[Above: a small quilt that I made a few years ago. It was good to come back home and see this item hanging on the wall. I'd forgotten all about it while I was away caring for my Aunt Claire (who lived almost to age 90, missing her birthday by just 27 days - see previous post) . The curving designs on the pale patches of fabric are derived from the rims of neolithic Korean pottery; I'd screenprinted them using paper stencils held together by ordinary sewing thread. To me the quilt feels somber but also full of energy - a good welcome back after a difficult couple of months.]

Not surprisingly, I’m looking now with fresh eyes at quite a lot of things, including my own goals and how I’ve been spending my time. It does seem to me that maintaining the Etsy fabric listings here has soaked up an extravagant amount of time and labor (my own). I didn’t mind in the beginning, when that labor had a large creative component. It was fun deciding how to color-sort the fabrics and writing the new computer code to do it. But now that the job has settled down into a clerical routine…

Here’s the problem. I can pre-select a bunch of locations on Etsy (i.e., sections of Etsy shops that sell fabric) and write programs to analyze and group in various ways all the fabrics for sale in these locations. But I don’t control the images in Etsy shops; and I have no way, without getting deep into the some kind of horribly computation-intensive computer-vision software, to filter out automatically those images that don’t work well as part of a group. (There are about ten types of such images; read about them here.)

To produce attractive pages of fabric, I’ve been doing the filtering manually. In other words, each update of the morecloth site has required someone (me!) to review thousands of images, arranged on hundreds of pages, clicking on all those I can’t use. It’s depressing doing this – saying “no, no, no, no no” – and also a very tedious and eye-straining task, one that takes an entire workday or more. I want to be using my talents to celebrate cloth and design. It pains me to spend whole days squinting at pictures I have to reject (and often for seemingly really petty reasons).

And here’s yet another problem. Even if I could persuade a group of Etsy fabric sellers to organize their shops in a way optimized for my purposes (i.e., with some designated sections that never include images of the kind I can’t use), I’d still have to deal with the issue of duplicate images. Because most fabrics (even niche fabrics like those created for the quilting and home-sewing market) are mass-produced, there’s bound to be a lot of overlap in the offerings of Etsy fabric stores. For buyers and sellers this may be a good thing. Buyers can choose where their money goes and shop at the stores they most want to support. Sellers can compete on the basis of customer service, shop aesthetics, and social networking, rather than on inventory alone. But for someone with my particular interests, duplicate fabric images are a time-consuming nightmare!

Aesthetics and data-mining possibilities have been, for me, primary factors driving this morecloth experiment. I want to create compelling online displays of fabric (which, to me, means, among other things, web pages without duplicate images). I’d also like (through statistical analysis of fabric listings and images) to uncover trends in fabric color and design. This information may not be timely enough have any commercial value, but it’s altogether fascinating from a social and cultural point of view. Because, of course, printed fabric, like all other art forms, both reflects and shapes the culture in which it lives. I want this information, but I don’t see how to get at it without some automated computer-driven way to identify listings and images that are essentially duplicates.

Any solutions? I’m momentarily at a loss – not sure how to revamp this morecloth project so as to reduce the amount of drudgery and recover my original motivation. I’m open to suggestions, willing to swap ideas and collaborate in various ways. I’d also be happy (if there’s any interest) to distribute (under an open-source license) all the working computer code for this site. Thoughts, suggestions, and wild ideas most welcome….

[Above: a pair of colorful necklaces that my Aunt Claire found in a thrift store.]

For the past two months I’ve been away from this blog, away from my home, and only intermittently able to access the Internet, I was called away, abruptly, to care for my Aunt Claire during the last weeks of her life.

While I’ve been able, minimally, to keep the morecloth site running and updated, I’m still not home yet and only just beginning to catch up on email. (If you’ve written to me or commented here and not yet got a reply, I’ll be answering soon.)

[Above: fabrics from shirts that Claire sometimes wore]

My Aunt Claire loved color, fabric, and decoration – and, not surprisingly, also loved to sew. She left behind, among other things, a well-oiled machine and a closet full of bright and pretty clothes.

[Above: the view from my Aunt Claire's kitchen window.]

In this now deleted post, I talked about wanting to write, if not balanced reviews, at least appreciations of some books on textiles that have mattered to me. (I’d even gone so far as to sign up with Powell’s and Amazon’s book-selling “associates” programs so as to get permission to use their bookcover photos.) And then I changed my mind, lost my enthusiasm for the project.

Partly because the permissions extended to cover photos only, leaving me only words to convey a book’s inner content. (A real handicap in cases where the book’s strength lies as much in its images as in its text!) I also felt uneasy about the conflict of interest inherent in praising books on which I stood to earn a commission, however small. So…no book reviews, at least for now.

