Navigation
Powered by Squarespace
« How much snow? | Main | Leinster, cooler than ice. Literally. »
Thursday
Dec232010

A faster way to decent panoramas?

We said goodbye to one of our colleagues today, after a few weeks with us in Dublin he is returning to Denmark, so after our regular post lunch trip to Starbucks we wandered down to the coast to see how deep the snow was. I had my camera with me and while we stood on the footbridge over the railway line I took a few shots with the intention of stitching them into a panorama.

When I shoot panoramas with my dSLR I use Autopano on OS X to create the panoramas, and when I shoot them with my iPhone I use the AutoStitch app from the the app store. When I got back to the office I realised that with the iPad's camera connection kit I could pull images from my dSLR and stitch them together with my iPad. And then I had a bigger thought. My dSLR is set up to write RAW images to one memory card and JPEG images to the second card. Why not process the RAW images using Autopano on my Mac, process the JPEG versions of the same images using AutoStitch on the iPad and compare the results.

So without further ado, two rendered panoramas from the same shots. (click the images for a higher resolution versions)

JPEG versions of the shots rendered using AutoStitch on an iPad
RAW versions of the shots rendered using Autopano on a Mac

So now for some technical stuff. My Nikon D300s shoots with an image resolution of 4,288 x 2,848, the JPEG versions of the images are 3.2 MB and the RAW versions are 12.6 MB. While images of this size present no problem for Autopano on a Mac it was a different story with AutoStitch on the iPad. AutoStitch is an iPhone app and the images my iPhone 3GS takes are a lot smaller than the images my Nikon takes, and AutoStitch seems to struggle with the higher resolution images. I could only select 7 images before AutoStitch crashed, but after rebooting the iPad I was able to select and render all ten images.

While Autopano on the Mac can pretty much make full use of the available resolution, AutoStitch seems to be limited to lower resolution output. After rendering, the JPEG image output by Autopano was 15,818 x 4,839 and 19.2 MB, the AutoStitch image was 4,130 x 1,210 and 1.2 MB. I did no selective processing, the two output images used the application defaults, but Autopano has added some exposure compensation to the final image. The only additional step I ran was in Autopano to render the larger version down to the same width as the AutoStitch version.

So what about the actual results. Well processing the images using Autostitch on the iPad is surprisingly quick and produces a pretty decent panorama. The panorama produced using Autopano on the Mac isn't perfect, the sea is underexposed and there is some colour 'funkiness' on the diving platform, but that can be fixed given some time. However it has done a much better job of de-ghosting the people on the street, and was able to use the much greater dynamic range of the RAW images which can clearly be seen in the cloud detail on the right hand side of the image.

The quality of the image produced using Autopano is clearly better, but it used almost 130 MB of image data, a computer running OS X and €120 worth of software. The AutoStitch image used 30 MB of data, an iPad and £1.20 worth of software, plus it took less than two minutes to produce and can be done almost anywhere. I'm actually quite impressed with what can be done with this iPhone app running on the iPad, and given how quick the whole process was I will have to start taking more panoramic images.

PrintView Printer Friendly Version

EmailEmail Article to Friend

Reader Comments

There are no comments for this journal entry. To create a new comment, use the form below.

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Post:
 
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>