PictoSense
In my previous post I had mused about adding sense to the media library. I was considering how one could make a media libary - tagged for people, locations and time - available in an entertaining way. The departure point for this thought process was the growing number of family and personal pictures that I currently have and manage but for which the tools are so arcane that my family would not enjoy using them.
Since then the idea has continued to pop into my thoughts demanding refinement. As a follower of GTD principles I figured that if I was to ever stop my brain from tinkering with this I’d better start jotting things down. Hopefully that’ll save me from having to keep track of things in my head and this constrantly rummaging around up there.
So where does this leave us? Being a visually oriented person, my first stop is the main screen and how it should work.
The PictoSense screen layout
Borrowing ideas from existing applications I managed scape together a first stab at a screen design without too much effort. Below you’ll see one of several screen mock ups that I’ve come up with: “map room”. Other designs, yet to be illustrated, will follow the same idea but instead of having a world map you see the “image carousel” or the “relationships map“.
The map room screen
The main features of this screen are a map with tagged locations, a timeline browser and a keyword browser. All 3 elements will interact with each other such that changing say the zoom level on the map will influence the timeline and keywords displays.
Features of the world map
The user is able to zoom and pan the map (1), switch between several views (2) such as Sattelite, Map or hybrid. On the map the user will find either markers (icons or thumbnails) representing locations where pictures were taken. The location of the markers is derived from the picture’s latitude and longitude meta data fields.
The number of markers on the map is constrained by all of the following criteria:
- Area of the map: you will not see any markers for locations not directly visible on the screen. (This may sound silly but you would not believe what I’ve seen coded across my professional experience.)
- The “current window” of the timeline browser: You will only see markers for images for which the capture date (an Exif metadata field) falls between the start and end dates that the “current window” specifies.
- The selected keyword(s) if any: You will only see markers which relate to the selected keywords. If no keywords are selected (or available) the map contents will not be constrained by this criteria.
Markers on the map can be either pictures or icons. Clicking on a marker will pop-up a larger version of the image along with selected meta data information (e.g. caption, keywords, capture date, etc.)
Features of the timeline browser
The user is able to see a timeline representing all the images in the library. Within this timeline he can set a “time window” which acts as a constraint for the world map, carousel or relationships map. The timeline has a histogram representing the number of images captured at a given point in time.
The timeline will be constrained by the earliest and latest caputure time of all images in the library. The time interval markers (eg. year, month, week day) will vary depending on the time span of the library. Several implementations are possible (see example 1, example 2, and a really neat one).
Features of the keyword list
The user is able to see keywords of all the pictures within the current “time window”. The size of the keywords is relative to the number of times each keyword appears within the set of images. Selecting one or more keywords will constrain the number of markers on the map. Unselecting a keyword will remove the constraint.
The selected keywords will not cause the size of the histogram to change. They will however cause the histogram bars (dates) of the related pictures to be highlighted. This will occur for the entire timeline and not just the current “time window”. This gives a visual form of feedback to the user of the times at which selected pictures were taken without modifying the height of the histogram which would be distracting.
Further thinking
The screen design and features are very early in their lifecycle at this stage. There are many more things which need to be considered and which during the writing of this have tickled my brain. What options should be provided, what happens to slections on changes in the “time window” etc.
For now though, there’s enough on paper for my brain to stop playing tricks on me by calling this topic to the forefront at the most inappropriate of times.
Tags: Computer Interfaces, Geeky, photos, tools, Usability
Adding sense to the media library
I’ve had quite a bit of time on my hands the last few days - being sick tends to provide that oh, so valued commodity which tends to escape us as active adults. I’ve used the time to get through some of the backlog of my snapshot database.
The libary
As you may recall, I’m pretty anal when it comes to the family’s picture archive. When we purchased our first digital camera what feels like eons ago, I quickly realized that we would end up with fairly large library over the years. And as with any largish library it would need management. So I set about tagging the images, a tedious but necessary evil if we were ever to make sense of the thing 20 years down the line. I won’t go into the details of the workflow here and just mention that I tag pretty much anything for: who took the shot, the person in the shot and the location it was taken at.
Recently I’ve started wondering how this data can be put to good use. My current toolset can very nicely tell me what I generally need to know, where something was taken, by whom etc. It will only tell me those things because I know the tools. That’s fine as long as I’m the only user, but that’s not really the point of the library. How could this library be made accessible to others (e.g family and friends) and if so what are the salient features I would put at their disposal given the meta-data at hand?
After a little pondering a couple of things seem essential: a timeline, locations, and people. Whomever would have access would like to have the ability to sort and filter data based on these criteria. These are fairly simple features which I’m sure can be found in a number of user friendly pieces of software.
Architecting a new libary
Yet how do you bring those to the user in an entertaining way? The idea of entertainment also seems a key criteria. Whomever would view our library would likely not be looking to trudge through thousands of pictures. Rather, this person would be interested in the story that the pictures tell.
Figuring out what the libary viewer, for lack of a better name, would look like and how its usage could be desinged is what’s been prickling my mind. The way I go about figuring this out is to answer the following question: “What are the storylines one could derive from the meta-data in the library?” It’s not the easiest of questions to answer and If you’d like to participate then please feel free to comment…
Time and space:
For one we can use locations pictures were taken over time using the location information and the time a picture was taken. For this I envisage a map of the world with a bracketed timeline. The map would display pictures, or rather some form of reference to pictues for a given time bracket. These references could be bubbles which vary in size with the number of pictures that were taken around a given location for instance. The timeline would function very much like the one found on Google Finance. It would allow you to set the width of the bracket (i.e how many days, months etc. to take into account) and allow you to slide that bracket over time. When you slide the bracket the references on the map would change in accordance.
The timline and map would work in conjunction to filter the pictures to those that you’re interested in. Other features would include the ability to zoom in and out on the map and the ability to click on a reference (bubble) to further filter the pictures. The pictures themselves could be listed in several ways although my personal preference would go to a filmstrip type of listing.
Relationships
Another idea would be to use the people in the shot. As each picture has an author and one or more people in the picture one can establish a sense of relation both between the author of the shot and his subjects as well as betwen people in the shot.
The first thing that springs to mind in this regard is a graphical representation of relations between the people. The strengh of the relation would be relative to the number of times any 2 people were in the same shot together. Something like the Visual Thesaurus would be an interesting start piont The central weight points could be around the authors as they’d could be considered to be in every shot - that’s debatable though, one should see how this works out in the data representation.
One could thus navigate the picture database solely based on the relations of people. Selecting one or multiple people would filter the number of people in the shot.
A second feature which could be added is the timeline. This could be achieved in very much the same way as with the “Time and Space” idea such that selecting a time-frame would restrict the number of people shown.
Further afield
I’m sure that there’s more ways to view this information, but for now my brain is fried - I am sick after all… To be continued then.
Do leave me a comment or two in case you want to contribute
Tags: Computer Interfaces, Geeky, photos, tools, Usability