[Above: rutabaga slices carved for printmaking purposes and a digital textile design using motifs created by printing with the rutabaga slices on paper.]

Back in 2008 Spoonflower altered forever the landscape of small-quantity on-demand fabric printiing. By offering digital printing at prices affordable to artists, hobbyists, and small crafts businesses, Spoonflower changed not only what can be done (and done by whom) with cloth, but also what’s likely to be done, both for pleasure and for profit. Commercial inkjet printing for fabric has been on the horizon for decades – see Danielle Locastro’s essay at first2print.com for an analysis oriented toward the garment industry – but only recently available at artist-friendly prices. In the past designers who wanted to see a repeat pattern printed on fabric had to: (1) sell or license the pattern to a business, (2) pay $100/yard or more to a commercial digital printer that bundled digital editing and/or color-matching services in with the printing, or (3) resort to low-tech printing devices. And by “low tech” I mean everything from finicky consumer-priced inkjet printers to silkscreen to linocut blocks to Javanese wax-printing tjaps to rubber stamps to paper stencils to the lowly potato (or rutabaga). Spoonflower’s affordable printing service changes everything; it’s what’s known, in business terminology, as a “disruptive technology.”

[Above: two more digital rutabaga prints, rendered in randomly chosen colors from Spoonflower's downloadable color swatch

Every printing process has its special aesthetic appeal, and fabric designers will no doubt continue to print and experiment with all the traditional low-tech methods. And, of course, they’ll continue to produce repeat patterns with the mass appeal needed for profitable mass production. But the services of Spoonflower (and its emerging competitors in the affordable custom-printing marketplace) will surely help shape the future of printed fabric design. Patterns bubbling up from hobbyists and artists/craftspeople will mingle in unforeseeable ways with those created strictly for mass production. That’s the nature of disruptive technology: game-changing, fascinating, and potentially disruptive. As far as “disruptive” goes, I guess I’ll just say that I’m glad I haven’t invested all my hopes in a business making old-fashioned rutabaga prints! The Spoonflower-enabled competition might be fierce.

Well, that’s an example (albeit a contrived and ridiculous one) of how affordable high-tech fabric printing could undercut a hypothetical low-tech printing business. What about Spoonflower’s new design marketplace? What effect will it have on person-to-person sales of digital repeat patterns outside the Spoonflower marketplace? Up till now these patterns have mostly been created and sold for non-textile purposes – as backgrounds for web pages, textures for 3D-computer-modeling, etc. But there are some at designs at patternhead.com and at istockphoto.com (find them at istockphoto by searching on “seamless pattern”) entirely suitable for printing on fabric. Will people go to the trouble of searching out these designs, paying for them, and uploading them to Spoonflower, if they can get similar designs right at the Spoonflower site with no trouble and no extra payment at all? The Spoonflower design marketplace benefits designers by providing them with a free place to show their work to a highly interested audience, by paying generous royalties on fabric actually printed, and by allowing them to retain copyright to their work. And, at the same time Spoonflower poses a serious challenge to designers hoping to set up shop independently in the just-barely-emerging person-to-person textile design market. It will be interesting to see how this challenge plays out.

I will be writing more on this. And also on Spoonflower’s textile design software that, I read somewhere (can’t recall where, just now), may be in the works.

Fading and blending with restricted color

The three fabrics above, all quilting-weight cotton from the Northcott “Visual Arts” series, appear to be printed in the normal commercial way (i.e. using rotary screens, a small palette of colors, and, maybe, halftones for the areas where one color fades to another). These color-blending areas – light yellow to dark in the top two images, light green to dark in the bottom one – give the cloth a vaguely antique or hand-colored look. I love these subtle blends, however mechanically and artificially achieved. But getting the fading colors to read as fresh and contemporary (rather than simply as nostalgic and fake) has called for some clever design decisions.

Each of the yellow-background fabrics is overprinted (in appearance, if not in reality) with a highly complex but single-color (black) pattern. This pattern catches and holds the eye, diverting attention from the mechanical means by which the background effect is achieved. And the crisp, unfaded black lines announce up front that this is a piece of contemporary fabric, not imitation vintage. In the green fabric, a grayish pattern, derived, maybe, from a high-contrast photo of a woven texture, is printed over a solid yellow-green ground. This pattern looks modern/postmodern (frankly photo-derived) and it also draws attention away from the mechanics of the green fades going on inside the round medallions.

