Physical Echoes @physicalechoes - Tumblr Blog

Brief video footage of Physical Echoes in the DPL

A selection of brief clips so that the external examiner can see the installation in action.

Some Physical Echoes from the past week.

Creative Technology Showcase Opening

Note: while this post might seem a little like me giving myself a great review, what I really want to get across is how blown away I was by the response I got at the Creative Technology Showcase opening last Thursday evening. Thanks to everyone who came along and made the event a success, and thanks to Sunjoy, Ewa, and David for photos.

Following a long period of waiting and gradual development, Physical Echoes finally got its first outing last Thursday evening and was well received by all. The opening event was the culmination of all the hard work put in by the MA students who finish at the end of this week, and was cause for many sighs of relief following a busy time for all involved.

The first part of week saw me transferring the installation from the third floor Digital Media Arts Project Space, into the Digital Performance Lab (DPL) on the ground floor. The equipment that I had requested all arrived promptly, and I set to work rearranging and recalibrating the piece for the new environment. The revised arrangement went together just as I had planned, although some lingering glitches and bugs made me worry for the integrity of the piece (not to mention my assessment results).

My trepidation would turn out to be misplaced, as a crowd gathered about the installation – following a suitably nail-biting wait before ‘firing it up’ – and the atmosphere turned to one of playfulness and fascination when people began making sounds and interacting with the piece. The machines’ movements, my light-hearted encouragements for noise making – helped along by some very enthusiastic and obliging participants holding wine glasses – gave the installation the feeling of a performance, something that was emphasised by my late decision to only run the machines in short bursts of 30-60 minutes.

The jovial atmosphere, coupled with the machines’ slightly ramshackle aesthetic helped smooth out some of the rougher edges in the piece. The occasional mechanical issues and software glitches could be rectified quickly and weren’t dwelled upon by the intrigued audience, making my occasional interventions relatively low-key and non-interruptive.

Different observers responded positively to different aspects of the piece, including the interactive concept, the design and operation of the machines, and the images they created. This presented an interesting triangle of enquiry regarding what I consider to be the real crux of the work. Further to this, after being told that I should sign the resulting drawings, I questioned whether or not I was really the creator of them. I was the designer and builder of the machines, but it was the machines that marked the paper, and those markings were only made in response to the sounds being generated by the audience. Whilst I can consider the installation to be a balance of all three points, I would be reluctant to jump to any conclusions about the true artist behind the newly created images.

Free-standing wooden boards for revised presentation.

Alterations for an updated presentation

At the time of my previous post I had just presented the ‘what I’ve managed to finish’ version of my installation. I had planned to follow up this post with more reflections on the development process. Instead, I have spent my time working on the documentation – as well as several other projects – and planning how to present the piece in a format closer to my original plans.

To ensure that the drawing canvases remain permanently well lit – without any chance of observers blocking the light and interfering with the camera tracking, as was the case in my previous presentation – I will be using overhead spotlights, pointing as directly as possible at each machine. I will also be using back projection screen, thereby eliminating any chance of observers blocking the projection, too. These measures should help maximise the opportunity for proximity to the installation, allowing observers to have a more intimate interaction with the piece, something that has been emphasized in the feedback I have received.

I will be making use of free-standing, wooden boards – as opposed to wall-mounted foam boards – so that I can move my equipment to the rear of the installation, allowing observers to get close to the machines. Their will also be space between the machines and the projection screen, meaning that the Kinect can also be put further back, behind the drawing canvases, emphasising the effect of the position-tracking video mixer for the projection. I’m planning to equip the drawing machines with charcoal rather than marker pens, eliminating the problem of the pen nibs drying whilst the machines are idle. If charcoals turn out to be unsuitable, I also have the option of using retractable marker pens. The machines will also be slightly elevated on a small stage, allowing observers to get a closer look more easily.

Conceptually, the piece remains much the same, although I want to emphasise and increase the potential for ‘up close’ interaction with the installation. Additionally, I want the interactive framework to be as robust and responsive as possible, following the interactive design ethos of Tom Igoe, who champions audience exploration and interpretation of a prepared interactive environment, rather than force-feeding the envisioned concept; something that I hope will produce some of the most surprising and positive responses.

