Chris Gagné

Delight customers. Create value. Do good.

3.63 Gigapixel Panorama of Bay Bridge Lights, San Francisco CA

Here’s a panorama I shot of the Bay Bridge Lights art project on March 22nd. Like my other gigapixel panoramas, this was taken with a GigaPan EPIC Pro motorized pano head, Canon 5D Mk II, a 70-200mm f2.8 USM IS L II lens at 200mm, and a Canon 2x extender. (PS. I’m hoping to up the ante on this one by shooting a higher-resolution image; if anyone has access to 400mm+ lenses of good quality, please let me know… I am capped out 3.63GP with  2x 200mm but with a doubled 400mm or 800mm lens this could be 14.5GP. With a doubled 800mm lens this could be 58GP, quite probably the largest panorama shot in San Francisco.)

There were 420 images at 21 megapixels each. Exposure was 6″ at f/8 on ISO 100. The white balance was fixed at tungsten (3200K).

There are a couple of issues with this pano. First, there are some focus issues, particularly at the left. This is because I was balancing between reasonable exposure times (lower exposure time = less motion blur + less time in the cold) and overall sharpness. I think I would have reshot this at f/11 or higher. Second, there are some vertical alignment issues in the cables. I’ve done what I can in the GigaPan Stitch software, so I think the only thing I could do at this point would be to export the entire image to RAW, export projected images, and manually consolidate them in Photoshop. This would be trivial Photoshop work for smaller images, but multi-gigapixel images are a bit much for PS to handle gracefully (especially on a laptop, hardware specs be damned).

This is my first time using the “premium” GigaPan Stitch.EFX project. EFX adds basic color correction, including black/white levels, gamma, exposure, temperature, tint, and saturation. At first, this seemed like relatively minor functionality to try and charge an extra ~$70 for. However, trying even these basic adjustments in Photoshop turned out to be an exercise in in futility and frustration, particularly because previewing the changes took forever. Do these color correction features hold a candle to what one could theoretically do in Photoshop, the GIMP, or (ideally) something like HDR Efex Pro? Not even close. However, these alternatives just can’t handle the file sizes. (HDR Efex Pro won’t even open the file.) One possible approach would be to export a smaller version of the whole panorama, devise a color correction formula, apply this formula via a batch process to the input images, and then stitch. Unfortunately, I don’t think this would work for HDR Efex Pro.

Despite the fact that the implementation of these controls is likely anything but trivial, they are very common controls in other imaging software (including free alternatives) and I think they’re just table stakes for stitching software. It’s very, very difficult to produce good-quality images without the ability to adjust the colors. (For a comparison, here’s an uncorrected version of this same image: http://www.gigapan.com/gigapans/126089)

The folks at GigaPan Systems have been extraordinarily supportive of my work, including providing extensive technical advice, pre-release software previews, and other goodies. They’ve created a coupon code for me to share with my blog’s readers: GAGNE10. Use the code to get a 10% discount for all GigaPan products (hardware, software, print, gear, accessories) excluding Kolor products.

Any questions or comments? Post them here!

 

2 Comments

  1. Out of curiosity, what is the source of the unusual noise/posterization under the bridge?

  2. Chris Gagne

    January 21, 2014 at 1:15 pm

    Kurt, it’s likely an artifact of the (arguably aggressive) color correction work I did in Stitch.Efx. I’d love to hear other people’s experiences, but I’ve found that it’s difficult to do a lot of solid post processing with files this large. This issue is especially compounded on a laptop.

Leave a Reply

Your email address will not be published.

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code class="" title="" data-url=""> <del datetime=""> <em> <i> <q cite=""> <strike> <strong> <pre class="" title="" data-url=""> <span class="" title="" data-url="">

© 2014 Chris Gagné

Theme by Anders NorenUp ↑