30 Jan

The most recent update to Adobe’s Lightroom introduced a new feature that has me very excited, Boundary Warp! Here at LLW we are all for large scale, hyper detailed panoramic images and we have developed a solid workflow to achieve optimal results.

When logistics allow, we shoot multi-row panos either on our Really Right Stuff pano setup or on a Gigapan Epic Pro in order to capture well aligned frames with minimal parallax error. In other instances (especially when traveling by air) we need to be more expedient and resort to hand-held capture. With practice, especially for single row panos, hand-held capture is quite feasible. The largest challenge are scenes that contain strong vertical elements (I.e.: trees) spaced on near and far focal planes, which would generate strong parallax errors in the final stitch due to the camera not being rotated around the entrance pupil of the lens.

Typically we process our images in Lightroom first, synchronizing settings across the whole batch, and then we export the individual frames to PTGui to perform the alignment and stitching. PTGui is extremely fast even with gigapixel class images, and has a great variety of output projection choices that allow us to create great panoramic images. As an example, the little planet projection below was captured hand-held as three 360 degree rows and then aligned, stitched and projected with PTGui with very little effort.

Ft. McAlister, GA - little planet

Ft. McAlister, GA – little planet

Today, though, I want to show you a great new capability of Lightroom that I find very exciting and that is not available in PTGui.

Let’s start by selecting a set of images to stitch from the library module:

Merging to Panorama

Merging to Panorama

Here we are selecting 11 frames taken with the EOS 5DS R during our recent trip to Monument Valley. From the pop-up menu we select Photo Merge -> Panorama.

Panorama Merge Preview

Panorama Merge Preview

After a few minutes to process the eleven 50 MegaPixel images Lightroom presents us with a preview very similar to what we would get from PTGui. Here we selected a Cylindrical projection as we are dealing with a single row pano with an horizontal field of view of almost 180 degrees. Until a few days ago the choice of projection would have been all that we could have selected. Note that while the capture was pretty well aligned on the horizon, the borders of the image are not rectangular due to the re-projection of each image in the output canvas. This is unavoidable, and the common solution is to crop the borders out to produce neat edges.

This would be fine in this case, but in captures that are less precisely aligned the amount of ‘scalloping’ of the edges can be significant, and thus the amount of image content sacrificed in the final crop could be more than desirable.

The latest version of Lightroom adds an incredibly nifty tool to the pano merge dialog, Boundary Warp.

100% Boundary Warp

100% Boundary Warp

By setting the Boundary Warp slider to 100% the irregular edges of the image are stretched to fill the empty space in such a way that the resulting distortion is unnoticeable. Note the rail on the right side of the image, which remains straight before and after the warp. In a simple mouse drag we have a perfect image border, and we did not have to sacrifice any image content to a crop!

We did crop the rail on the right for aesthetic reasons in the final image, which totaled in excess of 177 MegaPixels for a printed size of 92″ by 40″!

Artist Point, Monument Valley

Artist Point, Monument Valley

Here are some quick peeks with the loupe at 1:1 magnification to show the kind of detail available in the final image:

Incipient HooDoos

Incipient HooDoos

Mittens in the Distance

Mittens in the Distance

Vertical Face detail

Vertical Face detail

Buttes with Snow

Buttes with Snow

The image quality achieved is impressive, and the detail in the final print is beyond anything that we have achieved in hand-held captures before! Now if I could just stop grinning and order another 8 terabytes of disk storage…

%d bloggers like this: