Skip to main content
Search form
  • 00:12

    HUGH MCGRORY: Data is ones and zeros.So is everything in the digital world.And when we look at data and see that datahas turned into bar charts, and that's it,we think there is a lot more room for improvement there.All of our work at Datavized is trying to look at data

  • 00:34

    HUGH MCGRORY [continued]: and see how we can turn data into information to arriveat a system that works better for humans to beable to understand and communicate to other humanswhat the data is telling them.

  • 00:47

    DEBRA ANDERSON: Datavized Technologieshas partnership with Google News lab.And working with them, they're helpingus support the development of two new data tools.They're called morph, for visualizing data as art,and data sonification, which is turning data into sound.

  • 01:05

    HUGH MCGRORY: Let's just dive in.What I want to show you how to useis a very simple tool to make art from data.And the problem we're trying to fix hereis a lot of people are really scared of working with data.So we hack that and use the generative algorithmto build something that lets you make

  • 01:27

    HUGH MCGRORY [continued]: computer animation or graphic design in lessthan two minutes.When you come in, it asks you to either upload a spreadsheetfrom your computer or select a sample.In here, you can choose different kindsof charts, graphs.You can fill in those fields, click Evolve,

  • 01:49

    HUGH MCGRORY [continued]: and we're in a little tree.Every time you click, it will make seven babies of itselfthat are slightly different randomly.And you just click as much as you want until youfind something you like.When you find something you like, hit Export, select that,

  • 02:10

    HUGH MCGRORY [continued]: and it has made an animation out of the stepsthat you chose as you warp through.Our tools are built to be products.They're built in a way that lets you connect your own dataand do this for yourself and build your own experiencesvery, very quickly.All of our tools share one thing in common in that all

  • 02:33

    HUGH MCGRORY [continued]: you need to get going is a spreadsheet.We're trying to take just a really simple approach, makethings easy.It doesn't have to be complicated.And with a tool like this, your kids can work with dataor your grandparents can work with data,or anyone can work with data because the algorithm doesthe work for you.

  • 02:55

    HUGH MCGRORY [continued]: And morph let's you turn a spreadsheetin to graphic design animation or interactive visualizationin about two minutes and allows youto export the final product in a varietyof professional formats.

  • 03:15

    HUGH MCGRORY [continued]: And it all works from a cell phone, or a desktop,or an iPad.

  • 03:21

    BRIAN CHIRLS: For morph, I'm developing all the softwareand participating in the design.So that involves a number of steps.Typically, for projects like these,we run a number of experiments.So we build a lot of experimental code startingwith sample data sets to see what kind of output we can get.In the case of morph, it's generative graphical output.

  • 03:45

    HUGH MCGRORY: Oh, and you can change the length of time.

  • 03:47

    BRIAN CHIRLS: So we try a bunch of different things.That helps us to refine the design parametersand it also helps with the coding process.

  • 03:59

    HUGH MCGRORY: Select that.And it has made an animation out of the stepsthat you chose as you warp through.

  • 04:09

    BRIAN CHIRLS: Coding is a craft like anything else.And I find it's helpful often to havea version you can experiment with,and mess up, and throw out, and start over, and make a cleanerversion.

  • 04:22

    DEBRA ANDERSON: Let's test this in our software.

  • 04:24

    BRIAN CHIRLS: So the first step in the coding processis ideally if you can understand the parameters of whatyou're trying to build.That's not always possible, especiallyif there is a creative or very innovative component.You may not know what you want ahead of time.And that will depend on what are the needs of your client,

  • 04:46

    BRIAN CHIRLS [continued]: if you have a client, or what are the needs of your users.So in that case, it's often helpful to dosome experimenting.So we try to break the problem into smaller pieces and runthrough it, and see what works and what doesn't.

  • 04:59

    HUGH MCGRORY: What kind of data can this have?

  • 05:01

    BRIAN CHIRLS: And then there are other issues like performance.Performance is a big one, especially whenyou're dealing with graphics.So the experimenting process is an opportunity for usto see what the performance parameters are going to beand how we can adjust the volume of data, whether we'reaggregating the data, and what the output is

  • 05:24

    BRIAN CHIRLS [continued]: going to look like.

  • 05:25

    DEBRA ANDERSON: Working across these two projects,we have a collaborator and project mentor.His name is Alberto Cairo.He is a Knight Foundation fellow of digital journalismat the University of Miami.Working with him, he brings his data visualizationand information design expertise to our team

  • 05:46

    DEBRA ANDERSON [continued]: and helping provide really constructive feedback,asking key questions around the end user experienceto ensure that the application is effective and clear in termsof how to use it.

  • 06:00

    ALBERTO CAIRO: The goal is to make people be or become moreinterested in playing with a publicly available datathat they can download from any sourcesand try to transform that data into visuals that lookbeautiful, and attractive, and fun.The background of the tool used to be white.The interface looked a little bit different.

  • 06:22

    ALBERTO CAIRO [continued]: And then at some point down the road during the discussions,we decided to change the background color of the toolto black, and then use much more intense,much more pure colors in the graphics generated by morphto make those colors really pop out from the screen.

  • 06:41

    DEBRA ANDERSON: So as we develop these tools,we're able to take that feedback and iterate on that quicklyand update the versions of the tools.And this is really key to software developmentand also to designing really high quality products.

