What's the deal with electronic swatches?

[jsrnephdoc]

I've been involved with or read several threads here recently in which the need to compare "in the cloth" swatches is emphasized, not just to confirm sett size or the feel of the cloth, but even just to know what actual COLORS it contains. I've never worked in IT or fashion design, but I'm pretty sure camera and computer display technology have reached a point where, assisted perhaps by blessings from Pantone, where everyday technology should make it clear what one will unwrap once the package arrives. Yet we read over and over that one cannot tell from computer images what actual colors will appear in that cloth.

Now, that would NOT eliminate all error. For example, a kilt I ordered a few years ago was made with fabric the kilt maker ordered correctly from the mill, but the mill sent a different tartan, after which the kilt maker used it to make me a kilt and sent it off without reviewing what I had ordered (and what they had requested, correctly, from the mill), and I was horrified to unwrap something EXTRAORDINARILY different from what I expected.

(To it's laudable credit, the kilt maker donated that kilt to me, and while it emerges from my closet less than each of my alternatives, I've come to like it—a little bit).

Perhaps the answer to my question is simply quantitative, not qualitative. There are just TOO MANY tartans to create faithful electronic images of them all, especially to post on databases that have evolved over decades and even over centuries. And, of course, that assumes also that the typical purchaser isn't using a 1960s era RCA CRT TV to make his or her purchasing decision.

But that brings up another possibility. Assuming that a potential cloth purchaser is willing to have it made using readily available yarns already dyed, hoping to obtain cloth of a relatively uncommon tartan or one one not woven for a long time, would it not be possible for SOMEONE to use a small loom to weave swatch-sized samples before the purchaser agrees to have 30 meters of cloth woven?

I ask this based on curiosity far more than hopes I could motivate an industry, although it's certainly been said before with laudable intent, to wit: "Some men see things as they are, and say why. I dream of things that never were, and say why not."

[Father Bill]

Every computer monitor phone screen and television prefers and emphasizes certain colours. My TV loves reds and makes them blindingly bright, also effecting any colour with a smidgen of red in it. It's so obvious that I don't trust its accuracy, and I normally see less red than others being red-green deficient. My family and guests see the same aberration I do, so it's not my colour deficiency.

Interesting, so does film as in old fashioned cameras. Agfa for instance was famous for emphasizing greens, and was therefore often preferred by nature photographers.

[BobF]

I have two monitors side-by-side that show colors noticeably differently. And I can’t tell you how many eBay auctions I’ve looked at where the seller takes several photos of an item and it’s difficult to discern what the color actually is, they vary so much from shot to shot. Add to that the difference you’ll get depending on whether the light source is natural or incandescent or fluorescent or LED. I’m always happily surprised when an item arrives and it looks better in real life than it did online. Often in real life the blacks are blacker, as the screen on, say, my iPad for instance is never truly black until it’s turned off. When I design posters for the local opera company, I’m always a little apprehensive about how it will look in print, partly because of the calibration of my monitor, and partly because with the advent of digital printing, I don’t believe that printers open a can of a specific Pantone color ink anymore.

[jsrnephdoc]

Thanks both to Father Bill and BobF for their helpful posts, but I still suspect that "every" should be replaced by "many" in their responses.

Today, many people's incomes depend on accurate color reproduction (I'm thinking of those who work in the magazine and fashion industries), and I suspect that, guided by Pantone color samples, they can do an amazingly good job of making sure what starts out on a screen won't surprise anyone when it appears in print in a magazine or in a piece of cloth. That may not come cheap. Apple's separate monitors BEGIN at $1600 but offer many calibration techniques available to those patient enough or savvy enough to make use of them (and I confess I'm not among those possessed either of the skills or budgets to qualify). But I DON'T believe that aspiring to that need be answered by "it's not possible."

But, I don't know anyone who works for Architectural Digest or a Paris design house, so I could be spouting nonsense. I'm pretty sure I'm not, nevertheless.

And, of course, for the tartan seeker, I'm asking two questions. The other relates to the notion of obtaining actual cloth samples woven on looms capable of making swatch-sizes samples. Perhaps those don't exist except as hand-looms of the type Peter MacDonald is expert at using. I haven't the foggiest notion regarding that.

[geomick]

problems with images -

Monitors -
many types of monitors with different colo(u)r response do to construction variability. They could be corrected through a profile tool, but very few general users ever do that. Photographers (hobbyist or professional) will calibrate their monitors to be correct when displaying their choice of colo(u)r space, but won't be correct for any of the others (sRGB, AdobeRGB, etc). Many photographers will even calibrate their printers to actually print correctly, but most people don't do it.

Light source for the photo -
If you use daylight, you get a continuous spectrum for your illumination, but the colo(u)r temperature will change depending on time of day, cloud cover, etc. There are calibration devices you can either photograph prior to the intended subject or have in the corner of the frame that let programs like Adobe Photoshop correct the image for the true colo(u)r temperature of the light. If they use one.

If the light is not daylight, Xenon flash or old style incandescent bulbs, then the spectrum is not continuous and there will be spikes of brightness for certain sections of the spectrum that are overly bright and other regions that are overly dim. This means the image will be biased by the light source. This is especially common for LED or fluorescent lights. You can attempt to optimize colo(u)r temperature, but it will never be quite right for non-continuous light sources.