I’m planning to have a localised microphone for each of the machines, rather than the previous ‘ambient’ arrangement that attempted to gather sound from the entire space. By having a microphone on each machine, I hope to make my lecturer’s fascination with the idea of having a ‘conversation with the machine’ possible. The machines’ responses will be dependent on the observers’ proximity to them. This localised interaction will also be suitable for a larger exhibition space that I am sharing with other artists.

With these changes in place, I am hopeful that I will have finally done my piece justice. Although there are still several ways in which I could expand and improve the concept - possibly collaborating with more technically experienced artists or with some additional financial backing - this will should be the first time that I can really stand back and observe people interacting with the installation. Once I have recorded my observations, I will be set to complete and finalise the documentation, no doubt armed with more questions that I can take forward into further research.

Deadline presentation

Last Wednesday, I presented ‘Physical Echoes’ to my class and lecturers, beginning with a ‘disclaimer’ talk before I set the software and hardware into action. This talk was a self-conscious reminder to the my audience that due to various setbacks – detailed throughout this blog – the work that I was presenting was not quite finished. I made sure to highlight the compromises that I had made in order to meet this deadline.

While last Wednesday was the original deadline date, our programme leader has been very generous in offering us all the opportunity to extend until early 2013. From what I know, we have all been working on quite ambitious projects, experienced setbacks, and kept ourselves busy with more than just the course content.

As an added incentive, an end-of-year show is being arranged for sometime in December or January, giving us a new target for finalising our work. In the opinion of my lecturers – and myself, sort of – my work is pretty much complete, with just the presentation format requiring some final consideration, in order to present what I have created as well as I possibly can; I ‘owe it to myself’, apparently.

I will be spending the next few weeks developing my formal documentation further; something that promises to be a lengthy report by the time I finish it off. Alongside this, I will be making a handful of tweaks – as I am able to – that should eliminate some of restrictions that I was unable to iron out in time for last Wednesday, making my installation more presentation-friendly for the end-of-year show and truer to my original aims.

I will be following up this post with a series of more detailed reflections on some of the conclusions I have come to after working on this project, expanding on points such as:

Working solo on a diverse and large-scale project – it occurred to me recently that the last time I worked on a project with a long check-list of different, diverse elements, I was working as part of a 7-strong arts collective, not individually. It has been a testing experience.

Researching, developing, learning, and creating simultaneously – this is something that I tend to do anyway, but the process has been amplified with this project. When I started, I knew that I was definitely capable of producing about 60% of what I was planning to do. I have had to learn how to do new things, have learned that certain sections of my knowledge were incomplete, and I have had to keep up a reasonable amount of momentum and progress as I have gone along, in order to make sure that this knowledge was being funnelled directly into the project’s development.

Much of this has already been dealt with in the blog, although I will be looking back with an enlightened perspective to assess the pros and cons of my approach.

Kinect Compromises

Today I've been working on putting all of the individual elements of the installation together for testing simultaneously (as mentioned in yesterday's post). I had hoped that using the Mac Mini would solve my issue of not being able to simultaneously use three webcams and a Kinect - all four will run on my iMac, but not my MacBook - however, it seems that I'm having the same problem.

As a somewhat excessive compromise, I've decided to use both the Mac Mini and MacBook simultaneously for the time being. Rather than embedding Zachary Seldess' Kinect blob-tracking Max patcher into my main program on a single machine, I've been running it on the Mac Mini, wirelessly sending OSC data to the main program on my MacBook.

While this is a fairly inefficient and long-winded method for achieving my aims, it is working fairly well. One additional benefit of having the main program running on my MacBook is that I'm able to tweak parameters on my display without disturbing the projected material.

Testing with everything running has highlighted another previously mentioned issue of blocking the canvas illumination when moving in front of the projector. This limits how close observers can get to the machines and thus reduces the effectiveness of the position-based video mixer. For the time being, I've marked a cutoff line, which is quite restrictive. Because of this, I will bow be investigating independent light sources for the canvases, although the time-frame for acquiring them will be most likely dictated by my remaining budget at this late stage in development.

One final thing I have noticed today is that the occasionally breakable machine is still faltering from time to time, which is fairly frustrating. I'll consider making a few more modifications and adjustments, I think the heavy use this week has worn down my repairs on the already worn-down hardware!

Tying Everything Together

I've finally reached a point where all of the separate elements are coming together. I've learned a lot about dealing with delays and chiselling out smaller issues but it's nice to be able to see the bigger picture - quite literally - again.

