Testing in Studio

Today I tested my interactive piece in the studio room, W119. There was a white board and white projector screen available which I was able to use for the clear background. From testing the piece, it allowed me to see if it worked properly, as well as to see if there are more improvements that can be made. Here is a video of me testing my piece:

As you can see, on the screen I was captured as a black silhouette as planned. However, the background wasn’t entirely clear as it picked up the edges of the board and screen. Therefore they came up as black blocks/lines and obstructed the shapes from falling. To fix this problem, I need to ensure that the screen I use is big enough in order to fill the width of the camera. As you can’t entirely see the screen, here is a screen recording of the test:

From this, I feel that I need to edit the speed again to make the shapes fall faster. I also think I need to edit the size of the shapes more so that there are some bigger ones rather than them all being small. The programming works as expected as when the shapes are on a white pixel the descend and when on black they ascend. By carrying out prior testing, it has allowed me to further improve my piece to make it more suitable and successful. I am pleased with the outcome and will make further amendments so it is ready to display on the public screens in Weymouth House.

Improvements – Shapes, Size, And Speed

Some suggested improvements I received when testing my piece was to possibly add more of each of the shapes. As well as to possibly make the speed at which the object fall slightly faster, and to make the size of the objects more varied.

Here is the code for initialising a new object and setting the size range, of which I changed the range to random(10,30):

Screen Shot 2015-01-13 at 17.02.10It was also suggested that I should have more objects, which I altered easily in the object arrays, where I can choose to have 0-infinity objects:

Screen Shot 2015-01-13 at 17.04.39

Changing the speed at which the objects descend has to be done in the individual classes constructor. I edited the yspeed so that objects descend and ascend slightly quicker:

Screen Shot 2015-01-13 at 17.06.50

I feel that these improvements have benefited my piece, highlighting the importance of carrying out user testing. I decided against turning my piece into a game as felt it would be too easy. As users could easily put out both hands and prevent the shapes from falling, like this example…

Screen Shot 2015-01-15 at 13.10.58

From this, I will go on to test my piece in a studio room, capturing more of the persons body rather than just the top half. Allowing me to see if there are any further refinements I could do before displaying my piece in the public space.

Improvements – Mirroring Piece

An improvement I received when testing my piece was to mirror the video capture. It came to my awareness that at the moment it captures the video the wrong way round, so when users put up their right hand, on the screen their hand goes up on the left. This makes the interaction quite confusing which could deter people from interacting with my piece. I have found this improvement quite difficult to do and have had quite a bit of difficulty with the programming. Through reading online forums, I managed to gain some help from others who had similar problems. Here are the steps I went through and the various problems I encountered:

Screen Shot 2015-01-13 at 12.58.02To start with I managed to get my piece to mirror by adding this code:
Screen Shot 2015-01-12 at 12.23.50

This got the video capture to mirror but meant the brightness thresholding wouldn’t work correctly and so the video capture was in colour. This meant that obstructing the falling objects didn’t work as well as it was harder to detect the threshold on colours. Also, I felt that the black and white capture looked better as the coloured shapes stand out more.

As I was having difficulty, I decided to start my piece from scratch as originally I had used the brightness thresholding example. I felt that starting from scratch would help me to solve the problems easier as I would be writing out all the code step by step, in a way that I understood. I managed to resolve the problem by adding in videoMirror and a for loop to go through each of the video pixels. Then, instead of using the normal formula of x+y*width to access the array of pixels, I set it equal to [(width-(x+1))+y*width].

Screen Shot 2015-01-13 at 10.28.16

This flipped the video capture and the brightness thresholding still worked correctly. However, this caused the an error with the falling shapes. Now, when you put up your right hand to obstruct the shapes, it obstructs them up the left hand side. This meant I had to flip the shapes too, which I managed to do by changing the x co-ordinate of the shapes to (width-1-x).

Screen Shot 2015-01-13 at 13.04.08

Carrying out user testing with my family and friends was very useful and now my piece is mirrored correctly. This shows that I have followed an iterative design process as I created my piece, got feedback and then made improvements. I have learnt a lot more by carrying out this improvement, especially in relation to the advantages and disadvantages of hacking/programming. I feel that now I have programmed from scratch, my knowledge and understanding has advanced and I’m more confident with resolving any further problems I may encounter.  Next, I will go on to explore some other suggested improvements, such as adding more of each shape and altering speeds. Then, I will go on to test my piece in the studio room, capturing full body movement using a white screen.

