Monthly Archives: October 2010

Project 6.1 – The Dispassionate Eye

Marey's 19th century gun-camera

We are surrounded by images and imaging for which there is no discernible author – rather, the image seems to have been automated, pre-determined by some program or script. This project will be an experiment in seeing through such an automated eye.

For class Monday, devise a list of instructions for creating an “un-authored” image. Bring your list printed legibly, and a camera. We’ll spend the first part of class shooting as many images as we can based on people’s instructions, and then view them digitally to discuss where meaning is made in this mode of looking, and how an automated image can leave more or less room for interpretation.

Design your instructions with our computer lab as a starting point, and make them capable of being followed with a simple point and shoot camera. Your instructions can involve detailed technical directives (“Wait exactly 30 seconds before taking the shot”), randomizers (“Close your eyes and shoot”), linguistic instructions (“Walk to the corner of 4th and Peabody”),  names (“Aim your camera at Zach’s left foot”).

They can be broad (“Pick a direction and walk a while”) or specific (“Point your camera north at 15 minutes after the hour.”)

Your instructions should include directives about where the photographer should stand, perhaps when (“Wait for the traffic to clear”), but certainly should also direct the composition, the frame. Consider that you are still framing a subject even though you won’t be shooting the picture yourself, and give enough instructions to ensure clarity of purpose.

The actor in your script should have no confusion or questions. Your script should produce something like a reliable result, though also allowing for some variety or even accident.

Your end goal should be to create a script that generates an interesting meaningful image no matter who shoots it. Possibly a particular function might even be implied by your script/algorithm.

(We’ll approach this assignment a second time next week using only the computer, and not the camera.)

Project 5 Infoviz – The Next Step

This is not the homework for Monday 10/11. Read this post for that.

Once you have your data, your job is to design and create, using Illustrator, an economical visualization of your data. For your sample set, it would be ideal for you to represent the results of all three sub-queries. You should definitely show at least two.

The goal is to help us learn something about your data, to tell a story in a way that contains ZERO extraneous information.

You may use any method of visualization you like, as long as you create it in Illustrator. (There’s no trick to using Illustrator for these things, btw – just manually set your dimensions and locations for each object.)

There are few standard modes of visualization that you would do well to start with, even if you decide to depart from them in some way. The hardest part of this project is choosing which mode best suits your data. Some steps follow below to guide you in this process. You might consider browsing the whole gamut of types included on the site for the popular info-viz tool Many Eyes, which is as good a taxonomy as any.

Along these lines, my guess is that you will be best served in this project by either the Scatter Plot or the Matrix Chart, though the Bar, Line, Stack or Treemap might also help you.

Here’s how I could go about the process:

1. You would be wise to create an Excel spreadsheet to start your project. Through sorting by column for each axis of your data, you can learn trends you might want to bring out in the image. Also, you can use the spreadsheet later to set measurements and coordinates for use in Illustrator.

2. Ask of your data: what stories do you want to tell?

– if comparison over a temporal sequence plays a role, then you might want to consider a linear format with a strong X-axis. Scatterplots, Bar and Line charts all might be well suited to showing changes over time.

– if quantity in comparison plays a strong role (even metaphorical quantities, such as “Bad, Very Bad, Worst”) Then you’ll need to get the viewer to feel that change through relative spaces. Treemaps can work well for this, as can most anything with circles.

– Prioritize your stories: If you’re shooting to show all three sets of data, decide which two are most interesting in comparison, and which one might work well as a separate layer. That’s how Scatter Plots can work, for example – the X and Y axis get compared to one another through each dot’s plot on the map, while the size or color of the dot gets read first as its own layer.

– Any given chart can be altered through removing an axis or taking one away.

3. Once you’ve had a look at your data, isolated the stories, do some sketching by hand. Try out some different scenarios. It may even be that you can think of other ways of framing your sub-queries that would produce more useful results.

4. When you’re ready with a design, start an Illustrator file by establishing a scale for yourself (ie, 1 inch = 10x more of something). Then go ahead and determine the exact measurements, color and location for each item in your database before creating it in Illustrator.

Lastly, some examples for inspiration and information:

Project 05 – Infovisualization (initial homework)

(FINAL PRINT DUE OCT 25. More of the project will be explained next week.)

Read this short article for class Monday.

Choose a dataset – textual, image or otherwise – of a substantial size. Devise a query of your set that you can carry out manually or with computer aid, a query that produces between 25 and 50 results. Then design at least 3 secondary queries that you can pose of those results.

Bring to class your dataset, your query, your three sub-query questions. In class, we’ll design a visualization that reflects the answers to this query and subquery.

Some examples would include:

DATASET: The Holy Bible, King James Version
QUERY: Where does the word “measure” appear?
SUB QUERY 01: What book does it appear in?
SUB QUERY 02: What is being measured?
SUB QUERY 03: Is this for real?

DATASET: My Facebook friends
QUERY: Whom did I add just to be polite?
SUB QUERY 01: Where did I know this person from?
SUB QUERY 02: How many mutual friends do we have?
SUB QUERY 03: Would this person listen to Prince?

DATASET: Google Image search for “first”
QUERY: What are the top 40 results?
SUB QUERY 01: What color distribution is represented across these 40?
SUB QUERY 02: How would I categorize these images?
SUB QUERY 03: What words come after “first” for each one?

I’ll present in class some examples of how you might visualize these results, using Illustrator. We’ll have some time to sketch and begin working on the project.