Large-ish Format w/ Digital Back and Stitching

[bieber]

Today I finally got around to trying a technique I've been meaning to play around with. I used my digital back with an adapter on my P2, set up and composed a still life on the ground glass, and then made a series of digital exposures of it in pieces using the rise, fall and shift on the rear standard. Once everything was stitched together (which took a while), I ended up with an image with an effective sensor size of about 3.2x3.5 inches, although I ended up cropping in significantly from there to get my final composition (went from ~200MP to ~100MP in cropping). So I'm not exactly using the full capabilities of the 4x5 camera, but I'm getting a lot closer than I've ever been with a digital camera. The biggest limiting factor is just how much movement I can get out of the bellows, since trying to translate the center of the frame out to the edges is a pretty extreme movement, and doubly so if I used any movements to begin with. I might get a sliding adapter at some point to give me some more flexibility on that front. I tried to do everything I could with front standard movements so I'd still have room to move around on the rear standard. And of course I have to avoid using tilt or swing at the top of the rear standard in order for my movements to stay in the same image plane, but if I really need to I can use the coarse tilt, since the rise/fall and shift dials are above it.

Here's the final composition in full size: https://www.flickr.com/photos/tehbieber/41378191884/in/dateposted-public/

Just ignore the horrifically posterized preview image Flickr made. The download sizes all seem to be pretty decent though. I'll attach some very down-rezed versions of the original stitched image and the final composition here, too.

178257

178258

[bieber]

Anddd, of course the forum software also posterized it to hell and back

Edit - Actually, I guess it's my browser. Renders fine everywhere in Firefox.

[B.S.Kumar]

Which lens did you use? A wide angle bellows would solve the problem with restricted movements.

Kumar

[bieber]

I just have a 150/5.6 for now. I didn't really need any more area for what I was shooting this time, but I do have a bag bellows and I'm thinking about using it next time I try this. On left/right shift I can actually hit the limits of the camera pretty easily, my P2 is only marked for 3cm left shift, and it won't physically go much farther than that.

[B.S.Kumar]

A WA bellows is perfectly usable with a 150mm lens. You can also combine front and rear movements for even greater coverage, but since the lens position changes, perspective also changes. That will probably result in weird artifacts in a studio setting.

Kumar

[bieber]

Yeah, for stitching I need to keep the lens fixed throughout the different exposures, so I'm just sampling different parts of the same image circle. I do make sure to do whatever compositional movements I can with the front standard though, so the rear one has as much room as possible to move around

[Joshua Dunn]

Bieber,

I am currently experimenting with this as was well. What back are you using?

-Joshua

[bieber]

I'm using the credo 40. One of these days I really want to get my hands on one of the leaf or phase one 50mp backs with the CMOS sensor though, the double shot process really shows off all the flaws of the CCD sensor. In particular I've been getting some noticeable centerfold artifacts, so I try to make sure I get some overlap in my images and hugin does a pretty good job of blending everything smoothly

[bieber]

I was shooting with an extra bellows and intermediate standard today just for the magnification, but it turns out that also helps a lot when it comes to having some freedom to move around. Even with the front standard tilted all the way forwards, I was able to get a really wide range of movement on this one. 3cm left, 4cm right, 3cm down and 4cm up, all without really straining the bellows. That got me pretty darn close to 4x5 in the original capture, although I lost some of my edge room to stitching artifacts. Ended up with a usable area of about 4.25x4.1 inches, which I'm not complaining about: https://www.flickr.com/photos/tehbieber/42180117392/in/dateposted-public/

178439

[Nodda Duma]

For reducing fixed pattern noise, dark frame subtraction is a technique astrophotographers have been using since the dawn of digital imaging to remove this..essential when all your information is in the lower 25% of the histogram. You may want to check that out.

[bieber]

I'm pretty sure that's what the back's doing when its noise reduction kicks in after a long-ish exposure. The problem is even with that, medium format sensors heat up quickly, generate a lot of noise when they heat up (at least from this generation, some of the new ones are amazing), and worse yet the differences between the four separate sensors that make up the "sensor," which are normally controlled for by a correction factor, seem to start to grow beyond the level you can neatly correct for.

Attached to a MF camera, I can shoot dark scenes like this and get beautiful clean images, as long as I'm using nice short exposures with strobes and keeping the ISO low. The problem is that when you use it with the contacts disconnected, you have to manually signal the back to wake up the sensor before making the actual exposure, so (a) the sensor has to be awake for much longer than it normally would, generating heat, and (b) since the whole thing is manual and controlled by human reflexes, it's not repeatable enough to even attempt a flat field image. All that being said, with a little bit of redundancy in the stitching everything seems to be turning out pretty okay

[Nodda Duma]

Sounds very familiar. Astrophotography exposures can run minutes-long, and a major concern is, as you can relate, sensor temp. One technique to address this is to enclose the entire camera in a cooling box with a TEC ( an extreme solution but very effective), and the other is to generate a database of dark frames catalogued not only by exposure time but also sensor temperature. I’ve found this makes a significant difference. Not only do dedicated astrophotography imagers report sensor temp, but even my Canon 40D provides that info. So you might be able to get at that data likely through a software interface.

Hope this helps! Sometimes cross-pollination of ideas from other hobbies can be useful.

[bieber]

Interesting, I'll have to see if I can find any sensor temp data hiding in the exif