Erik Sanner Home Visual Other About

the way forward (nakasendo)

a few weeks ago i was working on a way to make a dynamic video piece with some of the video clips i took during matthew-and-my walk from kyoto to tokyo. we had always intended to make some sort of collaborative new media art out of our experience. i think we're close to being able to make something which will enable both of us to want to work on it, convey something of our journey, and pay homage to hiroshige. matthew and i are talking tomorrow hopefully so i'll know more then how he feels about this latest iteration of something-we-might-want-to-do.

anyway the technical aspect of it is that i want to be able to combine video in a way which has been in my head for a long time, but which i haven't seen people using. and i wanted to figure out how to do that. first i looked at some of the existing built-in ways you can combine layers of video in flash, to make sure i wasn't making it more difficult than it needs to be, just seeing if maybe i could use those built-in settings and achieve the same (or a suitable) effect:

none of those options were really what i had in mind, so i tried to figure it out another way. either i wasn't doing it right or flash is too slow and too processor-intensive to do what i was trying to do. here's a still result from as close as i was able to get to achieving the look i wanted:

basically what i want to do is ask every pixel, at random, to choose whether to show the top layer of video, or the layer below that, or the bottom layer (combining three layers). in the above image, every third clump of pixels (not at random, and not single-pixel granularity) is showing the corresponding video layer, shifting over during the next frame, and so on. this is something which took me many hours spread over a few weeks to do, but it still wasn't what i had it mind, just as close as i could get it with the time and know-how i had. so i was very excited when on the second day of the processing workshop, will figured out how to do it in processing (exactly what i had wanted to do), and walked me through it. here's a still from that result:

this is something i have wanted to see (and create) for a long, long time. however, for the piece, conceptually, i wanted something more than just random-moments-in-the-walk. what i want to do, and need to talk to matthew about, is include something of the longing and the now-that-we've-been-and-it's-over distant almost dreamlike-wistful quality of the memories of the walk. the training was as time-consuming and as intensive, in its own way, as the walk itself. so i want to do some of the training walks again, and i know matthew is still walking - i want both of us to start shooting little video clips, and integrate those into the piece, which would combine footage of my walks (probably all in and/or around new york city) and his walks (probably all in and around christchurch, new zealand). with luck we'll clarify between us tomorrow night if we're interested in pushing forward in this direction or if we want to do something else. either way i'm excited to feel like progress is being made and something wil come out of the nakasendo project this coming year.


this past weekend, spent both days learning a bit about processing, an "electronic sketchbook" and an "open source programming language and environment for people who want to program images, animation, and interactions."

i found out about the class from the dorkbot nyc mailing list.

the workshop was held in teacher will's apartment, as a fundraiser for parts and crafts, a collective which teaches kids (and adults), believing "building, playing, and experimenting are the best ways to learn, but, beyond that, we believe that they are also
the best ways to live."

a picture from saturday, will's computer is hooked up to the tv so we can see what he's doing:

a screengrab after making a program which would draw lots of randomly-colored circles of random translucency and move them around in random directions at random speeds:

when judson and i collaborated, he kept telling me i might really like processing, and be able to do some of the things i want to do in it, maybe easier (better) than using flash. i was really glad to brush up on my programming and start seeing what can be done in a different environment. i don't know how comfortable i'll be with processing outside of the context of a class, or if it will replace flash for me, but i like having the option of using it, and i want to learn more.

the occasional collaboration with isaiah

maybe twice a year (seems like becoming a regular thing) isaiah and i end up in the middle of nowhere with a bunch of people and we end up putting a short video together. here's a still frame from something we made two weekends ago. always low pressure, very free-feeling, ultimately fun.

installed and viewable

color study is about combining additive color (e.g., luminous projection):

with subtractive color (e.g., pigment):

to understand what the different combinations look like (e.g., yellow video projected onto orange paint vs. orange video projected onto yellow paint).

color study is up and you can see it for the next month or so, here are details:

installation went like this:

put up the projector:

projected lines where i wanted to paint the colors:

got some help painting:

which resulted in only minor mishaps:

and eventually a completed painting:

the next day, after it dried, i took a picture of the painting:

then tried to match up the colors as best i could, and removed the background:

put the picture of the painting into a video file where it rotates back and forth:

the video of the picture of the painting gets projected back onto the painting, completing the piece:

if you've never seen color study, i hope you make it to taller boricua, and i hope you enjoy it. again, here are exhibition details:

november 20, 2009 –january 9, 2010
tuesday through saturday 12-6pm,
thursday 1-7pm
monday and sunday closed

taller boricua galleries
1680 lexington avenue, nyc
(212) 831-4333

6 train to 103 street

accessible for individuals with disabilities
free and open to the public

the shifting definition of my role in my artmaking

(and the skills i need to build.)

all art is arrangement, right? music = sound arranged through time, sculpture = form arranged in three dimensions, painting = color arranged in two dimensions, etc. and typically a painter will have a palette and a brush and a canvas, and recently many of us also use computers and video and so on, so now we've learned how to use some digital tools to go along with those in the brick-and-mortar universe. now, all of a sudden, unrelated to any brick-and-mortar-world or digital-world drawing tools, i need some math.

when i first started programming, i realized that math would start coming into it. but i never took to math classes, and i usually avoid math as much as possible. however i realized it will come in handy to start learning some as i was working on a new version of "color study."

