A&HA 5128: PERIODIC UPDATE #8

Artifacts To Date:

My generative textiles are fabricated and in from WOVNS! I had three pieces made according to the algorithms that I have shared in previous posts.

This slideshow requires JavaScript.

To clarify, the textile manufacturer does not accept code files from Processing, but rather I ran each program many, many times to create several variants of the graphic, and then I selected versions that I wanted to have made. I sent three .png files to WOVNS and they take care of the rest after confirming my desired colors.

I am initially underwhelmed by the way that the colors translate in reality compared to the screen. I really should not be surprised, but since this is my first experience, I did not realize that small change in color choices would not actual materialize in textile form. My favorite graphic had some color variation that is apparently too subtle to show up when woven.

The company does offer swatch blankets that express the actual color palettes fully so you know what you are really getting. Perhaps if I decide to continue such work I will need to pick those up.

57e23a2538a5d88c22c0ae85362f0cf6_original

Advertisements

A&HA 5128: PERIODIC UPDATE #7

Artifacts To Date:

As planned I was able to design one more algorithm for a third piece. Here it is:

Form 11-16-2017 at 22.09.00s

And it’s code:

// "Regular Vertical Bars with Random Squares" | Dylan Ryder | December, 2017
// Talma Quality, Pastel 2 Palette
size(3984, 3000); // 46-48" x 36", 84 DPI
background(#F9FAF9); // Blizzard
noSmooth();
noStroke();

int w = width / 30;
int h = height;
int h2 = height / 10;

// large columns
fill(#FFE4F3); // Pinking
for (int col = 0; col < 30; col++) {
 rect(w * col + w / 8, h * 0.05, w * 0.5, h * 0.9);
}

// small columns in 2 rows
fill(#FFEEF8); // Tutu
for (int col = 0; col < 30; col++) {
 for (int row = 1; row < 3; row++) {
 rect(w * col + w / 8, (h * 0.9) / 3 * row, w * 0.5, h2);
 }
}

// small squares randomly placed in grid
int w3 = width / 60;
fill(#877193); // Beta
for (int col = 0; col < 60; col++) {
 for (int row = 1; row < 8; row++) {
 int rand = int(random(7*row));
 if (rand <= 3) {
 rect(w3 * col + w/8, (h * 0.7) / 7 * row, w * 0.5, w * 0.5);
 }
 }
}

// generate a unique filename for each generation based on datestamp
save("Form "+nf(month(), 2)+"-"+nf(day(), 2)+"-"+year()+" at "+nf(hour(), 2)+"."+nf(minute(), 2)+"."+nf(second(), 2)+"s"+".png");

Adjustments to Timeline:

I sent generative prints of all three works to WOVNS to be fabricated on November 17th. I should have them shipped out in around 10 days.

A&HA 5128: PERIODIC UPDATE #6

Artifacts To Date:

I worked this week to rebuild my Scratch prototypes with the Processing language, which is best for creating graphics that can be translated into textiles on a jacquard loom. WOVNS requires all colors to be expressed with specific hexadecimal values in any selected palette. Coding with that kind of color specificity is not possible with Scratch.
The first rebuild used a different algorithm to achieve the desired generative design. Here is an example generative design:

Form 11-13-2017 at 22.08.52s

Here is the code:

// "Two Agents Crossing" Dylan Ryder, November 2017
// Designed in Processing
// Talma Quality, Pastel 2 Palette

size(3984, 3000); // 46-48" x 36", 84 DPI
background(#F9FAF9); // Blizzard
noSmooth();
noStroke();

int side = 100;
int x1 = 2000;
int x2 = 0;
int y1 = -100;
int y2 = height / 2;

// Vertical Agent
fill(#877193); // Beta
while (y1 < height) { 
 // Y axis traveling
 int r = int(random(7));
 for (int i = 0; i <= r; i++) {
 y1 += side;
 rect(x1, y1, side, side);
 }
 
 // X axis traveling 
 r = int(random(3));
 int shift = int(random(2));
 for (int i = 0; i < r; i++) {
 if (shift == 0) {
 x1 -= side;
 rect(x1, y1, side, side);
 } else {
 x1 += side;
 rect(x1, y1, side, side);
 }
 }
}

// Horizontal Agent
for (int x = 0; x < width; x += side) {
 pushMatrix();
 translate(x * int(random(2)), 0);
 fill(#BFCDE5); // Moon
 //fill(#74AED7); // Ciel
 rect(x2, y2, side, side);
 popMatrix();
}

// generate a unique filename for each generation based on datestamp
save("Form "+nf(month(), 2)+"-"+nf(day(), 2)+"-"+year()+" at "+nf(hour(), 2)+"."+nf(minute(), 2)+"."+nf(second(), 2)+"s"+".png");

The next rebuild also used a slightly different algorithm to achieve my desired generative design. Here is an example of that generative design:

Grid 11-16-2017 at 22.26.37s

Here is the code:

// "Squares in an Imperfect Grid with 8-bit Fill Variability"  
// Dylan Ryder, November 2017
// Designed in Processing
// Talma Quality, Pastel 2 Palette

size(3984, 3000); // 46-48" x 36", 84 DPI
background(#BFCDE5); // Moon
noSmooth();
noStroke();
color c = #F9FAF9; // Blizzard
fill(c);

int w = width / 30;
int h = height / 25;
int r; // horizontal random gap between squares
int vRand; // vertical random gap between rows
int stroke;

for (int row = 0; row < 18; row++) {
 vRand = int(random(30));
 for (int col = 0; col < 18; col++) {

// decide if the shape is filled or not
 if (int(random(256)) == 255) {
 noFill();
 stroke(c); // Color
 stroke = 5;
 strokeWeight(stroke);
 } else {
 noStroke();
 fill(c); // Color
 stroke = 0;
 }

r = int(random(-20, 20));
 // draw the square
 rect(w * col * 2 + w / 2 + r, h * row * 1.35 + h / 2 + vRand, 
 (w * 3 / 4)-stroke, (h * 3 / 4)-stroke );
 }
}

// generate a unique filename for each generation based on datestamp
save("Grid "+nf(month(), 2)+"-"+nf(day(), 2)+"-"+year()+" at "+nf(hour(), 2)+"."+nf(minute(), 2)+"."+nf(second(), 2)+"s"+".png");

Adjustments to Timeline:

I am feeling motivated to make this work a triad and plan to design one more algorithm for a third piece. I plan to get that done this weekend and if I am satisfied with it, I will have all three works fabricated.