libpng12-dev
• libjpeg62-dev

The key is that you need the -dev versions, which I didn’t download at first. With those in place, configuration was a breeze. Then it’s just a ./configure; make && make install, and then you’re off to move /etc/ziproxy/ziproxy.conf.dpkg-new to /etc/ziproxy/ziproxy.conf, set it up as desired, and launch!

I’m moving some music between computers, and had to reset the password for an iTunes account, because no one had any idea what it was.

The name of the page on Apple’s site to reset it?

iForgot.

I use the history in my browser often, to find sites I’ve visited. It’s precisely because I use it so often that it annoys me when it grows large. Especially as I moved my site around, I’m seeing lots of pages that no longer exist pop up in the history when I go to visit a site. I’d like to purge the old ones.

Tools -> Options -> Privacy has a setting, “Keep my history for at least __ days,” but note that “at least.” Mine’s gone way past the 7 days I have set.

By entering about:config, I discovered something: that’s called browser.history_expire_days_min, which I’d set to 7. But there’s also a browser.history_expire_days, still at 180. (Along with a browser.history_expire_days.mirror)

It turns out that the _min tells Firefox when it can delete an entry, but it won’t unless it’s short on space. Firefox 3 apparently makes this automatic, but wasn’t designed for people with a severe case of OCD who want things to age out after a while.

Setting history_expire_days (and history_expire_days.mirror) to 14 solved the problem: now things get saved for at least 7 days, but not to exceed 14. No more 6-month-old history entries!

Edit: A lot of people are searching for history_expire_days.mirror, which is fairly cryptic. This portion of the source code was indexed by Google*, and defines browser.history_expire_days.mirror as “a preference whose value mirrors that of browser.history_expire_days, to make the “days of history” checkbox easier to code.” So just set it to the same as history_expire_days if you’re changing it!

* It’s open source, so it’s not like it’s hard to find the source code.

I forget who initially mentioned the concept to me, but I’ve bumped into the concept of Lomo cameras a few times now. As I understand it (which is to say, not completely!), they’re essentially cameras which produce images which are a bit less than perfect: they tend to have strong vignetting, excessive contrast, and so forth. But all of these imperfections combine to leave a fairly neat effect.

I stumbled across a few interesting tutorials, one on how to emulate the Lomo look in Photoshop, and another on introducing a glow effect in Photoshop.

As a bit of a perfectionist, I didn’t like the idea at first. I’ve always been careful to get everything just right: you want to eliminate vignetting, maximize the dynamic range, get contrast and saturation “right,” but not too high, and so on. But the post processing has a really charming look. “Timeless” might be a good way to describe them.

I found that vignetting can actually be a helpful effect, pulling your eyes into the center of the frame by ensuring that there’s nothing to look at on the edges. It has a neat side effect of making whatever’s in the center of the frame brighter:

I think it’s fair to say that it makes the photos look “old” in a way, too:

You can view my complete collection of Lomo-ized photos here… I think this collection may grow over time.

One thing about photography, at least for me, is that the whole process is enjoyable. Sometimes the pictures I take end up being total rubbish, but I enjoyed taking them anyway, so I’m not too upset.

Another thing about photograpy is that it’s very much a skill. The equipment you use does matter somewhat, and there’s surely such a thing as “having an eye” for a good composition. But the majority of what goes into making a good photograph, I think, is practice: the skills you’ve learned and the techniques you’ve practiced.

It was a gorgeous day outside this afternoon, so I decided to have lunch outside. And I kept spotting photo opportunities, so I went in and grabbed my camera.

I decided to practice three things while traipsing through my back yard this afternoon:

• Using a polarizer
• Using fill flash
• Manual focusing

A polarizer on a camera is really no different than wearing polarized sunglasses. It does a couple neat things: one of the more obvious ones is cutting glare. A slightly less obvious one is that it boosts contrast. The problem with a polarizer is that almost everyone uses a circular polarizer, which means that it has to be turned to just the right spot to accomplish anything. And most lenses (save the ones with four-digit price tags) focus by rotating the front, which means that as soon as you recompose a shot, the polarizer is going to rotate. So what I wanted to work on a bit was the ability to quickly determine the best settings for the polarizer, along with practicing keeping a couple fingers on the edge of the polarizer while focusing, to prevent it from spinning, so I don’t have to readjust it each time.

This isn’t a good shot any way you spin it (no pun intended), but here are some comparison shots of a rosebush in the back yard:

The lighting was about the same in both cases (full sun shining down), but in one case, I had the polarizer adjusted correctly, and in the other it was wasn’t. (Which accomplished about the same thing as not using one.)