Below is a picture of the projection setup for Physical Echoes (shot at iPhone-in-the-dark quality). What's most significant about this image - although not visible in the photo - is that the running Max patcher incorporates all of the drawing machine control and automation, position-based video mixer and essentially everything required to make the installation run in its entirety! The pens aren't attached here as I was focusing on setting up the projection (the 'echoes' visible in the photo were recorded on Monday for test footage, mentioned in the previous post) but so far, everything is behaving.

Tomorrow, I'll be heading back into the space to set up a Mac Mini - finally acquired, yesterday - loaded with all of this work to hook up absolutely everything (including the Kinect) and test that it'll all run sufficiently on the machine. One issue I'm dealing with at the moment is keeping the canvases sufficiently lit in the low-light conditions.

While it was easy to project white rectangles onto the canvases, this restricts how close you can get to the machines without blocking the projection. If the light is blocked, it doesn't just affect the main projection but also the software's ability to track the pen blocks, which is crucial to smooth operation. To remedy this, I either need to impose a limit on the movement of observers, or seek independent light sources for the canvases. Raising the lights in the space isn't an option as the projector isn't bright enough to display a clear image when it isn't dark.

Despite this minor issue, it feels good to be where I should - arguably - have been weeks ago!

Minor delays and some out of context footage

Just a quick post to note that, for the past fortnight, I've been delayed a little in finishing off the installation. Following my initial slowdowns with the projector issue, I've spent some time in hospital and had to temporarily deconstruct my hardware for another event in the exhibition space.

I've spent a couple of days - slowly and carefully, on doctor's orders - setting up my work again, refining a couple of points in the process and recalibrating everything as necessary. I have replaced the reels on the slightly temperamental machine, used an adhesive to keep them better attached to the servo 'teeth' and propped the front edges up very slightly. So far this seems to have fixed the occasional breakages, which is great.

It feels a lot like I've been sidetracked by small technical details recently, I'm really looking forward to looking at the bigger picture again as I finish off the last elements of the installation.

Contrary to that notion, here's some isolated footage direct from on of three machines' webcams. As I explain in the titles, the lighting conditions aren't quite what they will be, the recording frame rate seems to have gone slightly awry, the pen is fairly dried out and I forgot to turn on the tracking automation straight away, meaning it goes a bit too far right on the first pass. I'm also being exceptionally noisy in the video (just tapping on the mic, essentially).

Without sound, the other machines, the video enhancements or the dynamic mixing, this doesn't quite capture the feeling of the piece (photos in previous posts might help) but here goes:

Position-based Video Mixer Calibrated, Faster Servo Replaced

I decided to take a break from working purely with the drawing machines today and to have a quick play with the projection-based elements of the piece. One part of this is the position-based video mixer, which tracks observers’ movements in the space (using a Kinect) and adjusts which machine the projection 'focuses' on, based on their proximity to each machine.

Given that it seems I’ll be waiting a little longer to have the projector functioning, I’ve started making use of a large (very large) television as a stopgap. With this in position, showing a mixed feed of the machines’ cameras, I have been able to measure the range of the Kinect’s view of the space and calibrate the video mixer based on where observers can be seen. I’ve currently got the ‘sensitivity’ set quite low (the full range is mapped to the mixer) meaning that to completely focus on one machine you need to get quite close to the machnes, and for an equal balance you have to get close to the back of the room; the total area tracked is about 4m². From here, I will be able to make adjustments for a more subtle effect – based on some feedback I’m hoping to get in the next fortnight or so – although I’m already quite please with this low-key, dynamic addition to the piece.

The main parts of the projection element left to be completed are the video transformation (something else that needs to remain subtle) and actually being able to use the projector, in order to get it positioned properly on the wall above the machines.

In other developments, I’ve replaced one of the continuous rotation servos that was running much faster than the other five being used in the installation. I had wondered whether this was causing the reels to come off one of the machines quite often, however this problem has continued even after the replacement. The other things that might be causing the problem – from what I can see – could be the ‘teeth’ on the servo attachment, which feel as though they may have become slightly worn down and are not keeping the reel attached very firmly. Additionally, the reels themselves – taken from the same cassette – do not feel quite as sturdy as the other four used in the installation, and tend to bend much more easily, as well as letting the thread slip into a gap between the central shaft and one of the circular sides (I’m clutching at straws with my terminology here, apologies to any engineers that might be reading this post).

