Time Lapse Photography

To continue with my recent theme of technology articles, it seems appropriate to touch on another interesting photographic method-- time lapses. You’ve seen them on TV more times than you can count, but I doubt you have ever created one. Depending on the hardware you have, it may be a simple or complex endeavor, but it can be quite fun and can potentially reveal something that would be difficult to observe otherwise.
One of my favorite time lapses is shown in the video below and reveals significant motion of an iceberg in Alsek Lake in Alaska over a the course of about 2 hours. This scene is particularly interesting to me, because as I sat on the shore of this incredible lake with my friends for several hours, we would occasionally note that the massive icebergs had moved significantly right before our eyes without any perceptible motion. So in the remaining day light hours, I captured this berg’s motion and confirmed that it was indeed moving quite a bit right in front of us.

Panorama Creation

We’ve all seen beautiful panoramic photos that capture sweeping landscapes and have probably experienced 3D tours on real estate websites that let you look in different directions around the rooms of the house. Although these technologies are widespread, they are not widely used by the layperson to share their own experiences of the world. Given my interest in exploring different ways to tell stories digitally, I thought it would be worthwhile to explore the realm of digital panorama creation.

The basic process of creating panoramas can be summarized in a few steps, each of which I will explore in more detail:

1. Capture the scene by taking multiple overlapping photographs
2. Stitch the photos into a panorama
3. Share the panorama with viewers

JAlbum Usage

Today, there are more ways to share photos online than ever before, from Facebook to Flickr to Picassa to Snapfish and literally hundreds of more options. However, almost all of these online photo sharing tools have a few negative aspects in common. First and foremost, the provider of the tool makes the rules. That means that if Snapfish goes belly up, say buh-bye to all of your albums. It also means that they decide on the look and feel of the album- from the size of the images to whether metadata is displayed. Second, some photo sharing tools, such as Facebook, claim to own your photos after you upload them which is completely ridiculous. No matter how you look at it, what you gain in convenience with these tools, you more than lose in control over your photos.

Years ago, I used to to code photo albums by hand in HTLM, but since those early days of the web, I have found a stable, free photo album tool called JAlbum. It is an extremely flexible, easy to use tool that can be run locally on your computer or remotely on their servers. The basic premise is that you build a photo album which is static in its very nature. This means that the album is an entity unto itself that can be saved, copied, moved and as long as HTML remains in existence, will be a permanent, unchanging record of how you wanted your photos to appear in the album. Basically, it is the digital equivalent of the old-time, “real” photo albums that sit on your shelf at home. Contrast this with a transient, fleeting view of your photos that online tools provide. As far as I’m concerned, my photos-- my memories-- deserve better treatment than the online tools can provide.

As with any software tool, there are some tips, tricks, and workflow techniques that can really make the difference in JAlbum’s effective use. Since my usage of JAlbum tends to come in waves, it only seemed appropriate to document the steps and the configuration that I find useful.