Processing Forum, 2014. Mirroring a Webcam in a Sketch. Processing [online]. Available from: http://forum.processing.org/one/topic/mirroring-a-webcam-in-a-sketch.html [Accessed12 Jan].

Processing Forum, 2014. Flip the Video Capture on the X-axis.Processing [online]. Available from: http://forum.processing.org/one/topic/flip-the-video-capture-on-the-x-axis.html [Accessed11 Jan].

User Testing – Family and Friends

I have presented my piece to my family and friends in order to see whether they think it will be successful and to help me gain feedback and improvements, as well as to help me decide which objects to use (shapes/stick figures/software icons). It is important to gain feedback from users as it means I can improve my piece to make it more appealing and suitable for its purpose. It relates to the iterative design process as I am engaging with users in order to help develop and improve my piece. Before showing my piece, I came up with some qualitative usability questions to ask and here are the concluded responses from the multiple users…

20150111_132825

  • Which object do you prefer?

The majority of people preferred the shapes as they thought this idea worked and looked best. There were a couple that liked the stick men figures as they thought it looked more unusual and interesting. None of the users preferred the software icons.

  • Discuss what appeals you to interact with the piece?

The majority felt that the bright colours stood out a lot and helps make the piece look more appealing. They also liked the fact it involved camera interaction and thought this worked well.

  •  Discuss what you like about the piece?20150111_132938

The majority liked the black and white video capture contrasting with the bright coloured objects and thought this worked well. They liked how they could use any part of there body to interact  and obstruct the objects. The simplicity of the piece also meant anyone could easily interact without any difficulty.

  • Discuss what you understand about the piece?

Everyone understood that it is camera interaction where you can obstruct the falling shapes. However, some couldn’t understand what the point of the interactive piece was. When discussing the concept of surrealism and the blurring boundaries between the real and the imaginary/simulated, this helped their understanding but younger family members portrayed the piece as more game-like.

  • What improvements would you suggest?

A common suggested improvement was that I could experiment with adding more of each the shapes. Also, another common suggested improvement was to make the range of the sizes of the objects more varied. Some users suggested that I could make the piece more game-like and have a point system so that if an object falls, they loose a life. They felt that if it was to be made into a game, the objects needed to fall faster in order to make it a bit harder, as well as possibly to use shapes as decoys. For example, you only loose a life if a square falls and the other shapes are just there to distract you. A big improvement that has come to my attention is that the video capture needs to be mirrored. At the moment, when you lift up your hand, the screen flips it so it is the opposite side. This makes it a bit confusing for users to interact with as the video capture is reversed.

Moores (2005, p.112) suggests:

‘the ways in which individuals make sense of media products vary according to their social background and circumstances, so that the same message may be understood in differing ways in different contexts.’

This can apply to the user testing I carried out on my family due to the varied range of ages from 12-80. Each individual makes sense of the piece in their own way and denotes different meanings and messages. In particular, I found that the younger members saw it being game-like and the older members (grandparents) did not really understand it. This can relate to Jenkins idea of the ‘digital divide’ and how different age generations access and use of new technologies varies. Due to my target audience all being a similar age (students), I feel that there will be some more common denotations from interacting with my piece. However, as Jenkins suggests it is hard to understand the complexity of audiences. Due to this complexity, I’m aware that individuals may denote the messages and concept of my piece in multiple ways, peoples behaviour may differ and my expectations of the audience may be challenged.

Overall, testing the piec20150111_132925e with family and friends has helped me a lot. I have now decided that I will use shapes as the falling objects as this was the most agreed upon response. Also, I am pleased with the visuals and think the bright colours on the black and white video capture is an element that is appealing and liked by users. Due to some of my family being younger than my aimed audience, some didn’t fully understand the piece and the concept. However in the public space, due to my target audience being students, they will hopefully understand the piece better. From carrying out this testing, it has helped me to gain lots of feedback which will enable me to improve my piece to make it more successful. The improvements were all very useful and I will take each of them on board. I need to ensure that next I edit the code to mirror the video capture as this could make it confusing for users and put them off interacting with my piece.

