Keiichi Matsuda

Leave a comment

Starting to reinvestigate Augmented Reality for my masters project I came across Keiichi Matsuda, an Architecture graduate who has created films that look at how we might see things around us in the future.

What is most amazing is that he started visualising this AR enhanced world 4 years ago…

Augmented (hyper)Reality: Domestic Robocop

The latter half of the 20th century saw the built environment merged with media space, and architecture taking on new roles related to branding, image and consumerism. Augmented reality may recontextualise the functions of consumerism and architecture, and change in the way in which we operate within it.

A film produced for my final year Masters in Architecture, part of a larger project about the social and architectural consequences of new media and augmented reality.

From the website

Cell is an interactive installation, made in collaboration with James Alliban. Commissioned by Alpha-ville for the 2011 festival, cell plays with and proposes alternative landscapes in the technological ether surrounding our everyday movements. As our identities become deliberately constructed and broadcast commodities, our projected personae increasingly enmesh and define us. Cell acts as a virtual mirror, displaying a constructed fictional persona in place of our physical form. Composed from keyword tags mined from online profiles, these second selves stalk our movements through space, building in size and density over time. The resulting forms are alternate, technologically refracted manifestations of the body, revealing the digital aura while simultaneously allowing us escape from our own constructed identities.Cell uses Microsoft’s Xbox Kinect to track visitors as they interact with the installation. It was built in openFrameworks, an open-source toolkit originally built to teach artists and designers creative coding. Microsoft have supported the project from the early stages, working with Brighton based company Matchbox Mobile and the openFrameworks community to build a new code library (or addon) specifically for cell that supports the Kinect For Windows SDK. This is an important development in the field of interactive art. Providing openFrameworks users easy access to the official Kinect For Windows SDK, places the technology directly into the hands of a large international community of interaction designers and new media artists.

ASU1 – extra

1 Comment

Reviewing the footage late last night, I really wanted to have another go at 3D motion tracking, I had the inspirational idea that I would like to put the title of the film, ‘Just one more place to look’ within the scenes themselves and for the character to walk past them, so have spent all morning, trying to make this work.

In After effects there are many ways of tracking, you can track the motion, the camera, or 3D track either, so I have gone through each of the parameters to attempt to track the background live action scenes in my film, I started with the rotoscoped section, but AE couldn’t even start to analyze that and then brought the original live action sections into AE, thinking that would be better, but although it can track in 2D, I’m mostly doing that by hand, and the 3D doesn’t get it at all… see screen below for what I’ve been seeing all morning!


But still wanting to add the title of the film and to get some more motion tracking in, I decided to hand track the final fade out and put the title at the end.

This is again reinforcing the cyclic nature of the endless walking, round and round, with the perpetual ‘one more place to look’ motto for my main character, I think it’s an improvement on the original, but this is where another crit would have come in handy!


….and no Youtube I would not like you to stabilise my shaky footage – it’s supposed to be like that!

shaky foot

ASU final film

Leave a comment

Here is my final moving image piece for the ASU1 module of my MA.

Roughly titled ‘Just one more place to look…’

 I couldn’t leave my film alone on the final morning, so  please also view this page for a final update that I squeaked in just before the deadline!

ASU Film complete? – maybe

1 Comment

I am just rendering out my final film for this ASU module, I’ve re-recorded the voiceover and as well as adding in an extra phrase I have separated the verses of the poem out so that they sit in a better rhythm with the visuals.

Tweaking the visuals in after effects to lessen the greyed out sections so that the viewer is not too disorientated.

When analysing the film today I decided to remove the section where the character walks away from us down the small alleyway, it just didn’t fit with the look and feel of the other sections, so not to totally discard the hours rotoscoping it, I used this as my end credit sequence and motion tracked the walk in After Effects with my name and the piece of music that I used in the background of the piece.

ss_tracking ss_trackingText

I really like the way the film looks, I love the style of the animation itself, all sharp lines against the out of focus bokeh of the background.