1. Download JAlbum and install it on your computer.
2. Start JAlbum.
3. Configure directory preferences
   • Set the “My Albums location” under Tools->Preferences and then the Album tab. This is the location where JAlbum will create your albums. In older versions, it used to create the album in the same directory as the photos, but now that it can accept photos from many directories, it makes more sense to keep all the albums in their own separate location from the photos.
4. Create a new album if this didn’t happen automatically by selecting File->New Album Project
   • The Album Name will also be the directory name that holds the album in the album location that you set in Step 2, although you can override that location if you need to. Generally, I leave this alone.
   • Add some photos to the album by clicking on the “Add” button in the toolbar in the top of the main window.
5. Modify the album settings by selecting Album->Settings
   • On the General page, set the Output directory to a value of “.” Yup, just type in a single period into the whole line, which tells JAlbum to output the album files directly into the album folder instead of a subdirectory.
   • On the Pages page, set the “Image linking from thumbnails” to “Link to originals via scaled images”, since it is always nice to be able to download the original file, but it is much easier to flip through an album that has scaled down images
   • On the Images page, I leave the default values alone. In particular, I want the JPEG quality to be 100%, Thumbnail size should be pretty small at 130x130 pixels, and Images should be a medium size at 600x600 pixels.
   • On the Advanced page, I’m sure to select IPTC Caption as one of the Comment sources, since I want to be able to extract all the wonderful metadata that I previously added to the photos.
6. Install and configure the BluPlusPlus skin. Skins provide a very customizable look, feel, and functionality to your albums. I prefer BluPlusPlus for its simplicity and its ability to handle metadata well.
   • Since BluPlusPlus is an add-on to the default skins, you will need to download it from the JAlbum skins website and install it per the directions on the website.
   • Select BluPlusPlus for the skin and “expedition” for the style on the left side of the main window
   • Now back to Album->Settings and there should be a BluPlusPlus page, which has several tabs of its own. On the Slide tab, enable display of some nice metadata by entering the following for the Slide Caption Comment:
     • Iptc.Object Name||Iptc.City, |Iptc.Province/State |Iptc.Country/Primary Location||Date||FileName|Dimension|FileSize||Iptc.Copyright Notice
7. Create the album by clicking on the “Make album” button in the bottom-left corner of the main window
   • After the album is created, you can “preview” the album which really just means that you are looking at the local copy of the album. Since the album is its own static entity, this preview should be the same as the “live” version, but it is housed locally on your hard drive instead of on a web server.
   • If you edit the IPTC metadata after you initially added the photos, you may need to delete the corresponding images in the My Albums directory so that JAlbum is forced to really read the metadata.
   • Finally, upload the album to a webserver to share with the world.

Keep in mind that these are the steps that matter most to me and that there are infinite ways to build JAlbums and that there are tons of online resources to help you figure out all the different steps. Have fun!

Tethered Balloon Flight

My interest in flying a tethered balloon with a camera system started back in July of 2006. Since then, I designed and built a sophisticated camera system, purchased a  7 foot diameter helium balloon, and have talked about flying it over Boulder to take aerial photographs. Almost 4 years have passed with enough yak shaving and procrastination for me to vow to finally do this on not one, but two New Year’s resolutions. I can finally and proudly say that I flew the balloon and the camera system over my South Boulder neighborhood and it was a tremendous success!

My initial interest in the subject goes way back to my Master’s Thesis that involved the design of a micro air vehicle control system for DARPA. The justification for the thesis work was that there is a strong need within the military (and other agencies) for visual knowledge of an environment that is best provided by an unmanned aerial vehicle (UAV). However, UAVs can be difficult to operate, so much research has been invested in developing autonomous UAVs that can operate on their own, which is exactly how I was funded for my research project.

Flash forward to 2006. I no longer live in Boston, but I am still working in the aerospace industry. Instead of designing control systems to fly aerial vehicles, I am designing control systems to point scientific payloads on satellites in earth orbit. I pick up my first issue of Make magazine while visiting my parents and am immediately blown away by the advancements in low-cost, robotic-like electronics that are available to the average hobbyist that used to be top-shelf military grade components. I subscribe to the magazine and request all back issues, which included their premiere issue focused on kite aerial photography. I have an epiphany. Kites can fly above the earth in a relatively fixed location and orientation without any sophisticate control system, which allows people to focus on building camera payloads to take incredible photographs. I am convinced that I have short-circuited all of the work that has gone into UAV control system development. I realize that there isn’t always wind to fly a kite. Duh. I think back to my childhood and remember helium balloons. I am a kid again and having fun with a balloon. It is the answer. I am convinced that the answer is to fly a helium balloon with a camera instead of investing precious time and money on developing sophisticated autonomous UAVs. I research the market and learn that three companies visibly cater to this type of market: Southern Balloon Works, SkyDoc Balloons, and Allsopp Helikites. All of these companies do impressive work and had my epiphany long ago. But they don’t design precision pointing systems for NASA, I do. I convince myself that I can buy a balloon from one of these vendors, build a sophisticated camera payload, and point it better and for cheaper than anyone else using low-cost hobby electronics.