The shiny rose leaves reflect sunlight back into the lens, causing the leaves to have really white reflection/glare. And the bright sunlight being reflected back also causes the darker portions to be blown out. (Plus, both shots are underexposed a bit, but that’s something I would have cleaned up in Photoshop, except that these are unprocessed shots.) By aligning the polarizer right, it blocks most of the glare from entering the lens in the second shot. The shot ends up looking much more evenly-lit, without the absurdly “contrasty” effect of the first one. As a bonus, the colors “pop” a lot more. Granted, it’s still an underexposed shot of a rosebush without any flowers; it’s hardly a good shot either way. But the second one is enormously better.

I tend to hate using the flash on the camera. 95% of the time when it’s used, it ends up dominating the shot, setting the color temperature its own way, creating awful shadows everywhere, giving everyone red eye, and overexposing things nearby while underexposing the background. One thing I love about my 50mm f/1.8 lens is that I can generally use it in indoors instead of flash. (And when I do use flash, it’s typically bounced off the ceiling.)

But it turns out that the flash is actually good for something: “fill flash” is using the flash to “fill in” the shadows of a shot. Taking a picture of someone under a tree? It’s a great shot, but it’s a little dark. The key is that you dial your flash way down, so when it fires, it’ll spill a little more light on the subject, but is hardly as overpowering as it would normally be. This is something I don’t use often, so I wanted to practice a bit more.

This birdhouse is hanging on a tree, leaving it somewhat in the shade. And as you can see, it’s got sun spilling through on the ‘back’ (rightmost) part of it, which makes the rest of the shot appear even darker. (I couldn’t help but clean both of these shots up a bit in Photoshop, so this one ended up looking semi-decent. Straight out of the camera, it was way too dark on the side facing us.)

Here it is again, but with fill flash used. This honestly isn’t a great example, because it ended up being too strong. I can tell that flash was used, which takes away from it a bit. I should have dialed the flash down even more. Here it’s bright enough to overtake the nice natural light. In an ideal world, the use of fill flash would fall somewhere in between this shot and the “original” above it.

Fill flash has a lot of challenges, though. (One of them is the one I mentioned above: getting the level just right!) Another is that, if you’re not really, really careful, you end up with the same problem as using normal flash: you get awful shadows.

This shot was somewhat of an experiment, because it was in a shaded area, and it would have taken something like ISO800 to get a halfway decent exposure, but the lighting would still have been really plain. I used ISO400 and dialed down the flash’s strength. (On the XTi, it’s in the menus as “Flash exp. comp.,” for “Flash exposure compensation, and it regulates the on-board flash.)

Besides the fact that a photograph of a mushroom with a… what is that thing?… wouldn’t be a great shot anyway, it’s very obvious that a flash was used on this one. For one, it bounces off of the shiny thing right back into the lens. It overexposed the stem of the mushroom a bit. But most frustrating for me, it left a harsh shadow underneath the top of the mushroom. It comes across as very amateurish.

Another thing about flash is that most cameras have a flash sync of 1/200 second, which means that you can’t use a faster exposure. One place I played with using fill flash was in intense sunlight, to try to minimize the stark contrast between light and dark.

That, for example. (Another shot that really didn’t come out as well as it could have: it looks a little “overprocessed” to me, even though I didn’t do much!) I had the flash fire as low as it could go there, just to try to minimize any shadows on the rose, since the sun was coming in from the side and was quite strong. (Plus, by throwing more light on the foreground, the background is darker by comparison: note that everything that isn’t the rosebush is essentially black.)

Yet another thing is that ISO matters a lot. Normally, I try to keep ISO as low as I can, but it doesn’t make a big difference, except for noise. If I get a 1/30 second exposure at ISO100, I might bump up to ISO400 to make sure it’s got a better shutter speed. (Since a 1/30 second exposure will probably come out blurred.) Especially when you can’t set your shutter speed any faster than 1/200 of a second with flash, you want to play with ISO.

But remember that ISO is basically how sensitive to light the sensor is. While the flash is of a constant strength in terms of actual light output, raising ISO effectively raises the strength of the flash, by making the sensor more sensitive to the increased light. Thus it’s important to try to get ISO right from the start, since changing it will most likely leave you having to play with flash strength again.

The third thing? 99% of the time, I let the camera autofocus. It’s faster and more precise. But every now and then, autofocus just doesn’t want to work. Consider this shot:

