someone went and made the open source camera

OMG LOOK AT THESE GUYS! LAZYWEB WINS! http://www3.elphel.com/

They have oem components that just plug into each other and are exactly what I wanted to make! I wonder if they'll sell me parts. look at the 10353+10342+10347+10369. add a kodak sensor and power and you're in business! They have a wiki! I'm stoked. Glad someone finally did this.

firmware...

I'm an idiot. Really. I realized what makes things happen after the bitstream comes out of the ADC, into the FPGA, and out to ram: a GPU is running firmware that writes the pictures out to disk. It's that simple. the firmware will also be charged with a couple other tasks:

Metering
Shutter control
Indicator lights
Server function

Notice what's missing from that list? There's nothing about driving an LCD or web browsing or any of that. I decided that because there are so many wifi devices, It makes more sense to make the camera into a Personal area network device. De-centralize it, integrate it with software on client devices. Make the camera into a limited access web-server, with a big enough drive, ssh services, and it can make up for inadequacies in storage elsewhere as well.

So, I've got the broad strokes of a design pretty much completely worked out in my head. Now for details.

last post?

Not really, but the dns changeover seems to have gone smoothly, and I'm sure the people at Stanford will want their page to be top-level on the new site. What I'd like to see happen is the blog go to a subdirectory somewhere, and maybe start a wiki with information that we have so far. Then again, because of work right now, my time is very limited. I understand sort of abstractly what has to happen to get an image first on the sensor (latent image) and then off the sensor.

So, the process begins with a little jolt across the base of the sensor, which zeros everything out. Shutter opens, photoelectric effect moves around some electrons, shutter closes. The pixels are all charged with a tiny voltage which is relative to the intensity of the light that hit the active area of the pixel. The voltage is unstable, and has to be read out quickly.

So the readout process begins. There's a register clock that determines how fast this happens, sends a voltage to the right pin on the sensor and causes it to dump a line of pixels down one of the output lines (of which there can be 1, 2, 4, or more) which hook up to the ADC, usually straight through. This is for a couple of reasons: Short circuit traces make for lower noise, ADC's are available with built in amplifiers, and most importantly, simplicity. If there is more than one output in use in a camera, usually that output will get a dedicated ADC.

Then from there it gets foggy for me. The ADC spits out a bit stream that goes to some custom logic that reassembles it into something resembling a RAW file (putting the bits all into the right order, because the way they're read out can be flopped or sectioned in any number of ways). "Custom logic" can be a prefabbed chip (Texas instruments makes a bunch that are almost cameras on chip) or an FPGA. This section also has the job of writing the image to ram and possibly on to flash.

The other possibility is that the whole assembly I've just described works as a peripheral to an embedded system that does the job of getting the images written to flash, playback, networking, etc. There are some single board computers that might work, but none fit the bill absolutely in terms of size or performance or features.

Have I got this remotely right? Is this thing on? Buller?

been thinking about this...

Thinking a lot these days. I've come up with a few more thngs that the camera needs to be and or have in order to be the kind of camera I'm thinking.

1) A screen. This wasn't in my original design, but it would be nice, and it opens up a lot of possibilities. In-the field editing, conversion from raw, and blogging of pictures, for one. Multi-touch screen and fold out keyboard, maybe, too.

2) Big internal memory (I'm talking really excessive, like a 160gb drive) plus removable storage. This is a camera to replace your computer.

3) Media center - Since it's a full computer, throw in MP3 and MPEG playback, just a port of something already out there probably.

4)User interface. This is so important it makes everything else look stupid and small. The touch screen has to be finger-usable, no stylus. Shutter lag has to be under 75ms, and should ideally be less than 50.

5)Ambidextrous- For the last few days, I've been without the use of my right thumb due to an accident working on my car. There aren't really any left handed cameras on the market. there's no reason we can't change that, especially in a one-off environment.

Also, I'm talking to some people at stanford about moving this site to their servers and letting them take over the domain. I'll still be doing my thing, but the blog will move to be a sub section of their efforts. This is a very good thing, because honestly I don't know a lot of EE stuff that I was going to have to learn, and they do. I have to talke to my hosting providers about that though. Seems the domain is locked for some reason.

I've had ideas for several versions, from the micro-mini stripped down version, to the one above that's a desktop replacement for people who only need email, internet, and GIMP. Video is a possibility, although the sensor I want to use is probably not going to support it.

not much work done lately...

sorry. been between jobs and hustlin' ain't easy. There will eventually be more work on this, but it's on hold while I get my life together. that's all.

a maquette and a shutter update...

so, first of all, the bad news. an LCD shutter is unfeasable for several reasons: they aren't clear enough when open, or dark enough when closed, and worse, because the light would be hitting it at all angles, some would leak through anyway. Which Brings me to the first real hitch in the design:

I need a mechanical (moving parts) shutter.

I was hoping to avoid this. It's an extra complication. There's a reason most camera manufacturers were proud of their super-awesome focal plane shutters- they're wonders of technology, little testaments to how awesome your engineering department is. There's no way I'm going to design one of these, it's way out of my league. Which means I need to find a commercial source. This is where I turn to you, internet. Where can I buy a single focal plane shutter in 35mm format? Electronic, quartz timed, the whole bit. I'll worry about sound dampening after I worry about having one that works.

And in other news, while I'm ruminating on the electronics design, I've been mocking up a few different body shapes. The first was a really simple box in the dimensions I gave in that first post, 5x1.5x3in. I didn't like that though, and so I made this new one, shown at the top alongside my M4 for size comparison. Yes, I like big dorky grips. And while it might get a little wider, and a touch deeper (I didn't maesure the depth, so it's a bit off), it won't get any taller. Honest. And the dorky grip? probably going to be smaller, molded. I'm still not convinced I shouldn't make the first prototype out of some really nice wood, but the sheetmetal is fast to bend a shape from, and less messy.

chip news...

So I've narrowed it down to two sensors- probably the IBIS-14000 from Cypress or the FTF-5033 from Dalsa. I like the Dalsa more- it has better fill factor, it's a newer design, and their spec sheet indicates that there's little problem with incident angles because it doesn't have microlenses. Also, you get a few more megapixels for your buck.

And in other chip news, I've found an ADC (analog to digital converter) that fits the bill- the LTC-2203 from Linear Technology. 25msps, 16 bit, 225mw of funkin'.

Other thoughts: Should I go ahead and design custom logic, or find a DSP that does what I need? Basically, the sensor sends voltages to the ADC, and the ADC spits out a datastream that should go to some high-speed ram (I'm thinking I should get a 1Gb chip, just to be excessive). Then that data needs to get written to a flash card/chip (and I'm on the fence right now as to what kind of storage to use, removable or not. Depends really on what is easier to implement). I have a feeling this kind of thing is already commercially available. No need to reinvent the wheel, and that part of the imaging chain has almost nothing to do with how the final image looks- once it's bits, it's bits, and it's not going to change.

I've decided I might need a shutter, too. I think it might be possible to get away without a mechanical one, though. If I can find one in approxamately the right size, there are those near-opaque LCDs that would be good. It only has to be 99.902% opaque to get a clean readout from a 1/2000th exposure. Which I think is doable... can't seem to find specs on any lcd's, though. it'll happen, and I'll keep you all posted.