Slashdot Mirror
Homemade Digital Cameras
Michael Golembewski writes "For the past three years, I've been taking apart cheap secondhand flatbed scanners and turning them into homemade large format digital cameras. They are well over 100 mexapixel in resolution, and produce results that are both similar to and significantly different from traditional digital and conventional cameras."
That is some eerie art! Is your initial part of your last name really pronounced "Golem" by chance?
Very cool effects. When I read the snippet I figured this was going to be something like the old "Make an E-size scanner out of any hand scanner" fraud that was popular for a few years back in the old days (remember stitching manually on a Pentium 200, anyone?).
For some reason I can't believe this works. I figured the scanning element (CCD) needed an intense amount of light to properly "read" an image on the bed.
The fact that you use duct tape to get everything "light tight" put a good smile on my face, as well as the fact that you even got this working. If you're thinking of selling artwork, I'll be the first in line (the lady and I realized it's time for more photo-prints in the house). By the way, the image taken of the actual camera doesn't seem very high res. Was this by choice?
Opened a path to new computer technologies and related devices
This has got to be one of the more fun geek mod projects. I'm definately assembling one. Got to keep an eye out on Ebay for a suitable camera. A great new use for my notebook. I'd love to try some shots at a carnivale. The spinning rides would make for some interesting shots. I'd love to try some high res landscapes though. Could also do some stunning macro photography as well.
Disable the motor (maybe just remove the carriage), and set it up to take pictures of things going past it. Cameras like this are used at finish lines at athletic meets. Interesting distortion. Might be an interesting project for someone still at school - I was once around the teachers trying to work out who won a 100m dash: some of them were a little bald(er) by the end of the day.
Prediction for end of Universe #42: Fencepost error in Quantum_bogosort.cpp
I find it hard to believe that he can get 115 megapixels out of that scanner. Since he is using a 4x5 camera, that works out to a scanner resolution of 2400dpi. That is the kind of resolution of high-end film scanners, not a cheap flatbed (whatever the marketing material says).
lhk
Nice. My first thought was that it would be just one of the many how-to-make-your-own-digital-camera articles that have been around for ages at Make magazine and other similar media... so a lot nicer and more creative than that. And now I want to find an old, old camera, with an old flatbed scanner and try some of that stuff myself.
Back in the early 90's Dicomed introduced a digital camera back for 4x5 cameras for commercial use. It used the same theory, slap a scanner on the back of the camera. You had to use hot lamps, a studio environment and nothing could move. But the results were stunning, 100MB files that could easily be printed on the cover of a 150 line screen magazine. Wonderful.. What is really cool about your work is that you are celebrating all of the motion artifacts that studios were killing themselves to remove when using this technology. Your stuff is pretty creative and fun, and most importantly makes one look at the world differently.
Interesting project, it reminds me of the digital camera I built with a fellow student as a degree project. We ground the top of a 256 byte DRAM using a grinding machine in the mech.eng lab and fitted a glass window. The DRAM capacitors discharge at a different rate when exposed to light. We mounted the chip to a PCB, cut the back out of a 35mm Zenith camera and mounted the PCB. Obviously the optics and chip were poorly matched, we were only using a small part of the lens.
Knowing the discharge rate of the DRAM and the time to load and scan all 256 elements you could get a black and white image. We used the camera for some image recognition work. One application was counting the number of cups remaining in a drinks machine hopper by edge detecting the image then counting the "lips" that we saw.
That was back in the autumn of 1986. We've come a long way.
The scanner software that comes with the scanner he's presently using shuts the thing down if there are hardware faults. All his mods count as hardware faults thus making the shipped driver useless to him. He discusses a closed source pro driver which is a bit better, but still not perfect for his needs. Then explains how he uses SANE to make the thing actually work like he wants.
That's cool -- an artist embarks on getting enough programming language to modify a program so he can use it like he wants to. That's owning your hardware in the purest sense. And it's made possible by the community that generates all that great open source software.
What changed under Obama? Nothing Good
You mean like Technicolor? Or like Autochrome? Or like three-CCD analog camcorders, digital cameras, and digital camcorders?
Pshaw. Sergei Mikhailovich Prokudin-Gorskii did it first.
Laws do not persuade just because they threaten. --Seneca
This seems to be actually repeating early developments in professional digital photography. The first digital backs for cameras like the Sinar, and the Arca Suisse were miniaturized flatbed scanners like that. Obviously they were really good only for still life. But still (back in 1992 or so) when I was in photographers school and we visited someone who had one of those backs, we managed a portrait of someone sitting very still. There were little smears where his breathing caused motion.