We are now in July 2006. I spend some of my free time furthering research on the topic, ordering a balloon from Southern Balloon Works, buying a Canons S3is camera, buying and learning to work a BasicStamp-based Parallax BoeBot, buying a camera cradle from ServoCity, and ultimately building a pretty nice semi-autonomous gimbaled camera system to fly on the balloon. Of course, there were issues or I wouldn’t be writing this in 2010. The real-time video downlink from the camera had numerous issues. I eventually worked out the kinks between the RC transmitter and the video downlink and wound up with a reliable RC/video system. I also got involved in mentoring a FIRST robotics team which was a valuable experience, but that also ate into my free time for such extracurricular nerdy activities. Ultimately, it came down to priorities and focus. There are a few things I can say with complete conviction and one is that when we make something in our life a priority or if it occupies our focus, it will be done well and it will be done quickly. So although you could say that requirements creep caused me to try to make the system too autonomous or I got caught shaving a yak, when it comes down to it, flying a helium balloon just wasn’t a priority and didn’t occupy my focus for the past two years despite the fact that I really wished it would.

Sometimes, things in life don’t go as planned and we have to accept them for what they are, decide to fight against them, or a little of both. When it came to this balloon project, I would like to think that I went with a little of both. First, I was fortuitous to become friends with Sam and Lars, who have helped me tremendously in literally getting this project off the ground. Second, I made the conscious choice to abandon the lost cause of a BasicStamp controlled “smart” camera system and to pursue a remote controlled camera system that was well within my means. What was amazing was that after making that choice, it only took three weeks to make the adjustments to the system, get my friends onboard with the project, and ultimately fly my gimbaled balloon camera system. You can probably begin to understand now why I am proud of this accomplishment.

Instead of explaining the as-built system in writing, I thought it would be easier to watch it in action. So check out the overview video below.

And now onto the premiere flight of my tethered balloon camera system.

As you can see from the video, I think the first flight of this system went very well. The as-built system functioned as I had hoped, but the balloon proved to be much less stable than I expected. In some ways, that actually makes things more fun, since it provides me with a justification to improve upon the system :-) It may not be the game changing invention that I dreamed of years ago, but I am glad that I dreamed and proud that I had the courage to try to make it come true.

Be sure to check out the photo album to get a sense of the quality of the photos. There are some fun ones in there! The photo below highlights the improved resolution that near earth aerial photography can provide compared to satellite imagery. Also, if you want to see this photo and two others in Google Earth, down the this KMZ file.

Kite Aerial Photography

What to do on a windy April day in Colorado? How about flying a kite high above an office building with a camera dangling precariously on the kite line in a poor man's attempt to take some cool photos? That's exactly what Lars and I did yesterday in my first foray into Kite Aerial Photography (KAP).

I first read about this in the premiere issue of Make Magazine several years ago. With my interest piqued about aerial photography, I began a slow quest that has led me to build a tethered balloon photography rig (still hasn't flown though) and to launch a camera to the edge of space twice so far this year. On the eve of this year, I mentioned the possibility of getting into KAP, so with the windy weather and a CHDK enabled camera, it only seemed natural to give it a go yesterday.

The winds were at least 15-20mph with gusts much higher, so it seemed likely that my kite could lift the weight of the camera. I had never attached any mass to a kite line before, so the first experiment involved wrapping the taught kite line around a carabiner and clipping a wrench to it. The wrench was easily hoisted into the air as it swayed all over the place in the gusty air. Two minutes later and the wrist tether of the camera was clipped to the carabiner and it was up in the air wildly taking photos! Unfortunately, the battery ran out and it was only programmed to take photos once per minute, so back into the lab for battery charging and reprogramming. An hour later, the camera was back up in the air taking photos once every 10 seconds! I quickly walked the line down, reviewed the photos to make sure it was all working, and back up it went on its completely unstable tether to a height at least 200 feet above the ground!

Considering how little effort (and thought) went into this little aerial photography experiment, I would say it went extremely well. We actually got quite a few nice photos, the best of which is an extremely crisp photo of our neighbor CASA with the Flatirons in the background.

Be sure to check out some more photos in the photo album.

Of course, I couldn't leave it at that, so last night I set about building a Picavet suspension to hold the camera level to the ground for the next time we have some nice windy conditions here in Boulder...

On another quick note, this trusty soldier of a camera is going to fly to the edge of space yet again this Friday as part of a documentary being produced by the Boulder History Museum. Let's hope for its safe return!

Going UP Tomorrow!

UPDATE! 9:36am- Up, up, and away!

