Photographers: I get it now.

I always knew photography is hard, but I’m starting to have a new appreciation for it. After Lydia went to sleep, we stuck a couple of socks that Kate finished into the light tent, put the camera on a tripod, and took some pictures.

What looks nice and bright to the eye, though, looks dingy and yellow in the camera. I guess this is because the tent is illuminated with hardware-store floodlight bulbs, not color-balanced daylight bulbs, and I don’t see the difference in the real world, because my eyeballs do the correction automatically(?) I did remember the magic words “white balance”, though, pushed “menu” on the camera, and pointed at a little picture of a light bulb (instead of a little sun.) That got most of the yellow cast gone, and we got some pictures that I think look pretty good:


…but even that is only after I opened the raw image in Photoshop, selected the background, desaturated it, and adjusted the white balance to favor the highlights. I don’t think it’s considered totally cricket by crafters to manipulate your photos so much, though, so I’m hoping to find a different combination of light bulbs and white-balance settings that won’t need so much manhandling.

Photographers: I get it now.

Basement light tent

Kate’s at knitting tonight, so I finished the light tent in the basement, using oak dowels, a paper dropcloth, and lots of black duct tape. It turned out to be huge, much bigger than I had anticipated, and brighter, too:

There’s a slit in the front of the tent for the camera’s lens to go into, and a sheet of white oaktag inside. It’s kind of big and unwieldy (in retrospect, I should have made something this size), but it looks pretty good inside. I took this picture of my nerdle-point pillow, in progress, and the even white lighting made it very easy to punch the item off the background:

Now to make this project pay for itself by sticking everything nearby into the tent, taking evenly lit pictures, and selling it on eBay. Like maybe the cat. Here, Squeaky!

Basement light tent

John’s Christmas/birthday wishlist: an XO laptop (one for me, one for a kid)

Originally called the “OLPC” (One Laptop Per Child), or the “$100 laptop”, the XO laptop is designed from the bare metal up to be used by children in developing nations, to bootstrap a worldwide generation of skilled hackers, entrepreneurs, and knowledge workers. The laptops are designed to be low-power, chargeable using a yo-yo like pull charger (no electric grid needed), to connect with each other and the Internet using a wireless grid, and to show the source code of all the running programs in a way that lets kids learn.

You can read more about the project in a New York Times article here. Sure, it’s going to raise lots of issues (what if they get stolen? What about goatse?), but as far as I’m concerned it’s going to open a floodgate of information and enfranchisement. In twenty years, your employability, no matter where in the world you live, is going to be based on two things: knowledge of computers and command of apostrophes and homonyms. Any kid with an XO laptop can start learning both.

The non-profit XO project isn’t set up for consumer sales in the US — their mission is to get governments to buy them a million at a time to distribute to kids. But the governments are hesitant to write those big checks until they can see demonstrated enthusiasm for the devices. So from now until November 26, you can buy an XO laptop to be sent overseas, and get one for yourself. That’s one for a developing nation, and one for us to play with in Sharpless street, so Lydia can join the Worldwide Hacker Army. Here’s a description of the program. So, dear friends and close family — anyone who has me on their Christmas and birthday list — you would make me very happy if you would click on the button below and contribute a little something, you will have made a nerd very, very happy. At the end of the contribution period, I’ll take all the proceeds, contribute the difference, and place the order.

Contribute towards a “Give one, get one” XO laptop for John:

Thanks, and merry Christmas! And happy birthday to me, etc!

John’s Christmas/birthday wishlist: an XO laptop (one for me, one for a kid)


My current needlepoint project is to, you know, create a nice sofa cushion. A pillow. With a two-dimensional barcode design on it. Here’s the canvas, half-transferred from my printout using black acrylic paint:

The design is in a machine-readable format called QR Code; codes like this can be found on your UPS package or pharmaceutical label. 2D barcodes can contain all kinds of information, not just numbers.

The QR code stitched into the pillow contains an encoded hyperlink to the Wikipedia entry for “Pillow”, so if you’re a Japanese teenager with a DoCoMo QRcode-enabled cellphone, you could snap a picture of the sofa pillow and immediately, you know, start reading about pillows. Here’s a picture with some of the black yarn stitched in:

This is the first time I’ve tried painting a canvas, not just marking the intersections with a pigma pen. So far, I’m learning that needlepoint has three phases, and that the one where you just stitch the yarn into the canvas is by far the easiest one. The first, prepping-the-canvas phase, requires some cognitive sleight-of-hand (mapping pixel-shaped blocks to intersections, then painting the intersections, is trickier than I would have thought.) And the third phase, finishing requires all the tools of a carpenter and the black arts of an upholsterer. I haven’t tackled that one yet.

So far, though, I’m having a blast. I’ll post more pictures when I get an area that has both black and white stitched into it.

PS. In case you’re wondering, here’s how I made the design:

  1. Went to and used their online tool to create a PDF of the Wikipedia “Pillow” URL.
  2. Made a screenshot of the resulting PDF, opened the screenshot in Photoshop, and scaled so that the smallest box unit in the semacode was exactly 2×2 pixels. I used “nearest neighbor” scaling to preserve hard edges. This resulted in a design that was 58px by 58px.
  3. Used the awesome KnitPro Web App to transfer the design onto a numbered grid. Before I uploaded the image, I increased the image’s canvas size to 96px wide by 120px tall, matching one of KnitPro’s existing sizes, so that KnitPro wouldn’t scale or antialias the design.
  4. Used a fine-point red pigma pen to break the numbered grid into 8×8 boxes, and did the same to the canvas.
  5. Carefully transferred one 8×8 box at a time to the canvas, painting the intersections of the threads to correspond to the black boxes on the numbered grid. When I messed up, went back and painted with white paint.
  6. Trimmed the canvas to size, finished the edges with masking tape, tacked it to a frame, and got stiching!