Sadly we did not experiment with more motion. I think the "experimenting" with motion is the interesting part (as far as photography is concerned). Some of the pictures on the site are enjoyable. Hacking it all together yourself is interesting too, at least for us geeks.
As for the comments in the style of "large format photography is only about the image quality"... it isn't exactly only about that. It is also about stuff like parallax control (putting buildings "upright" with parallel lines) and depth of field control (laying the plane of the depth of field folded through the scene in order to allow image to be sharp on other areas). All this can theoretically be achieved even with smaller formats, but due to mechanics it gets harder the smaller the format (Arca Suisse's 6x9cm cameras seem to be the smallest that still work very well, at least in my experience).
Therefore the "experiments" done with this hack to in a line a bit with stuff like putting ordinary photographic paper into a large format camera or using polaroids for transfer prints. The "long exposure" part of it is also a reference to the times way back, when due to old processes like the daguerreotype, portrait subjects were held up with wire constructions. Very cool, all of this hack, congratulations.
Heh, all of the pics I saw on the site were scaled down considerably, apparently enough to survive a ./ing, for now. It'd be neat to see a few of these images at full resolution, though. Maybe a bittorrent...
I remember reading about a similar attempt using EPROMs in late '80s. In a normal recycling procedure, these were erased by shining a high intensity UV light for a couple of minutes through the glass window that's present in the IC shielding. After making sure all bits are reset, the window is covered by a sticker or label. Then the data is 'burned' and the (typically 256Kb/32KB) ROM was inserted into your BBC micro (insert favorite hobby computer).
Similar to the process you describe, an image could be gathered by setting all bits to 1 (or actually 0, as EPROMs have negative logic), then waiting for bits to flip. I don't think a lense was used, just a diafragm (dot in piece of black paper). On average, an exposure took 1-4 hours (and resulted in a very poor quality black/white image for current standards) if the scene was bright enough. Lighting the scene with a UV helped a bit, though visible light worked surprisingly well.
Of course, the whole whing was hardly useful, but remember that - at the time - the average micro had little more periferals than a keyboard, joystick. Mac had a mouse and some MSX systems had support for an analog "gen-locked" camera, BBC had some support for light pens. There may have been handscanners for PC/ATs. So mostly any hardware that would allow translation from analog to digital was greedily welcomed.
This is somehow similar to mobile phone cameras. Some of them are very slow too:
m l
http://dimss.solutions.lv/samsung-phone-camera.ht
Dude, you should let me be your slashdot dupe checker. I'm pretty confident that this was on Slashdot the last time it happened. I was a bit impressed by this guys contraptions so I guess I remember the URL, here you go. http://www.sentex.net/~mwandel/tech/scanner.html
You can get a similar effect, at much poorer resolution, by processing digital video of a pan. You take a narrow slice out of each frame, and append them together into a scanned image.
I've played with this and got some images I'm very pleased with. However it's spurred me into wanting to hack some scanner hardware. Unfortunately I'm more comfortable with software than I am with the mechanical...
I wrote up the video to panorama stuff.
You have to set the scan area as small as possible.
I had to prop the lid open a tiny bit, which left tiny shadows, as if the chips were floating above a white surface.
Your assumption is based on the idea that he would learn about the same rate as you. This is wrong for several reasons
1. People don't learn at the same speed
2. This guy was learning something very specific
3. The dedication - this guy had a project he wanted to work, you were taking some classes
You should be able to code what ever you want - all it takes is the desire and dedication to do it - no formal classes are required. Prehaps the fact you never "messed with Basic" explains why you haven't learned much - you simply aren't interested enough to pursue it yourself.
My wife used to have a jewelry business, and I did all the product photography for her.
At first I tried using film; but the turnaround time, even with 1-hour developing is a drag, because it's tough to get everything Just Right, when you're dealing with highly-reflective and very small objects. So, we discovered that it's much easier to just drop the stuff onto a flatbed scanner and do a hi-res scan. The old HP scanner I had at the time had a really deep depth-of-field and a nice wide, diffused light source, so even non-flat pieces came out very nice. And, you could stick colored paper or cloth on top of the product for fun backgrounds - propped-up a bit if you needed to get the background out of focus.
Then that scanner died and I couldn't find another scanner that would duplicate the DOF and diffused light source. So I bought a digital camera.
But, boring story short: if you can find a scanner with the right DOF, you can do some really great macro stuff with it. 4000dpi at 1:1 shows a surprising amout of fine detail.
I have discovered a truly remarkable proof which this margin is too small to contain.
combine this with some really wild slit/film configurations, and you can get some interesting images... check out what andrew davidhazy is doing with moving slit photography, especially some of this stuff. he even has some articles discussing scanner derived camera backs here and here
three can keep a secret, if two are dead - benjamin franklin