To try and combat these potential issues, I will be replacing the reels and using an adhesive on the servo’s ‘teeth’ to hopefully get a slightly better grip of the reel attachments.

Presentation Format

I had a tutorial session with David Hancock today, bringing him up to date with progress I've made so far with the project. One of the issues raised during this session was how I will to present the work to audiences. Given the quite complex nature of the installation - I'm still probably the only person in the world who could configure it to run - and it's occasionally temperamental behaviour, we are in agreement that the current version would probably not lend itself to autonomous exhibition in a gallery.

In addition, the potential problem with having enough differentiation between the responses of the individual machines to varying frequency bands - and making this clear to observers interacting with the machine - has lead me to consider breaking down the different functions of the installation into multiple presentation formats:

As David Hancock saw me in action at my Manchester Mini Maker Faire exhibition last month, he considered the 'assisted workshop' style of presentation something that could be transferred to this project; explaining the concepts and technology involved to people, before allowing them to interact and create their own drawings and take them away. The sequential automation discussed in a recent post would probably suit this format best, which was how I presented to David today.

Further to this, we discussed the concept of a 'controlled situation' method of presentation, perhaps incorporating a specially engineered composition or live performance - with the possibility of collaborating with other artists - that shows of the full potential of the installation (using the simultaneous, multiple response functionality that I originally proposed.) This could form part of a split event with the workshop-style presentation, perhaps at an exhibition similar to the Mini Maker Faire or FutureEverything's Handmade.

I will be spending some time over the next few days thinking about how best to work with methods such as these before making some firm decisions on how to proceed with presenting the piece.

Machine Wiring

Today, I have mostly been working on some documentation for the Major Project submission. I thought it best to include a diagram demonstrating how all of the drawing machines are wired up. A couple of bits might seem a little convoluted, and I've met some electrical engineers who would probably freak out if they saw this, but, as I've got things laid out, this works perfectly for me:

One other thing I forgot to mention yesterday was the addition of a small abstraction in the 'main' Max program, which warns me when my Arduino is not plugged in. As it was, Maxuino would not warn me when the board wasn't connected, leading to problems when I assumed that it was and turned on the 5v for the servos! My abstraction 'listens' for the connection confirmation message from Maxuino abstraction, throwing up an unavoidable warning if that message doesn't come through.

Considering alternative operation

I’ve made a small amount of progress today (after some initial difficulties getting power in the space) whilst still trying to iron out a few small issues:

I’ve done some quick tests with the microphones to try and determine the optimum frequency bands for each machine to respond to. For the time being, I’ve got the microphones at floor level, angled upwards and in a coincident-pair stereo arrangement (photo below). I had originally planned to use a single omnidirecitonal mic, although constraints on my – almost non-existent – budget mean that I’ll be using this arrangement and summing to mono to get a generally wide coverage for the machines to respond to.

These quick tests have raised my concerns about the range of frequencies that the observers will be able to produce in the space and their ability to differentiate between the unique responses of the machines; I would have to select quite a narrow band of frequencies as it is, before separating. This being the case, I am considering alternatives ways for the three machines to operate.

Currently, the machines all run simultaneously, whilst also waiting for each other to begin moving between the two horizontal points to ensure that they remain in sync. If the machines are not going to respond differently but still behave individually, I could make them run sequentially, waiting for an adjacent machine's pen block to reach the edge before moving (see animation below):

I quickly implemented this function today and it behaves fairly well. I am not completely abandoning the originally-intended function just yet, however I will divert some time in to developing this idea as a backup option; maybe even presenting it as an alternative version of the exhibition. The only thing I’d need to consider with this version would be how to adapt the dynamic projection. Something to think about.

Roundup of the week

The projector is up but the lamp needs replacing. It could take some time to sort out so I'll be working with a large television/display for now and waiting to see what happens (more requests passed along a long chain of staff).

The power supply is functioning perfectly, sufficiently running all 9 servos without any real issues so far. Despite this, one of machines seems slightly faster than the other two i.e. it reaches each side faster. As a result it is prone to breaking slightly (the reels fall off after extended use). I'll be looking into this, I think it could be an issue with friction on the pen block rather than anything to do with power, although I'll explore the problem to make sure.