Hurrah it’s finished rendering.. oh blimey it’s come out at 21GB.. better re-render that then…

I’m not sure if I want to put in tracked text sitting in the background, firstly against the poem I don’t know what to put, even though I really like the idea Mark Wilsher had for using the narration as the main character’s internal monolgue and the appearing text as another voice, or question, but secondly I don’t know if it needs it now the extra narration is there.

If I have time I may do a short test, I only envisaged the text being one or two words at the most and now that I’ve tracked the credits it will be a similar process, although it does add more nested comps into the equation…

Left to do…

Evaluation proper

Learning Agreement

Hand in!

Alpha layers, sound and final feedback

Leave a comment

After finishing the digital ‘cell’ animation (all 1200 of them) I put them back into after effects into separate sequences, which I bring into a top level composition to assemble. It was at this point I realised that one section of live action had not been rotoscoped… Unfortunately with the final feedback session due the next day, I knew I would not be able to turn the extra 138 layers into vector animation in time for the crit, so decided best plan of action was to carry on with the draft assembly, knowing I would complete the rotoscope animation later.

Next I assembled all of the alpha png files into separate comps to overlay over the full cell animation so that I could fade the background in and out, adding to the film another level of disorientation.


Making the alpha files is a fairly easy thing to do after I have done the full cell illustration, it’s just time consuming… I isolate the alpha layer and save it out as a .png, which supports an alpha layer when brought into other adobe software, and again, repeat, repeat repeat…


another 1200 files later and it’s ‘comp’ped up into after effects, ready to play with.



I had tried and tried to create a script for this film, but was unhappy with the results… After reviewing the footage against the little bit of music I felt it needed a poem or a short narrative voiceover and after playing with cut up poetry, looking on all manner of poetry websites, finding snippets of prose that I liked, sometimes just words, or figures, or rythmns, and after much deliberation I have been able to produce a piece to use for my voiceover…

Walking, walking not quite alone

Searching, looking on the way home

a shadow of me both revealed and lost

my life is split, no matter this cost

This route I take I know it well,

 if my memories could speak what mine would tell

of me,

or me,

of both of us

lost in cycles spiralling down

our fate is framed by constant frown

The call we answer isn’t real, it’s in our heads, it’s what we feel, day after day we tread this road,

 our never ending mental load.


Appearances can be deceiving, eyes, feet, walk, all misleading.

Life was once bright and clear,

but now I trudge along in  fear

of who and what and where and when

we know our lives can’t fly again.


Now released into the wild

our minds are dark, unfurled, defiled

but walk this way we must repeat

 in case our life we find replete.     TT 2013

I worked on the sound in Logic Pro, I recorded my voiceover at home using my ZOOM h1, rendered out only the audio layer from the original scenes in Premiere and imported this live sound from each section, to this added the track from bonobo which I had in mind from the very beginning, I wish I could have recorded my own version, but time was running away.

Screen Shot 2014-01-12 at 15.50.14

I quickly output a movie file from aftereffects of the visuals so I had those to work too for timing, but I only had time for a quick mix, but after outputting my bounced file into after effects it was time to fade the background visual in and out – this is only experimentally done here – check the sound, check the timings were correct and render a copy for the feedback session.

The first render…


The feedback I received from this were mostly to be expected, the sound mix was extremely low and I think people found it hard to hear, but they liked the tone and inclination.  The visuals – I thought – looked great, actually people didn’t really feed back what they thought about how it looked, ie, the style. Comments were mixed with regards to the fading in and out of the background, mostly positive, but the fades were testing it here, so all good. I need to look at the voiceover itself, the poem I think is too short and as a viewer you’re left too long without the narration, so I’m hoping to add another ‘verse’ in, which should fit the timing better.

The text I had quickly added was disliked, even though the intention was to track the last line of each phrase onto the screen, but it was felt it would work better perhaps as a conversation as the poem is written from the character’s point of view, so this could interact with the the audio…and to try not tracking the figure, but for the text to be on a wall, which the figure walks past…

I will be implementing most of the changes, all positives to my finished film, time is short, but I am focussed on getting the best result out of this, although I am slightly worried it may end up as a bit of an ‘art’ film…which is not me at all…


Leave a comment


So much to do, I’ve borrowed another computer,  one renders whilst I work on the other for the bit of rotoscoping I forgot!