  • 06:56

    HUGH MCGRORY: One of the big limitations of datavisualization is visualization.So you have to look at a screen to understandvisual information where a sound can fill the entire room.So you could be like on the other side of the room,hear the sound change, and know that something has happened.So to us, that's just really common sense.

  • 07:19

    DEBRA ANDERSON: Sonification of datais the literal translation of raw datain, again, its rawest form, and translates that into audio.

  • 07:32

    HUGH MCGRORY: So music is math.Scales go up and down.So does data.So if you can give a numerical value of between 1 to 10and then have that play on a piano key,or go up and pitch, or change an intensity,then you can end up with a way of being able to hear data, as

  • 07:55

    HUGH MCGRORY [continued]: well as just see data.

  • 07:56

    DEBRA ANDERSON: Through the process and approachof data sonification, we take data from spreadsheetsand we transmit that through the sense of sound.And we can adjust that in terms of instrumentation,musical elements, nature sounds, anything that we can hear,

  • 08:18

    DEBRA ANDERSON [continued]: we can understand.

  • 08:19

    HUGH MCGRORY: Say if your data is rolling from 1 million to 8million for population, then 8 is the lowestand the highest pitch and 1 is the quietest and the lowestpitch, and everything in between is givena note on a scale and a pitch.So what we've actually found is that you can hear that

  • 08:44

    HUGH MCGRORY [continued]: and you can set that up by yourself.

  • 08:46

    BRIAN CHIRLS: Let's do a new project.And I will import this cleaned up spreadsheet.

  • 08:52

    HUGH MCGRORY: That's the approachthat we take at Datavized, we're lookingat this as a holistic problem and tryingto bring in things from the world of mediaand interactive to see how we can add on to and enhance

  • 09:12

    HUGH MCGRORY [continued]: the experience of being able to understand and communicatethis.Our Google tools that we've created are open source.There are a lot of benefits to things being open source.When you publish an open source tool,other developers can build on top of that architecture.

  • 09:35

    HUGH MCGRORY [continued]: So it can be taken off in directionsthat maybe wouldn't have been possible within a closedsystem.It's also closer to our ethos about the open weband accessibility.And open source does not mean that you can't build a businessaround these things, because a lot of open source tools

  • 09:57

    HUGH MCGRORY [continued]: have pro versions and enterprise accounts.So open source is a very good business model.It doesn't just mean giving everything away.The problem with data tools right nowis they are built by professionalsfor professionals.And that really limits the audience.

  • 10:17

    HUGH MCGRORY [continued]: An analogy would be if you look at the differencebetween Photoshop and Instagram.Photoshop works great for professional photographersand it lets you do everything that you might want to do.But for a beginner, it's terrifying.Instagram, the learning curve is like two minutes or less.

  • 10:39

    HUGH MCGRORY [continued]: And that's the type of tools thatinspire me-- things that are really easy, really accessible,where the interface and the algorithm does the work for youand preempts some of the decisionsyou're going to want to make.And there's no need for these tools

  • 10:60

    HUGH MCGRORY [continued]: to be complicated anymore.And part of what we do at Datavizedis try to build the software so that you don't have to.Well, you have to think very differently whenyou're building a tool for the general public.It has to be fast and it has to be easy,

  • 11:20

    HUGH MCGRORY [continued]: and it has to work on everything.It's very complicated to build simple tools.It's a lot easier to build really complicated tools.

  • 11:30

    SPEAKER: So each one of these is kindof the frame of the animation whenyou select little baby options.

  • 11:36

    HUGH MCGRORY: Yes, exactly.We're really passionate about the web.And if you look what works on the web comparedto what doesn't, we think that thereis a huge untapped potential of creativity in regular people,like just ordinary members of the public that maybe don't

  • 11:58

    HUGH MCGRORY [continued]: know how to use advanced software for animation,or data, or anything like that.Most people, including me doesn'tneed to know exactly what this letter does.Because the whole point of this is not knowing in advance.It's just experimenting.A lot of people think about technologythrough the lens of what it does today.

  • 12:21

    HUGH MCGRORY [continued]: And maybe some of the biggest companiesthat you could be working for in five yearshave not been invented yet.And technology moves very quicklyand technology gets simpler.And the less it asks of the end user, the more popular it gets.

Video Info

Publisher: SAGE Publications Ltd

Publication Year: 2019

Video Type:In Practice

Methods: Data visualization, Coding

Keywords: accessibility; algorithms; coding; data display; data visualisation; digital art and animation; digital sound; open source software ... Show More

Segment Info

Segment Num.: 1

Persons Discussed:

Events Discussed:



Hugh McGrory, CEO and co-founder, Debra Anderson, Chief Strategy Officer and co-founder, and Brian Chirls, Chief Technology Officer and co-founder, discuss their work at Datavized Technologies, a software company building data-driven tools for the immersive web.

Looks like you do not have access to this content.

Creating Open Source Tools with Google for Data Visualization & Sonification: Datavized

Hugh McGrory, CEO and co-founder, Debra Anderson, Chief Strategy Officer and co-founder, and Brian Chirls, Chief Technology Officer and co-founder, discuss their work at Datavized Technologies, a software company building data-driven tools for the immersive web.

Copy and paste the following HTML into your website