Tomorrow morning at approximately 8am MST, a weather balloon with my modified Canon SD800is, my SPOT GPS transmitter, and my Garmin Vista Cx will voyage to the edge of space!

The idea belonged to my colleagues Sam and Lars, and was soon added to my list of resolutions. After realizing that it was a reasonably easy goal to accomplish, we set about gathering all the parts and preparing to launch the camera to 100,000 feet above the Earth. For comparison, jet airliners fly at 35k feet and at 100k feet the curvature of the Earth and the darkness of space become apparent. Our goal is to photograph this breathtaking sight and return our payload safely to earth using off-the-shelf components that we already have. The camera was modified using the Canon Hack Development Kit (CDHK) to take photos at regular time intervals. The GPS will record the 3D path of the balloon as it ascends through the atmosphere, into the fast moving jet stream, and finally above 99% of the atmosphere where it will expand to a point that its latex surface will rupture. The payload will rapidly descend through the upper atmosphere until the air density increases to the point that the attached parachute slows the descent to a safe speed to its landing location about 100 miles downwind from the launch site. And all the while, the SPOT will be relaying the position of the balloon to us in real-time!

So if you are interested in following along, just take a look at the map below and be sure to refresh every it 10 minutes or so. I can't guarantee the exact launch time, so keep checking our progress throughout the morning.

IPTC Metadata

Interesting topic, huh? It turns out that metadata is one of things that is quietly in the background making things more efficient and productive without you ever knowing it. Every time you take a digital photograph, information is stored in an EXIF header that includes the camera settings (shutter speed, exposure time, flash, etc.) and the time and date of the photo. This makes it very easy to find when photos were taken without you needing to manually record the date, time, and camera settings the way photographers used to do SLR film cameras. This information is considered metadata because it is a small snippet of the file that is separate from the digital representation of the photo that is the bulk of the file.

Now, don't you wish that there was other information associated with each photograph, like location, who is in the photo, the photographers name, and a description? Lots of different software applications are attempting to deal with this by allowing you to "tag" photos or by including a description, and this definitely is a step in the right direction that will allow you to search for photos more easily in the future. But what if you stop using Windows Vista, .Mac, or Flicker? What happens to those descriptions and tags? They disappear! That's because they aren't part of the photo and they aren't based on an industry accepted standard. This is where IPTC headers come in. IPTC stands for the International Press and Telecommunications Council. This group recognized that photos needed metadata attached to them with descriptions, photographer, etc. so that they could be easily passed from one news organization to another. The reason this matters to us is that it provides a standard format for this metadata that will remain largely accepted throughout the world for the forseeable future.

Now the question becomes, how can IPTC be used to actually make my digital photography experience better? First off, you need to be able to add IPTC headers to existing photographs. I'm sure there are dozens of ways to do this, but I will just mention a few here. Originally, I used a free application called PixVue to edit and view IPTC data for photographs from within MS Window Explorer, but it has come and gone and come again, so I can't guarantee its availability. IrfanView is a popular, free image editing software that can also be used to view and edit IPTC data, but I found the process to be somewhat cumbersome. Most recently, I found that Microsoft has released a tool called Pro Photo Tools (previously called Photo Info) that is very similar in functionality to PixVue for editing and viewing metadata for groups of photos with the additional added benefits of being able to correct camera date/timestamps and geotag photos. I think this type of tool should be included in the Windows and Mac operating systems since it is such a basic task. I'm not going to go into the details of editing this data, but basically, you just need to select a group of photos and right-click on them to be presented with a dialogue box that lets you edit their values. The toughest part of this editing process is remembering to actually do it! However, once that information has been entered once, there will be no need to re-enter it again in the future.

Once you have populated your IPTC header data, you can then take advantage of its presence. There are 3 ways that I currently used this information and I'm sure it will only expand in the future. The first way the data is useful is when searching for photographs. If you use a search tool on your local machine, you can now search for terms inside the photos instead of just browsing through your own personal heiarchy to find photos. For example, it would be much easier to do a search for "Pat" within all the photos in My Pictures than it would be to browse through thumbnails of all those photos to find pictures of me in them. Second, I use JAlbum to create online photo albums of my photographs and once the IPTC is populated, the album software is able to simply extract the description and location information and post in on the web pages without me doing a thing! Similarly, when I use RoboGeo this data is also used to create descriptions of photos within Google Earth.