The automation program that keeps the pen blocks moving between two horizontal points has been improved and properly tested, with measures in place to make sure the tracking doesn't go awry after extended use. The basic audio to pen movement program from the previous incarnation of the installation has also been incorporated into this 'main' program now, simply controlling all 3 of the pen servos in unison for the time being.

Now that these pen servos are all running without problems, I will be working out the optimum frequency bands for each machine to respond to very soon. I had planned on positioning the microphones at ceiling level this week but have discovered I'll need a little more equipment to do so. Once they are in place, carrying out some audio tests will be much more straightforward.

Soon, I will also be experimenting with the options for video transformation that I assembled a while ago. Now that the machines will actually be drawing again, I can use some of the image 'isolation' ideas - that I researched and worked on at the beginning of the project - in practice, hopefully finding a suitable method.

I now definitely have use of a Mac Pro in the space - which will have to do until a Mac Mini is available - meaning that I should have no issues using all of my cameras and Kinect simultaneously. As I will be working on the projected element of the piece next week, I will try and merge both the position-based video mixer with the video transformation program once they have been calibrated for the space.

Interruptions and waiting for a third party

A couple more setbacks have cropped up this week. Firstly, I have been informed that I have to wait for a third party - the Balfour Beatty construction group - to assist with mounting the projector. Whilst this seems utterly ridiculous to me, I am at the behest of the university when it comes to this and must continue to wait.

Additionally, today I was interrupted by workmen whilst I was working on the installation in the exhibition space. I am not sure whether these men were from Balfour Beatty, The Peel Group or the university, but they were carrying out repairs of some kind in the ceiling, completely interrupting my workflow and effectively forcing me to end my day early. I have made an enquiry with my programme leader to find out more about this.

On a more positive note: the power supply I have acquired and modified to give power to the servos in my installation is behaving perfectly, ensuring that the Arduino can focus on controlling servo movements and stop randomly resetting when they draw too much power. My thanks to fellow student John for his assistance in soldering.

Minor problems and solutions

Since getting into the space, I have experienced a handful of issues that have slowed down the development of my installation. Despite this, I have been doing some research and acquiring new knowledge in order to try and solve these problems:

Problem 1: At the point where the floor meets the wall, there is a small skirting that would obstruct the movement of the drawing machines’ pens on the vertical axis.

Solution: To ensure an entirely level surface for the pen to move across – as well as for the pen block to run easily against – I have acquired three A2 foam boards (3mm thick) to attach to the wall, covering the skirting and sitting flush to the floor.

Problem 2: Rather naïvely – having never used large amounts of components before – I have experienced problems when trying to power and control all of the drawing machines’ servos simultaneously. I had been using the 5v pin on the Arduino to try and power them all, unaware of the current limitations involved.

Solution: After doing extensive research, I have finally found a suitable power supply that should be sufficient to power all of the servos used in the installation simultaneously. All that remains is to do a small bit of modification to the switching AC adapter I have purchased (DC 5V 8A) to make it suitable for connecting up to the setup I already have. (Now complete, pictured below.)

Problem 3: I have been having some trouble supplying sufficient power to all of the video input devices used in my installation (three webcams and a Kinect), despite using a powered hub.

Solution: As I have successfully powered all four of these devices before at home – using my iMac’s built in ports, rather than using my MacBook with a powered USB hub, as I am doing in the space – I am reasonably confident that the both the Mac Mini’s built-in ports and my hub should be sufficient to power them all. I am currently looking into ensuring that I have this machine available for the final installation, with sufficient time to test the setup.

Problem 4: I went into the space earlier on this week intending to ceiling-mount a projector. Unfortunately, the mounting equipment available would have been unable to put the projector at the height I require.

Solution: After a brief exchange of ideas with my programme leader and a member of technical staff at the university, I suggested a method of modifying the existing equipment that should make it suitable. I am currently waiting for word from the technical staff who are attempting to implement this modfication.

*Hopefully* all of these solutions will be in action by next week, allowing me to continue developing the piece. At the very least, I have learned some technical lessons over the past fortnight or so.

In the Space

I've finally made it into the space...and discovered a handful more issues. I'll be consolidating and expanding my documentation over the weekend while I wait for some more parts/advice in order to proceed.

This was a lot of stuff to carry on public transport:

Now that I have the space, I can iron out a few bugs:

Trending Blogs

Recently Viewed Blogs

Physical Echoes