No fades, no overlaps, no oblique angles

The next three prints , which by coincidence (or maybe not) all happen to be animal prints, use simple shapes, no blending and little detail inside the shapes, and no overlapping at all. In the middle fabric – the only one where some of the shapes are hollow – a few of these hollow shapes do touch each other or the big black target-like form. But none of these collisions lead to deformation. Nothing gets squashed; nothing overlaps. It’s as if all the animals, plants, and rocks (or whatever) are on parade and all taking care to stay a proper distance apart and turn their best sides toward us.

The charm of each design is that they do this so well, filling up the blank white space so fully and evenly. That and the inherent appeal of animals, however symbolized and simplified for easy recognition. There’s also the feat of fitting all these shapes together while keeping all the animals in heads-up position (or, in the case of the first fabric, at right angles to each other). The designers have done a lot with simple means.

At some point back in 2000-2001 I made the decision, for reasons both practical and philosophical, to stop using Microsoft’s Windows, along with a lot of other proprietary software. Everything I do these days – the morecloth site, my personal computing, the soon-to-be-launched site I’m building for online textile design – everything runs on free open-source software. This policy has worked well for me. The only only glitch I’ve encountered so far has come from the fact that fabric-printing services, all of which seem to depend on mysterious black-box configurations of unnamed software and hardware, tend to release clues about their secret processes only in proprietary Adobe Photoshop formats.

Of the four low-cost custom fabric printers I’ve looked at (Spoonflower, FabricOnDemand, KarmaKraft, and EyeCandey aka candeyshop.com), Spoonflower has been the most forthcoming with information. Last year Spoonflower published a downloadable file (in Photoshop .aco format) containing 523 colors (well, actually, 505 colors, since some turn out to be duplicates, at least when expressed in 24-bit truecolor RGB format) – let’s just say a downloadable file containing a big general-purpose selection of Spoonflower-safe colors. By “Spoonflower-safe (my term, not theirs) I mean that, if I understand things correctly, none of these 523 (or 505 or whatever) colors is “out of gamut” for Spoonflower’s printers. In other words, all of these colors can be printed, singly or in combination with each other without triggering the dreaded “remap colors” function that seems to reside deep in the bowels of all commercial printing systems.

I say “dreaded” because this “remap colors” function is undisclosed (at least to us end-user customers) and therefore unpredictable. The function doesn’t just remap the out-of-gamut colors to in-gamut colors; it remaps some or all of the in-gamut colors as well. In short, it tries to help us clueless customers out – and, no doubt, in many cases succeeds – while telling us, in effect, not to bother our pretty little heads about all that in-gamut / out-of-gamut stuff. Not my preferred way of doing business. In fact, as I’ve been getting ready to offer design-and-layout services to people wanting custom-printed fabric, my biggest worry has been color predictability – i.e., preparing files in such a way as to dodge the unwanted help of unknown color remapping functions.

Yesterday I finally set to work on Spoonflower’s Photoshop-formatted file of safe colors. (I hadn’t realized earlier that it could, in fact, be auto-translated into a format usable by gimp-2.4, the free open-source program I use in place of Photoshop.) Once I had the Spoonflower palette file in a format I could use, I still had to make some adjustments; I had to split the palette up into smaller ones, since indexed-color images require palettes of 256 or fewer colors. But now, with that detail taken care of, I have the means, so it seems, to auto-generate a large number of color schemes in Spoonflower-safe colors. (You can see a sampling of 252 such color schemes in the two images above.) I’ll be printing swatches to test things out and, if all goes well, uploading some color scheme files here that others may find useful.

(To see the Spoonflower-safe palette and a couple of smaller palettes that I extracted from it click on the following link.)

Anyway, I’d been getting impatient with my own piecemeal efforts to put together color combinations of fabrics at morecloth. It seemed to take hours just to arrive at twenty or so semi-attractive combinations. And then it occurred to me that I was, in effect, using whole pages of fabrics at morecloth in much the same way that color-scheme software uses the single colors in a palette. So a wrote a little color-scheme generator to work with morecloth’s pages of color-sorted fabrics.

I asked it to auto-generate some color combinations, and, in a matter of seconds, it came up with fifty-eight of them – most of which looked better to me than the ones I’d chosen by hand! (Is this progress or a reflection on my color-picking abilities? It’s sort of embarrassing to be beaten by a machine, even if you’re the one who programmed the machine.) You can see the fifty-eight combinations here (or just click on the “color combos” link at the top of most morecloth pages). I’m working on additional color-combining rules and plan to have more and more varied combinations in the next site update.

(1) Come up with come up anything definitive – any model way of sorting or presenting fabric. There’s plenty of room for other approaches and other websites built around the wealth of fabric at Etsy and elsewhere.