76 hours and counting…


I’m at a point now where I am just about ready to import my animated frames and their alpha counterparts into after effects, it’s been somewhat of a slog and I’m still not there yet, but, I thought I would quickly just stop and look at what I had done.

so far,

1148 full combination vector artwork plus live video… done

380 alpha vector images…done

still to do 768 alpha

if I use my rough calculation of about 2 minutes per image, totalling 2296 separate images when all are exported, that totals to 4592 minutes…

= 76.5 hours…

wow, that’s an impressive time total… but as the title states, and counting…!

and that’s not including the 4 nights of filming… editing… and still trying to write some sort of a script…

first full rotoscope sequence

Leave a comment




After the short test I was so pleased with how it looked I wanted to get a full 30 second scene totally animated, concerns over whether it would work for a longer section were on my mind, considering the amount of time I would need to put into a full test, but I decided to put the time in as I have ideas of overlaying the digital ‘cell’ animation over separate and different sections of the film, so it should be worth it.

To start with I edited down the section I wanted in premiere pro, this ended up being a 36 second piece from the subway footage. Then I imported that edited section into Photoshop getting it to render every third frame as a separate layer resulting in 300 layers and to each and every layer I undertook the process below.

Stage 1 – layer control

first of all I added a new layer above the layer I wanted to illustrate, so that I can use my digital cell with and without it’s original background, working in a  non-destructive process.


Stage 2 – vector path



Then on this empty layer using the vector pen tool I trace around the area I want to fill with colour, obviously starting with the full outline shape, which will be a lighter colour.



then the inner shadow areas, which are to be filled with a  darker colour. I work with a wacom tablet and pen for this quite precise result and like the clean edges it gives you, you have to make definite decisions about where the light and dark will fall on your digital graphic. I have used this style of vector drawing for years, usually in Illustrator to great effect.

Stage 3 – shade and light fill



Then you select all elements in the path that you want to appear as shadows, make it active in the path tool toolbox and fill it in using your darker colour and the paint bucket tool.



then select the lighter area, make it active, which will switch the areas and fill the outline shape with your lighter colour, occasionally small dart areas didn’t fill and would need to be done with a little more accuracy, but it’s easy to spot.



Stage 4 – save out



Set photoshop up to save as into the same folder, to minimise clicks, and save the ‘cell’ layer with it’s film layer as a single jpeg, bearing in mind the finished number would need to be 3 or 4 digits so start at 001, or 0001, as this will help keep them in the right order when you bring the images into your video editing program to put back together as moving images.

Stage 5 – repeat and repeat and repeat and repeat



I found on average it was taking me 20 minutes to complete this whole process for 10 frames and over the week I managed to fit in the 10 hours it required to digitally ‘cell’ animate my 36 seconds of footage.

I did split the photoshop file into 2 sets of 150 (original layers) at one point as with over 400 layers and 2.3Gb is was making my computer slow a little…

But, once they were all transformed into jpegs, I opened up After Effects and imported the 300 files, selected them all and right clicked to create new comp from selection ( this is where the numbering comes into its own) and after a bit of trial and error with how I selected them all (it did at first run backwards 😦 ) they came in perfectly all on seperate layers and for the right duration and this is how it came out…

During the long hours spent over my wacom I came across a piece of music by Bonobo called a beginning, an ending and it had the right mood, so I placed it over the video to give more of a feel to the piece.


Rotoscoping – a first test in Photoshop, After Effects

Leave a comment

After the crit yesterday I wanted to dedicate a morning to see how far I could get with an illustrative styled rotoscoping technique.


This is the still I started with and after setting up all of my shortcuts to minimise clicks, it took me 2 minutes a frame, from start to finish, and here is the result.

I am really pleased with how this looks, love the style and have realised that I have been visualising figurative work with this technique – flat colours/hard outlines – since I was last at college and found great success with painting and drawing, life drawing in particluar.