That's all for now!

Update, 3/6/2011:

After a recent round of computer updates, I found myself needing to edit some photo metadata. Initially, I used the PixVue installation that was referenced above, but Windows Explorer (the file browser, not the web browser) kept crashing under Windows 7. So I moved on to using MS Pro Photo Tools, but it interspersed the metadata somewhat haphazardly between the IPTC and EXIF headers. Finally, I discovered that Irfanview can be used in batch mode to edit IPTC data and I’m happy with this new method.

Immersive Images

Today's technology blows my mind. Not only are digital cameras so pervasive that they are found in almost every cell phone, but the quality of digital photographs and the methods of presentation are amazing. I have noticed a recent trend that is attempting and largely succeeding at treating the viewer as more than a mere, outside observer. Instead, new technology is immersing the viewer into the world in which the photography was taken. I consider these to be immersive images and thought I would discuss some of the established and up coming technologies here.

1. Virtual Reality (VR) Panoramas



2. Google Street View in Google Maps and GE



3. DigitalUrban Panorama Viewer for Google Earth



4. Gigapan in Google Earth



5. Microsoft Photosynth

---

VR Panoramas
The veteran technology here has to be virtual reality panoramas. These were made widely popular by Apple's Quicktime Virtual Reality (QTVR) standard and can now be seen on most realtors' web sites as an attractive way of presenting homes for sale. They provide a very inuitive and effective experience that truly immerses the viewer in the scene. Simply dragging the scene in any direction replicates moving your head and the ability to zoom is typically included as well. I have spent a little time learning how to create these objects and will discuss the methods that worked for me in a future post. In the mean time, you can check out my first QTVR based on a panoramic image I took at Walker Glacier in Alaska this past year.

---

Google Street View
Google took the idea of VR panoramas to a whole new level by having cars with automatic panoramic cameras on their roofs drive the streets of most large metropolitan areas. The result is Street View- the ability to "step into" the Google Map or Google Earth street and see what is actually there. Once again, the interface is intuitive, but I will let Google describe it for themselves. Once you have Street View at your disposal, it's amazing how useful it can be to travel parts of your route before you actually get on the road.

---

DigitalUrban Panorama Viewer
I recently learned that there is a way to embed VR panoramas in Google Earth so that a viewer can step into a photo realistic view of locations on the ground. I will be adding my Walker Glacier panorma shortly using the instructions provided on their blog. I love the fact that this allows a casual user to integrate two such powerful experiences together.

---

Gigapan in Google Earth
Gigapan expands on the notion of embedding VR panoramas in Google Earth by insisting that they contain gigapixels of photographic data. The basic idea is to use a robotic panoramic mount with a standard digital camera to capture dozens if not hundreds of photos that are then stitched together to form amazingly high resolution VR panoramas. Although the panoramas can be viewed on their website, it is a far more enriching experience to view them in situ in their Google Earth layer. The robotic camera mount will become available to the public in the near future, but it remains to be seen how many people will invest in such a technology and then generate content for others.

---

Microsoft Photosynth
Whereas Gigapan is based on an extremely deterministic, proactive capture of photos to generate a virtual representation of a scene, Microsoft's new technology named Photosynth assembles 3D scenes based on photographs that are scoured off the web. This amazing technology uses photographic keywords to locate photos of a particular scene and then it uses advanced machine vision algorithms to identify common geometric features in the photos and derive the location of each photo in the scene. And it does all this with no preconceived idea of what the scene is before hand! If you want to see first hand what I just described, check out their technology preview on their site. I imagine that the preview is intended to show the power of the technology and not the final user experience that will likely be similar to other VR photo viewers. Clearly, the power of this technology is that the amount of available web content available to generate an immersive experience is far greater than the number of intentional immersive images. It will be very exciting to see how this technology progresses and becomes integrated with MS's Virtual Earth in the near future!

Update (6/3/2008)
It looks like Google is trying to compete with MS Photosynth with their Look Around feature for select locations in Panoramio. It doesn't seem to be as immersive as Photosynth or nearly as adaptable to the everchanging landscape of the web, but who knows how it will grow in the coming months and years...