I wanted the camera to focus on the dragonfly on the big blade of grass closest to the camera… No, not the ones behind it. No no, not the lawn way in the back. You can manually set with autofocus sensor the camera uses (and, indeed, this is often helpful in tricky situations), but sometimes that’s not enough. What I’ve been working on doing is trying to get good at smoothly transitioning to autofocus mode. In the case of that dragonfly, it was jumping around a lot, so I didn’t have the time to manually select an autofocus point. The location of the auto/manual-focus switch seems pretty standardized on lenses, at least on the three I use: it happens to be right where my left hand’s thumb rests. (And right in front of the lens release, too, but I’m yet to make that mistake!) So I can flick the switch and rotate the lens until the image is nice and sharp. (You’ll want to make sure the diopter is set just right, since you’re going to be focusing based on what “looks good.” The easiest way is to use autofocus, and focus on something stationary and well-lit, with lots of contrast. And then, while looking through the viewfinder, play with the inconveniently-small knob next to the viewfinder until what you’re looking at appears perfectly sharp through the viewfinder.) It didn’t take me long at all until I could get the image most of the way there, but the hard part is getting it precise. There’s a region where it’s obviously out of focus, a “sweet spot,” and then it’s out of focus again. (Front- and back-focused.) But withing that sweet spot, there doesn’t seem to be a big difference. This is why you want your diopter perfect, and this is why you want to practice. Indeed, I’ve still got to practice more to get tack sharp, but I can at least reliably get pretty close. But I’m going not only for accuracy, but also for speed: if it takes me a while to get manual focus just right, I might as well wrestle with getting autofocus to work.

Sorry for what turned out to be an unexpectedly long post. But hopefully, it’ll be an inspiration: practicing various techniques is not only instrumental to becoming a better photographer, but it’s a lot of fun!

(Oh, and I’m 17 shots away from hitting my 9,000th frame on this camera… That works out to around 1,000 shots a month.)

I’ve never really supported McCain, but most of the election, I’ve thought that he’d be a big improvement over Bush, and a big improvement over any of the other Republicans. I felt that under McCain, we’d be a little bit better-off, but that under Obama, we’d be even more better-off. So I had a strong preference for Obama, but couldn’t get myself to be too bothered by McCain.

But a lot of stuff has caused me to swing from McCain-agnostic to anti-McCain.

• Sarah Palin. I liked her at first. She seems like a good governor, took on her own party over corruption, and was against the Bridge to Nowhere. Except that it turns out that she, herself, is under investigation for abuse of power as governor; she billed the state a per-diem for time spent living in her own house (to the tune of $40,000). And the Bridge to Nowhere I applauded her for opposing? It turns out she campaigned for the bridge, and only scrapped plans when the Feds decided to give them less money. Oh, Alaska kept the Federal money, spending it on other things. All these little things about her lead me to wonder: why did McCain pick her? What does it say about his judgment? Will he try to appoint Harriet Myers to the Supreme Court again? (For a little bit of humor, a Daily Show clip about Sarah Palin and “the gender card.”)
• McCain’s not being able to say how many houses he owned kind of blew over in the media, but left me somewhat confused. How do you not know that? For someone combatting images that he’s wealthy and out of touch, and combatting smears that he’s old and senile, not knowing how many houses you own was a big misstep. The Washington Post has the answer, by the way.
• Remember the nasty ’04 campaign and Karl Rove? Guess who’s working behind the scenes on the McCain campaign?
• His brusque Time Magazine interview.
• His recent ads. This is what shifted me from, “He wouldn’t be a good pick” to, “I actively oppose John McCain.” It’s the norm in politics to tell half-truths about your opponent. But McCain’s latest ads, instead of being half-truths, are truth-free. I posted a few days ago about them, in which he asserts that Obama voted to teach sex in kindergarten. He voted for a bill to teach kindergartners about avoiding sexual predators. And then he sent an e-mail out to everyone on his campaign mailing list (including me), starting, “Friends, You’ve surely seen the shameful attacks Senator Obama and his liberal allies have launched against our vice presidential nominee, Governor Sarah Palin.” Who’s calling who disgraceful? He also keeps talking about Obama’s plans to raise taxes, when a cornerstone of Obama’s plan is actually to cut taxes on the middle class. It’s defined pretty clearly on his website. And yet McCain persists in talking about how Obama’s going to hurt the middle class. Then Obama criticized the McCain-Palin message of change, saying, “You can’t put lipstick on a pig.” McCain went around lambasting Obama for insensitively calling Palin a pig. Except that Obama wasn’t actually doing that, at all.

It ended up blowing up in his face during his recent interview on The View. A recent New York Times op-ed calls him out on the same thing. And it seems that a lot of other news publications that are much less liberal than the New York Times have followed suit. And the Obama campaign is getting good at denouncing the attacks and injecting just a little bit of snark, like their recent statement, “In running the sleaziest campaign since South Carolina in 2000 and standing by completely debunked lies on national television, it’s clear that John McCain would rather lose his integrity than lose an election.”

I used to favor Obama but not really mind McCain. But I think McCain has gone too far—way too far—and I’m left not only rooting for Obama, but rooting against McCain. I shied away from the “McCain = Third Bush Term” people. McCain was different, I thought. I didn’t agree with him on all the issues, but he was a lot better than Bush. He had some integrity. At least, that’s what I used to think.