Emulating large format photography with a DSLR

Lessons I learned. Field-notes, part I.

One of my new year’s resolution is to chronicle my journey as an amateur landscape photographer. With these field-notes, I’ll be reflecting on how and why I take pictures.

So this isn’t a tutorial, it’s a collection of personal thoughts.

I’m hoping this process will help me learn, grow and ultimately become a better photographer. These field-notes could perhaps be an opportunity to spark a conversation, exchange thoughts or connect.

Or they might simply be a load of utterly useless crap.

Who knows?

Entre Sambre et Meuse. © Quanah Zimmerman

So, here are a few notes I jotted down as I recently ventured into the addictive world of panoramic photography.

By creating panoramic images, my aim is to try to emulate, with my Nikon DSLR, the result I’d get with a large-format camera.

I love large format landscape and architecture photography. My panoramic photography process is a personal attempt to emulate a result I’d (maybe) get with a medium to large-format camera.

I’m fascinated how large landscape images seem to draw the viewer in. That’s what I’m aiming for. I’m not looking to create wild 360 degrees panoramas.

Be consistent when taking the original shots.

The key to getting a smoothly stitched and correctly exposed panorama is consistency while taking the original photos: the camera is set to manual mode to avoid any shifts in exposure between shots and the autofocus is disabled.

The biggest challenge, when taking photos that will later be stitched together, is avoiding moving objects. This seems like a no-brainer but water, leaves and grass can create patterns of unwanted artefacts and aberrations that are nearly impossible to correct in post-production. This is the case in the Étretat photo, as the waves moved faster than I was shooting. A slower shutter speed would have probably smoothed them out. Some artefact weirdness is also visible under the bridges in the “Entre Sambre et Meuse” shot. I decided not to mess with this problem.

A panoramic tripod head would be a nice addition to my toolkit. But for now, as long as the angle isn’t too wide, I can manage just fine .

Image parallax occurs when foreground and background objects don’t align in overlapping images. To get the best results and eliminate parallax issues, the camera should pivot around the optical center of the lens, and not around the base of the camera.

But I’m not currently using a fancy panoramic tripod head.

The Étretat panorama is simply a series of eight hand-held shots. If I’m in a low-light situation, I’ll use my old trusted, standard issue, Manfrotto tripod.

If shooting a hand-held panorama, I’ll try to keep the camera level as I pan and the D800's virtual horizon feature comes in handy.

I’ll also be very generous in the overlap between shots, as it’s visible with the six original RAW files. Each shot includes well over 50% of the previous image.

These “hacks” don’t eliminate image parallax, they merely attenuate the effect and makes the stitching process a little easier.

The six RAW files that make up the Entre Sambre et Meuse panorama.

It’s important that all the original images have the same depth of field. I’ll typically be shooting these images at small apertures, anywhere from ƒ/10 to ƒ/13. I’ll generally stay away from the smallest apertures unless I’m trying to get slower shutter speeds and I don’t have my ND filter. Photographing with really small ƒ/stops will cause a loss of sharpness due to diffraction (and make any sensor dust really visible).

La falaise d’Aval à Étretat. © Quanah Zimmerman

When facing highly contrasted landscapes, I’ve developed the habit of constantly underexposing 1/2 to a full stop.

Should you purposely underexpose? This technique has been openly criticized by many photographers as it is liable to increases the noise in your image. If that was true a few years ago, the sensor quality has really evolved. And I find it easy to bring back some light in the shadows while still almost impossible to bring back detail in burnt highlights.

And I’ll only underexpose when shooting with low ISO settings. This keeps the noise down to a minimum.

By purposely underexposing, I’m not looking for faster shutter speeds. What I’m really trying to do is nail the exposure of the sky while still retaining some detail in the foreground. It can be a little tricky, notably when shooting directly into the sun.

In the future, I’ll probably try blending exposures but this means bracketing every one of the six to ten shots of the panorama. This would result in a fifteen to twenty shot composite. I’m not quite there yet.

In my recent attempts at creating these massive panoramas, I’ve been using a 35mm focal length. It seems to be the right balance. The next step might be shooting with a 50mm focal length. Maybe I’ll even try to stitch together a double row panorama. But I’m not there either.

Start the postproduction work on the cleanest possible image.

Each of the original shots is opened in Camera Raw. I adjust the white balance, correct the lens distortion and sync the settings. That’s all, at least prior to the stitching process.

I’ll merge these individual RAW files into a panorama with Photoshop’s Photomerge option but as with HDR, there are numerous third-party apps that also do a great job.

Once the panorama is stitched together, the file is saved in TIFF format so it can be opened up in Camera Raw.

The resulting panoramic image is then saved in TIFF format, so I can edit it further in Camera Raw. This time around, the exposure is globally corrected as needed, as well as the white balance.

For more flexibility, the image can be opened as a smart object in Photoshop. This way, any Camera Raw parameter can be edited at any time.

I’ll always start my post-production process by cleaning up the image. When shooting at small ƒ/stops, dust on the sensor will immediately become much more visible, mostly in the skies.

I’ll also get the borders and corners nice and tidy. Photoshop’s Photomerge does a pretty good job of stitching the images together, but while trying to partially correct the geometric distortion, the corners and some parts along the borders will be left empty, as illustrated by the jagged edge in the image below.

The image is corrected before any postproduction work is applied.

This problem can get worse as the panoramas widens. If you’re aiming for a larger angle, around or above 180 degrees, you’d probably be better off with a specific tripod head.

In both the Namur and Étretat shots, the angle isn’t really wide. Instead of cropping the photo, I clone stamped the missing areas. As with my postproduction retouching, I’m obviously taking artistic license here.

Even though probably no one will ever notice this level of detail, I still try to be as precise as I can.

I’ll typically zoom in at 300 to 400% to clean-up the image and add the missing edges.

I want to start the postproduction work on the cleanest possible image.

Be prepared to work with huge files.

I’m shooting with a full frame Nikon D800. Each image being a 7360 x 4912 pixel 35Mb RAW file, the merged result can rapidly get massive. — Really massive.

My final processed photoshop file of La falaise d’Aval à Étretat is over 2 Gb. It was the first time I resorted to save in .PSB format, which stands for Photoshop Big, also known as large document format. A PSB file apparently extends the PSD file format to a “maximum height and width of 300,000 pixels and the length limit to around 4 Exabytes”.

Who knows how much an exabyte is. But that should be enough for the time being.

There’s nothing special about saving in .PSB format as the option simply appears in the drop-down file extension menu when saving. The .PSB file thumbnails don’t seem to render in my version of Bridge, though.

The great advantage of these composite shots is the incredible level of detail. It’s a pure pleasure to work on a 8491 x 6860 pixel image.

This technique helps me slow down and really plan out the final shot:
This is the real lesson I learned.

This technique is neither an authentic large-format photo nor a real panorama, as the final ratio is somewhere between 6:4 and 6:5 (the traditional 35mm film ratio is 3:2).

But it does help me encompass a large landscape in a single image while avoiding a wide-angle esthetic and extreme geometric distortions.

But moreover, the process can be slow and unforgiving, similar to shooting with a large-format camera. This technique helps me slow down and really plan out the final shot.

This is the real lesson I learned in the process.

Are these field-notes helpful? Tell me what you think in the comments section, on the Twitters or on Facebook, I’d love to know!

Show your support

Clapping shows how much you appreciated Quanah Zimmerman’s story.