Jenkins, H, 2008. Convergence culture: where old and new media collide. London: New York University Press.

Moores, S, 2005. Media/Theory. Thinking about Media and Communications. London and New York: Routledge.

Initial Testing

I have personally tested my first draft of idea to see if the programming works correctly. As I am still unsure as to what objects to use, I decided to test them all to see which works best, here are the clips showing my tests:

From the test, I found that the object orientated programming all works as expected, when they are on a white pixel, they descend, and when they reach a black pixel they ascend. I did however find that in the software icons examples, they tended to fall in groups at times which is something I will have to resolve if I choose these objects. I think the bright colours on the white background work well as they stand out and make the piece more bright and appealing. As you can see from the tests I have carried out, the brightness thresholding in each example varied slightly. The threshold level in each is the same so this is due to the different time of day and the brightness of the room. Therefore, from this I have learnt that when displaying my piece, I need to edit the threshold to a level where it would work best, to suit the brightness of public space in Weymouth House. Next, I will go on to show my pieces to users, in order to gain feedback as to which object they think works best and other improvements they may have.

Processing Experiments – Popping Bubbles

When working on my piece, I decided to experiment with an idea that when the user hits the balls they disappear. I edited my existing code so that instead of ascending when they hit a black pixel, they now disappear. If I was to develop this idea, I would create it so they pop like a bubble, rather than just immediately disappearing. Here is a video clip showing the experiment:

I think this works well and with further improvements could look better. However, after talking to my audience, and showing them my progress so far they preferred my initial idea of being able to obstruct falling objects. I will continue to work on my idea, developing the programming for it.

Userbility Testing

In a recent s1419225895usability_testingeminar we learnt about usability testing and thought of possible questions to test the usability of an interactive piece example. Usability testing is when you test/evaluate a product with its aimed audience. It is important that I usability text my piece with the target audience as it will allow me to see its suitability and fitness for purpose. As well as allowing me to find out how users behave with the piece and what they experience from it (UX). Kuniavsky (2003, p.18) claims that

‘A good user experience doesn’t guarantee success, but a bad one is nearly always a quick route to failure.’

This shows the importance of carrying out usability tests, as without a satisfied audience my piece would be unsuccessful. He also suggests:

‘What makes a good experience varies from person to person, product to product, and task to task, but a good general definition is to define something as “usable” if it’s functional, efficient, and desirable to its intended audience.’

Carrying out usability tests will allow me to collect both quantitative and qualitative data to determine whether my piece is desirable to its intended audience. Quantitative data is that which deals with numbers, it can be measured and easily analysed and made into graphs/chart. An example of a quantitative question could be ‘Did you understand how to use the piece?’, the responses from this would simply be yes/no, hence allowing you to easily measure/analyse user responses. Qualitative data is that which is more descriptive, giving you individuals own personal opinions and views on the piece. An example of a qualitative question could be ‘Discuss what you like about the piece?’, hence giving you more detailed, quality answers. Quantitative questions are normally used when testing a big group of people as they could be good in order to find out who to target, however the responses aren’t usually very accurate and aren’t detailed enough for you to gain ideas on how to improve the piece and hence follow an iterative design process. Therefore, for my piece it would be best to use qualitative question when carrying out usability tests as it will allow me get more descriptive, useful answers as well as gain more feedback and improvements that I could possibly make. I need to ensure that I word the questions in the correct way in order to get qualitative responses from users as sometimes they can give yes/no responses which wouldn’t be as useful, therefore to do this I could use ‘discuss’/’explain’ in the question which encourages them to give a longer answer. From this, I have learnt a lot more about usability testing and quantitative/qualitative data. It will benefit my piece and means I will follow an iterative design process. Next, I will go on to create a prototype for my piece and test it with the aimed audiences.

Kuniavsky, M, 2003. Observing the User Experience : A Practitioner’s Guide to User Research [online]. Burlington, MA, USA: Morgan Kaufmann.