ASU shooting recce #1

Leave a comment

So decided to stop going round and round with the words and just to go and physically shoot some of the locations I really want to use, to see what happens, expected and unexpected!

First obstacle was to set up my steadicam rig, I bought a second hand merlin steadicam last year and found it tricky to set up with my camera and lens, just due to the sheer weight of the thing, my nikon + pro lens weighed in at 4.5lbs, which you obviously have to offset with the weights on the steadicam itself, I got some lovely footage after nearly 3 hours of fiddling, but my arm/elbow was absolutely buggered! all for maybe a minute of test footage. In the end I had used my 50mm prime lens to make the whole rig lighter, but of course this meant I was unable to record very wide shots… With this in mind I decided to get hold of a friends Nikon D3100 and just go with it’s kit lens, making the camera rig just under 2lbs – a good start – and again 2 hours of set up time to get it close, it was still way too wobbly, but I had to get out there and shoot, so although not as stable as it could have been, I decided to go with it and refine the steadicam later.

Steadicam settings used – mid=1 finish weight, end= 1 start, 1 mid and 1 finish weight, stage hole ‘N’ -3 mark for plate, 5 turns on the handle, fully opened levers.

Second obstacle, I needed a body! fortunately one of my friends was happy to help, although slightly surprised when he asked who we were filming and I said you!

Third obstacle, the weather, fingers crossed it wasn’t raining and we got lucky with that although it was absolutely freezing!

I was very happy to set off with just a good friend and my small photographic rig, this meant on the whole we could be very discrete with our filming and easily moved or relocated if we had too many problems. I only wanted to use the available lighting…

As we walked into the city, my friend suggested a couple of places he liked, so we went and had a look, unfortunately it was way too dark and this was when I had my first unforeseen problem, on this little Nikon, I couldn’t manually control the video settings, so although I could take a fair photo in the streetlit footpath, I couldn’t get the same results when filming… very frustrating that the camera had limited night video settings… I would have to hope that the other locations were brighter…

Also as the steadicam wasn’t perfectly balanced I would have to be super careful not to make anyone viewing the footage sea-sick! it’s a real art to make it smooth, one that I had almost got to grips with by the end of the recce shoot.

The next locations, the alley off of St Stephens was way too dark , but  the church in chapelfield and the market were good and bright, so the shots there are better, the market shots I am particularly pleased with, even though a couple of inquisitive teens on bikes, did make retakes necessary.

The subway underpass was also amply lit and steadicam wasn’t as steady as it could be, but the footage is on the whole good.

The final shot of the little lane off or Newmarket Street was where disaster struck though, the freezing cold had absolutely zapped the battery and I managed to get a shot ready to take, when it completely died and I couldn’t film a thing in my favourite looking location… Undeterred though I decided to use my phone’s video camera, just to test out shots etc, because I didn’t want to waste any more time and we were both there and the lighting was almost perfect, the cobbles were shiny from earlier rain and the lights looked as good as on my photo recce.

So here are the best shots quickly cut together with 3 of the smaller parts slowed down.

I hadn’t thought about sound at all, but I’m really intrigued with the slowed down sound that accompanies the shadow shots, I really like the way that has turned out, the effect it gives is perfect for the feeling of the film…

It is really frustrating that the dslr battery died on the last shots as my mobile phone couldn’t keep a still focus, but the way it looks up to then is how I imagined it in my head.

Looking at the footage now I know I need to refilm with a camera where I can control the video settings more, try and get a spare battery for it, balance the steadicam better, use more close up shots and I think this can be achieved by using my pro dslr with the 50mm lens on it, it will give tighter shots overall, like in the market scene, which is my favourite of all captured footage so far.

I also realise watching it back that I can’t just have someone walking in front of the camera to tell the story so an accidental bit of footage I recorded whilst walking through the underpass shows me that perhaps switching viewpoints to the character’s point of view could be useful in telling the story of the search…

Next steps, thinking about sound, introducing those extra pov’s,  finalising the script for the story and a full reshoot… not much at all then :0

Older Entries Newer Entries