"color study" consisted of stripes of painted color, with an image of that painting projected right back on top of itself, rotated ninety degrees, forming a grid. the grid shows every combination of primary and secondary additive and subtractive colors.

for example, you see projected yellow on top of painted yellow, projected yellow on top of painted blue, projected blue on top of painted yellow. the plan was:

here's the result of that, an image of a previous version of "color study" (picture of an egg tempera painting rotated ninety degrees and projected back onto that same painting):

showing a new version soon and wanted t
o make some changes (improvements i think). first, added brown. next, wanted to change the order of the colors. i don't like that the white is on the edge (in the version shown above, painted white is left edge vertical stripe, projected white is bottom horizontal stripe). here's the new order:

also i began toying with the idea of painting a pure painting and projecting a pure projection next to the projection-onto-the-painting grid - above in that same image you can see some possible layouts i was playing with.

so far, so good, role of the artist still "thinking about what to make" and "arranging the stuff to make it happen," all within the realm of what i expect to be doing when i'm working on a piece. realized that given time constraints and the scale of the piece (for the new version, i want to make it much bigger, but i won't have as much time as i'd like), it won't work to use egg tempera or oil paint. so, i went to the hardware store. it felt funny to let someone else mix the colors for me.

but "wanting to learn some math" is even weirder than "hey here i am asking other people to mix colors for me." the moment-where-i-thought-uh-oh-i-might-need-to-start-learning-more-math came after i decided to animate this version. at a crit several months ago, it was obvious that people understood the piece better if the projection was rotating. so, i wanted to rotate the projection (and back) continuously - simple enough. but i also didn't want any areas of the painting to be projectionless. this meant increasing and then decreasing the size of the projected stripes as they rotate - also no problem. however, i realized the rate of increase (and decrease) couldn't be uniform if i were both to avoid projectionless areas and keep as much of the projected color on the painted color as possible.

here are some illustrations of what i'm talking about. in each of these images, the 50% transparent square on top represents the projected square, and the 100% solid square beneath, always with perfectly vertical stripes, represents the painted square. when the actual piece is made, the projected square will be 100% solid, and the other square will be painted in the real world.
this first image shows a square rotated, without increase, showing the projection gaps i want to be filled up, which is made possible only by enlarging the square.

getting rid of the projected corners hanging off is no problem, i can just mask the projected square

to get rid of the projectionless corners, i need to enlarge it, also no problem

cropped, looks like this, exactly what i want, no problem so far right?

the problem comes in the early stages of rotating the top (projected) square. the rate of increase of size can't be uniform, it has to increase sharply at first and then taper off as the rotation approaches forty-five degrees, otherwise there are projectionless gaps early on. so i can do it by hand, and it works, and you see the "artist's hand" in the work, no real problem. but i'd like to know how it makes sense, how to make it work, mathematically, so that i can program it. right now i don't even know what question to ask. when i google "how do i rotate and enlarge a square so that it continues to cover a square beneath it?" nothing too useful comes up. i think i need to find someone who is used to teaching geometry. i think i need to start learning math. this is not really something i anticipated when i realized that more than anything else, i want to spend my days making art. surprise!

how we made "reflecting above"

today started putting together a presentation about the piece - maybe i can do an artist talk tomorrow after the screening - if not, i'll figure out a way to post it online

tomorrow details:


made some more shapes to add to the piece today. started by tracing shapes in a still from the video i shot and layered of lisa's installation:

then traced shapes from some of the pictures kazue and i took last week of our second test:

then traced shapes in documentation photos of lisa's installation:

which gave me these shapes to add to the piece:


hadn't really done any "new" coding for quite some time.

but it was great to sit down and be doing it - it's funny how sometimes you're stuck, so you put it aside and do something else, then with your mind still sort of vaguely aware there was a problem you had or you weren't sure what else to do, it becomes very clear, precisely when you're not really paying attention to it.

anyway what has been satisfying is that i came across a few problems, and i figured out how to solve them, but i didn't resort to google or asking friends for advice, i just said "i want to be able to do this" and found a way. end result isn't efficient or "professional" but it works and it doesn't crash and it's doing more than i had originally wanted/intended (in a good way).

this is all for the lisa kellner / kazue taguchi collaboration (which i'll post more about another time) - the virtual component. first step was drawing some shapes with curves:

i wanted to have these revealing some video i shot of lisa's installation:

problem was, a bit more of the video than just the shape was showing up - actually a really cool effect (which i would probably have trouble creating if i were trying to make it happen intentionally, but it wasn't what i wanted and in fact was, in this instance, specifically not what i wanted). here's a still:

and a video (looked cooler in motion):

turned out that wasn't programming-related, i just wasn't importing the shapes from illustrator into flash correctly. relief.

the next day or the next next day or maybe the day after that, i had it pretty much doing everything i wanted it to do, here's a short video clip:

but after a minute or so, i would get a scary error message:

every time something doesn't work, i get a bit nervous, i imagine i won't be able to fix it, or maybe nobody can it's just not fixable. and sometimes that's the case, which means - oh man - i have to do something else. but more often than not, it's just a matter of paying attention and trying to figure out what's going wrong. so i looked through my inefficient code and i found a few things which looked like they could cause problems, and i changed them, and now i'm all set, ready for the physical installation on thursday at the brooklyn arts council gallery. there are still a few things i want to play with (for example, i want to add more shapes, based on the stills kazue and i took last week). but i'm happy with what exists right now and it works, no errors, doing exactly what it is supposed to do. here's a screen-capture still:

looking forward to sharing it properly on thursday.