I made a 4" digital back, and you can too

[Kirk Gittings]

Since this is mimicking a virtual LF sensor, the advantages are LF's advantages; movements, great lenses, shallow depth of field, etc. You also wind up with a rectilinear image, whereas when you stitch images using a small format camera, you tend to end up with cylindrical or spherical projection images, which have a much different look.

I've done both and this is not a real difference with modern stitch software.

[TomR]

Another quick one from my front lawn.

2013-680 by Tom Rintjema, on Flickr
This image is over 15,000 pixels wide.

The main advantage of using a fixed lens and moving the sensor, over a panoramic head is you only get the lens distortion and vignetting once. If you take a series of photos with a 35mm format lens and stitch them together you have to correct for the lens distortions and vignetting for as many exposures as you take, which doesn't work out perfectly all the time. Also you don't get any of the parallax problems you may run into with a panoramic head.

[Jim collum]

I did a writeup on a tool back in 2004.. you can probably still find them in the used market... that pretty much did this.

http://www.outbackphoto.com/workflow/wf_48/essay.html

The original site of the tool is at http://www.patrikraski.com/#/studiotoolstm .. pretty sure he's not making them any more, but you can probably email and see

[TomR]

Thanks Jim. I've read that essay before. I see you ran into some of the same problems I am. I think it's important to note that I have been trying to do this on as tight a budget as possible. I've spent about $40 in parts so far. The monorail I bought used for $100 or so. I just like to tinker with things.

Don't worry though, I didn't destroy any of the components of the camera and I have ordered some film to use in it.

[Vaughn]

Actually, I was hoping to see an article about ripping the sensors out of cameras and making one's own large sensor for 4x5... ;o)

[Jim collum]

if i were a hardware guy, instead of software.. i'd of probably tried it by now

[Brian Ellis]

You certainly have my admiration, I wouldn't even attempt something like this because I'm sure I'd end up with a bunch of parts from a ruined LF camera and a ruined DSLR scattered around on the basement floor. And some of the pictures look great. I haven't studied what you did because I'd never be able to do it. But from skimming quickly, and assuming that I use a DSLR with a 3"x3" viewing screen and a TS lens, is it correct to say that what you gained here is an extra inch on each side of the screen (4"x4") and maybe a little more in the way of movements?

[Farside]

That's a VERY interesting thing to do.
Now I'm eyeing up a View II and a Canon 10D, both doing nothing.
Actually, since there's no permanent mods to either, the K20D would be better for end detail.

[erie patsellis]

Or, debayer a sensor ( see http://stargazerslounge.com/topic/16...-bayer-matrix/ ), make a similar mount and have a true b&w camera.

Even though it can be painfully slow at times, I prefer using my Dicomed scan back on either the Sinar or RB.

[paulr]

If you could biu OEM sensors from Sony and get a circuit board made, you could do something almost practical with a pair of sensors. If they're oriented vertically, and spaced just fractionally under an inch apart side by side, you could mimic a 3x4" sensor with just 4 exposures. If you're using 24 megapixel sensors, you'd have nominally 192 megapixels. The pixel pitch would be over 5 microns, so it wouldn't be excessive for the best LF lenses under the best circumstances. Even just using 12 megapixel sensors would have a lot of potential.

I have no idea how hard it is to get these parts, or how unreasonably hard the electronic engineering and assembly would be. Probably more like side project for a rich tech company than DIY.