(2) Turn morecloth into a for-profit undertaking. This is a side project for me, something I started impulsively when I learned, a few months ago, that Etsy had created an API (application programming interface), i.e. an easy way for programmers to get at the data on Etsy’s site.

(3) Create a mysterious “black box” situation (see picture above). The software behind this site is pretty simple. I’m not publishing the source code – at least not yet – because it’s in flux and a bit of a mess right now. But read on for an explanation of how it works.

The morecloth site is static: just plain xhtml pages that get regenerated and replaced every week or two. I host them on a low-cost shared server and generate them offline on my home computer, a desktop machine running GNU/Linux (Debian Lenny) that I got for $98.75 on ebay, shipping included. I do know about Moore’s law and falling technology prices, but it still kind of amazes me that this project is so affordable, that the economic barriers are so low.

To create initial versions of the morecloth pages I use a bunch of Python scripts, some of which invoke the ImageMagick image processing library. I then review these pages and edit them into final versions using more Python and ImageMagick and a touch of JavaScript. Here’s the procedure:

(1) Download fabric listings using the Etsy API. (How to query Etsy is explained at developer.etsy.com.)

(2) Extract Etsy listing IDs and image names from the downloaded data using regular expressions.

(3) Fetch images from Etsy and gather color data from them. I don’t save the downloaded images; I just use a Python script to grab them off the Etsy site and process them on the fly with ImageMagick. Invoking Imagemagick from the commandline (possible from within Python, thanks to Python’s subprocess module), I quantize the images (i.e. reduce colors in them to a manageable number), create histograms from the reduced-color images, and store the histogram data with the corresponding Etsy listing ID and image name. This may be a slow, clunky way of doing things – there are more direct ways to work with ImageMagic using Perl or C – but speed isn’t a big consideration, since all the image processing takes place offline.

(4) Group listings according to color. There’s no end of ways to do this. But first it’s important to reduce the number of duplicate fabrics. Viewers probably don’t want to see a page with a bunch of identical or nearly identical pictures. Fortunately, most duplicates can be eliminated without visual inspection, simply by having a script compare the histogram data.

Once the listings are thinned down to eliminate most duplicates, functions can be applied that map the listings to numerical values. One such function, for example, looks at the average color of a fabric’s image and returns a number representing the lightness or darkness of that average color. Given a population of fabric listings and numerical a function defined on those listings, the population can be divvied up into groups by various statistical procedures – many available through the “stats” module for Python. I’ve barely had a chance to play with the possibilities, but it’s clear that there are many ways to represent a population as a collection of non-overlapping sets and that, given several such representations, sets from one can be intersected with sets from another to capture a variety of color-related properties of interest in a population of fabrics.

(5) Clean up the color groups. Once a group is assembled, it may still have some problems. I still have to lay eyes on the actual images and discard anything that detracts from the group as a whole. I try to catch images that include highly visible watermarks or non-fabric objects (e.g., tape measures or coins used to indicate scale). I also try to spot any duplicate fabrics still hanging around or anything that just looks out of place – any fabric disconcertingly different from its neighbors. Some of these visual judgments may be susceptible to automation, but probably not all.

To do this clean-up operation I use Python to create a temporary web page (for local viewing only) for each proposed color group. This page includes some JavaScript and a form with a text box: clicking on a fabric image calls a JavaScript function that removes the image from the page and adds the corresponding Etsy listing ID to the text box. Later, when the form is submitted, a Python cgi script writes to disk IDs of all the listings deemed inappropriate for the group. After all the color groups have been inspected, another Python script creates the final pages that appear on morecloth, along with the bands of colors seen on the index page.

There are a few refinements and add-ons to this process: Python scripts for blending, splitting, and combining color groups. But that’s the basic idea behind morecloth.

Oops! I forgot to say why I’ve taken on this project. Well, the fabrics are pretty. The statistical stuff is potentially interesting. I feel generally friendly toward Etsy sellers. And – who knows? – maybe this will generate some interest in a separate for-profit textile project I’m launching later this year. (More on that later…)

Maybe I should add a note of thanks to the people who provide, one way or another, the images that appear here – to the Etsy sellers and to Liza Prior Lucy of Glorious Color, who gave me permission to use images from her site. I don’t work for any of these people – morecloth is a strictly non-commercial undertaking – but my fabric-sorting efforts wouldn’t get anywhere without a good supply of fabrics to sort. In the next post I’ll be writing a bit about my own motives and about the software behind my site.

Images in this article show items on sale at Etsy. Click on them to see the Etsy listings.