planning focus points and readability tests
This will be pretty dry and technical but some of my former students found this useful and perhaps someone else will as well!
The composite method I use for piecing together large-format prints with a high degree of detail relies heavily on planning out what parts of the ‘scene’ will be in focus and where the crucial areas of interest lie. This is also a distinguishing factor and plays into the conceptual creation of the image, and why a single frame (for instance, from a large format 8×10 film camera) won’t achieve the same effect – fundamental to the ‘readability’ of these images themselves is the rather neurotic control of depth of field and focus across the composite frame. This allows for a ‘flexible’ depth of field that can be essentially deformed across the image, allowing the simultaneous sharp focus of several elements at once that wouldn’t normally be possible with a single-frame image.
The second work I’ve planned for The Once and Future King is tentatively titled “The Impact”, and it’s just easier to copy and paste the narrative description from the proposal here to give a sense of the final product:
The Impact (Reunion: Prologue)
On a dark rural road in Ontario, the wine-red Jeep of one Randall Carter is found abandoned, door open, several large stacks of books balanced on the seats. Carefully folded on top of the books was found a military uniform, embroidered with Carter’s name and a series of unidentified insignia. The Jeep was discovered by Carter’s boyfriend, Charles D. Ward, after Carter’s protracted absence prompted Ward to retrace the route to Carter’s residence. Ward claims he found the Jeep by following the sound of music on the radio. The combined weight of the books was found to be 167 lbs, which coincidentally was approximately Carter’s own weight prior to his disappearance.
Instrumental to the work will be the ability to “read” a lot of the details of the “stuff” piled into the Jeep – literally read the book titles, but then also a lot of the signifiers and symbols present in the rest of the ephemera spread throughout. So the first function of this test shoot is to map out the focus points / depth of field across the scene, and make sure that everything that needs to be in focus, is in focus. The second function is to test the viewing angle with respect to the composite – rudimentary at this point cuz my tripod head sucks – but because we need to be able to see “inside” the Jeep the angle of the shots, and the Jeep with respect to the camera, is somewhat crucial and does complicate the composite somewhat.
Last, the test generally brings to light unanticipated problems with the shooting/composite process, and gives me a sense of “what I have to work with” – just how much detail I get straight out of the camera under uncontrolled conditions without any post-processing. As such these images haven’t been touched and a few look a little dark or desaturated for a few reasons I didn’t anticipate. Learning!
Anyway, some fairly large images and such follow, so click below to read the rest.
First things first, I generally scribble out a map of where the focus points will be, something like this:
This is actually one of the finished composites, marked up to indicate the points of focus. Key areas include the front passenger seat, the passenger-side sun visor and dashboard, the display on the stereo, and the front and back ends of the Jeep itself. The circle on the sidewalk indicates a potential placement for the figure – I’m actually undecided on this but I figured it’d be a good place to make sure the ground is in crisp focus. On a purely technical note – my new shiny iPad was really handy for this stage, I used the built-in camera, imported the image into GoodNotes and scribbled all over the damn thing. Was great to have when I was working on the test shoot itself.
Once I had everything sketched out, I took the top down on the Jeep, and proceeded to affix post-it notes to all the key focus points. They served two purposes – focus targets that were clearly visible, and as an added bonus, they had fine text from a local dental office on them, and that served well for “readability” evaluation across the image. Then I threw a few books in particular places, some other odds and end, and got shooting.
In this case I shot a matrix of 18-20 photos, basically 3 horizontal layers, 6 frames each, with the camera in a vertical orientation. I’m using a 50mm 1.8 G lens set at f5.6, and for the first test shoot the camera was roughly 10′ from the jeep at a slight angle to show the interior dash. (oh, and it’s my Nikon D800E, 36mp images … in case that wasn’t evident).
Here’s the first one, sans-red circles. There’s a Zoom.it link below the image so you can view the full-resolution version:
Zoom.it seems to alter the JPEG slightly, or had a colour profile issue, as the image looks darker and desaturated than the original. The first attempt showed promise, the detail was all there, and the print-size was about 60×90 @240DPI which is definitely respectable.
Satisfied the focus points were functional and stitched together alright, I tried a second test-shoot with the camera closer to the Jeep – about 5-8′ – to up the detail just a touch. Less wiggle room for too much distance to the left and right – in fact one of my initial composite attempts severely warped the sidewalk – but some trimming at the sides came up with the right image. It’s a little slant-angled however, that’s in part the tripod, in part the omission of some of the end pieces. Again, click the link below the image for a full-res version in zoom.it to see the detail:
I was pleased with the results in both, but the second probably saved me having to buy a new lens – I wasn’t sure the detail was up to snuff in the first, so I was considering an 85mm lens to shoot from the same distance … and still am considering … but the closer-up version worked just fine. There’s some issues with the angle and it looks a tad compressed but I think I can solve those problems.
Finally, to get a sense of what the individual frames look like, here’s two similar frames from both test shoots. These are full-res so clicking on them should give you the option of seeing them in their full 36 megapixel glory. Again, these are unprocessed, right out of the camera (well … they were RAW files originally):