Statcounter and random stuff

So remember how I said I'd gotten 208 unique visitors to date on Tuesday? It's more than 275 now. Blue is the unique visitor count. Check it out for the entirety of this blog's existence:

I use StatCounter. It's free, easy to use (add an html element to your blog) and has decent visuals. How accurate is it? I don't know. I recently started using FeedBurner (when Google bought FB and it became easy to use with blogger) which can do some site statistics too. Hopefully comparing them will be interesting.
---
Last night at around 01:00 I was doing some reading drills and had a monstrous headache. I thought "Maybe I should do an extra nap tomorrow..." At the end of the day though, I think it was a blood sugar thing. I ate some pie and felt much better and much more awake.
---
Evidently the Stop & Shop on 9 is open 24 hours. I'm also told it's closed Sunday. But seriously. 24 hours? Around here? Hooray places that are open after 20:00!
---
As part of a NASA project, I'm using a PIC as a prototype circuit. I got to be all nostalgic and literally run through a good portion of the labs we did for POE. Awesome. For my money, the hardest thing to do with a PIC is get your computer to believe it's actually plugged in.

vvvv

Upon a teacher's suggestion, I downloaded vvvv to help with some image processing I needed to do. It's a visual programming language that's pretty awesome at a/v stuff. I'm just going to toss in some pics instead of ranting for too long... I should note that it's kind of an annoying program and that most of my, uhmm, code(?) is just variations on code that others put online.




This program is hot. The parts of the image that look a bit messed up are because the camera is out of focus and has been substantially tampered with physically. Now this isn't actually what I need I think, but it's close. I need to differentiate color not brightness. This might be easier in the YUV colorspace instead of RGB... I need to look at if vvvv has support for such a transform and I need to look at the code of the effect node. The effect node is the one that has a bunch of inputs in the bottom left. It's the one that sets the brightness threshold and outputs the appropriate color. It's evidently the only node that you can easily get to the source for (a .fx file).

Sweet.

In case anyone wants to replicate this: I used this main patch, this effect file and this effect patch. I'm pretty sure that the second patch is a demo you don't actually need, but I didn't want to check and I wanted it available if it was needed.

This is interesting... I can't seem to title this post... I'll see if it works from a different browser after saving a draft...

Anyhow. Today was pretty sweet. Some of us were laser-cutting parts that we need for NASA out of thin plastic film for quite a long time. The way the machine works is rather neat; it just burns whatever lines you give it. One sets the speed it goes at and the power it uses. That's it. Then the laser goes along burning what you told it to burn and the ventilations system attempts to pump out all of the toxic fumes you generate. It doesn't fully succeed, but I'm going to go ahead and imagine that the smell would've been far worse if it weren't for the ventilation system.

So anyhow, one of the sheets of plastic we were cutting was really thin. I mean we had one that was five thousandths of an inch thick; that was the fattest one. The small one was to thousandths of an inch thick. We set power and speed lower for the thin piece of Kapton and then we started up the laser. It went along the bottom edge of the shape we were cutting and then promptly proceeded to miss the Kapton entirely.

This wasn't the laser's fault; it was positioning itself perfectly. The issue at hand was that the piece of plastic was really thin. Thin enough, in fact, to get sucked into the ventilation system. We had to shut everything down.

Eventually, everything was fixed and the now-wrinkled piece of Kapton was recovered from the ventilation system. Hooray!

...Something's being silly. I can't seem to get into my account from Opera. It wasnts me to switch my blogger account to a Google account. Oh yeah. It's always been a Google account. wtf.

OK. I got in. (I needed to tack on the @gmail.com). When I look at my post in Opera I can access the test box for the title, but I can't touch the bod. Also what I'd already written disappears. How odd. My hand is force. I will use the scourge of worlds: IE.

AH good. IE can't access the text box for the title either. Whatever. This shall henceforth be known as "The Post With No Name."

FPGAs

These things are sweet. Like, really sweet. The idea is that you write a program and the FPGA implements it in hardware (it makes a digital circuit with a portion of its hundreds of thousands of gates).

The parts I like are the cost and how much they jam on the board. All the stuff on the board can be used as either inputs or outputs (depending on what they are); also there's some mode-controlling stuff on the board, but let's ignore that for now.

I'll describe one that's just shy of $100. It has 8 LEDs, 8 switches, 4 push-buttons on the board itself. Then it has an additional 60 I/O ports that can be used for whatever. Some of these (24) are split off into groups of 6 pins that you can plug something called a Pmod into. These are extensions that they've already made for you such as video I/O or audio I/O or motor control etc. Oh yeah. It also has a 4 digit seven-segment display. This lets you make the simple squarish looking digital clock type numbers. And there's a slot for a 1/8 VGA LCD screen. Sexy.

Anyhow. I keep on coming up with reasons to buy one that aren't good enough to get me to do it. Case in point: for about $200 I can make a TV card for my monitor. I mean. OK, I could play Smash on my monitor. Or I could get a $50 TV card. But the FPGA can also do other stuff. But for $200 I could buy a new TV. *sigh* I need better arguments for myself. If anyone can come up with a compelling reason for me to get an FPGA other than their shininess, that'd be awesome.

I'm pretty sure this device has made me the happiest and most victorious-feeling for the most mundane tasks. For example, I felt accomplished after following a step-by-step screencast to make the FPGA count in binary and display the 8 MSB in LED form. And now, I'm failing to use the seven-segment display to make it count with numbers. Actually, I think I've gotten it to do it, but too fast for my eyes to resolve... I'll work on that.

NASA projects

So I now know what I'm working on.

I'll be on a team that's making an x-ray flux meter. This will pretty much count photons per unit time above some energy threshold. Ideally, we'd be able to figure out the energy of each incoming photon so that we could build up a spectrogram.

The other team I'm on is actually doing two projects. The first involves making a cheap x-ray imaging device by either hacking the software from a webcam or making our own camera. The second is something I've taken to calling ISPS - Inter-Stellar Positioning system. This is a lot like GPS except it can tell you where you are in space; it's somewhat harder than GPS b/c there's an extra dimension and we don't have satellites anywhere except around Earth. The idea is to use millisecond pulsars (neutron stars that send out x-rays at an extraordinarily steady interval) instead of the satellites. We're not actually even sure if this can be done, but hopefully it can and hopefully we can put together some software that could do it.

Anyhow, I'm super-excited. W00t!

Cool stuff from NASA

As some of you may know, I'm working for NASA at Olin this summer. On Friday we went down to Goddard Space Flight Center to talk to some people and get some projects to do for the summer. There's some cool stuff; check it out:

We're working with the x-ray group so you might see some obvious correlation...

-Interstellar GPS: Millisecond pulsars that emit x-rays at extremely constant intervals could be used to determine your position in space. We would study how to implement such a system and perhaps do some of the hardware/software too.

-Mounting Mirrors: X-ray mirrors need to get in space without breaking; this one's a lot of FEA analysis and stuff.

-Hacking a webcam: Hacking the software from a cheap webcam to get crude x-ray images on the cheap.

-Flux Meter: Make a hand-held USB peripheral that can measure x-ray flux.

-3d XRD study: NASA recently developed a 3d x-ray diffractometer (used for identifying minerals), but they don't currently know how best use the data. This one's sifting through data and figuring out how much useful information we can extract and how.

Those are our projects; I'll be working on two of them and will know which tomorrow.

On a completely separate note, I got to see something cool when our plane was breaking. You could see air resistance in action in the form of water vapor. So cool!