The camera (film or digital) -
As was mentioned above about Agfa being overly sensitive to green, film is based on using organic dyes to sensitize the silver halide particles to a certain wavelength region of light. Many old films were especially bad with red. If the balance of all the dyes are correct, then the film will record an image which displays correct, provided it is balanced for the right type of light (daylight film vs. Tungsten lamp film). Digital sensors have a similar issue. They use a sensitizer layer to render 1/4 of the pixels sensitive to red, 1/4 sensitive to blue and 1/2 sensitive to green. The colo(u)r of an individual pixel is taken from the intensity of the 4 pixels (2 green, 1 red, 1 blue) used to calculate it. The higher end the camera, the more likely the pixels will be well balanced, but it is possible to have dead pixels on a sensor and the camera needs to know that when it calculates the image. The RGB model is based on the sensors in the human eye. When the program parameters correct for dead pixels and for the colo(u)r temperature of the light, you get an accurate image. But most people leave their camera on Automatic White Balance, so it will adjust the setting based on how the camera perceives the image.



tl/dr - lots of factors which can all be corrected for, but it will only be completely right if all are done correctly and very few do

[jsrnephdoc]

tl/dr - lots of factors which can all be corrected for, but it will only be completely right if all are done correctly and very few do

Thanks for expounding on that. I still think my question is pertinent, because there are plenty of examples where people's incomes depend on making EXACTLY such effort. Yes, that doesn't help the customer who doesn't make similar efforts, but if the mills at least did so to create their "jpeg swatches," for at least SOME customers surprises would become less frequent.

I suspect my alternative proposal (one-off weavings of non-inventory tartan swatches on small looms using in-inventory yarns) might be an alternative to a 10 or 30 meter weaving that turns out regretfully for the customer, but I don't know whether such infrastructure even exists.

[figheadair]

Thanks for expounding on that. I still think my question is pertinent, because there are plenty of examples where people's incomes depend on making EXACTLY such effort. Yes, that doesn't help the customer who doesn't make similar efforts, but if the mills at least did so to create their "jpeg swatches," for at least SOME customers surprises would become less frequent.

There are a number of of factors to consider when viewing jpg, or any other graphic format, image:

1. The graphic production will depend on the software programme use to produce it.

2. As far as I know, no graphics' package is calibrated to Pantone colours. Instead they use RGB or CMYK codes that match a range of shades of each colour. Some graphic packages, such as the one I use, allow the user to adjust/add in the shades.

3. The, as others have said, there is the difference based on the viewing hardware. Take any graphic, look ate it on your desktop, laptop, tablet and phone, and you will see that each will show the images shades slightly differently based on the graphics card or colour generator used.

I suspect my alternative proposal (one-off weavings of non-inventory tartan swatches on small looms using in-inventory yarns) might be an alternative to a 10 or 30 meter weaving that turns out regretfully for the customer, but I don't know whether such infrastructure even exists.

Whilst this is a nice idea, it is wholly impractical. An 18" x 12" sample would have to be hand-woven. That would take me the best part of a day's work with the associated cost.

[OC Richard]

When the vagaries of photography and monitors and screens has really jumped out to my eye is with tweeds.

I got some swatches to hand and then compared them to the images seen on various sites.

Maybe one site's images looked very close, but perhaps on somebody else's laptop they wouldn't.

A tweed suit or jacket being worn in a film or television show might appear different in every shot.

There's even a single take, a shot from Downton Abbey, where one character, dressed in a tweed shooting outfit, walks across a room and as he does his suit changes from grey to brown.

So we can't rely on any image, be it photo or CGI, be it on television or on a computer or on our phone or printed.

Nor even in person! Tartans and tweeds can look very different in indoor light. You have to take the garment or swatch outside the shop to know what it looks like.

[jsrnephdoc]

There are a number of of factors to consider when viewing jpg, or any other graphic format, image

and

Whilst this is a nice idea, it is wholly impractical. An 18" x 12" sample would have to be hand-woven. That would take me the best part of a day's work with the associated cost.

Well, perhaps not COMPLETELY impractical. We come at this question from polar opposites of experience and ability, but I would not anticipate such a task requiring YOUR skills, knowledge, or care. Such an undertaking would not be needed often, and would not even need to be done with great skill. The major question would be color. In my own case—the Robertson Red (white line)–even that might not be a major issue, because it's easy enough to see and squeeze commercial samples of the "foundational" red (modern) tartan on which it would be based at any kilt shop or via easily available swatch samples (that perhaps to choose between 13 and 16 oz fabrics, or mills, for example). The mill could dispense (at customer cost) small samples of the same yarns they use for the common Robertson (Red), accompanied by their most commonly used white yarn of the same weight, to a chosen hobbyist with a home loom. Now, I have NO idea what the hobbyist would earn nor what commission the mill would receive for recruiting and using the hobbyist, but I suspect it would be orders of magnitude less than the cost of 30 (or even 10) meters of single width cloth. My vision likely would be that the mill might be HoE, given their proven ability to do well with a herringbone selvedge (which would NOT need to be woven in to the sample). The customer would then have at least a good idea of what the COLOR arrangements would look like in the actual cloth and not be disappointed by that and the cost of the entire custom weave.

I know only one hobbyist with a loom. She was my first wife, and while we parted amicably 48 years ago, we've not even traded greeting cards for a few decades. But, I might be able to ask her what she'd charge a sponsoring commercial enterprise for such an assignment coming from a stranger, and how long it might take her once she had the materials in hand.

Certainly it would be less expensive than requiring the mill and end-user to share equivalent computer skills and hardware to examine on-screen images. On the other hand, while I HAVE seen a YouTube video from Lochcarron that illustrates their skill in creating a new tartan and offering several computer print-out options of their design to a customer, my guess would be that the mill could provide a standard swatch of the in-inventory tartan at VERY low cost (accompanying the hand-woven swatch) and make it clear to the customer that the hand woven swatch is not meant to convey ANY information about the quality of the weave, but rather just the arrangement of the colors. Of course, not every such request would be as simple as mine, but I still think the idea has merit and even possibility.

Thanks again. By the way, this idea came to me a few days back, as I was re-watching your trans-oceanic interview hosted by Rocky from USA kilts a year ago, discussing the HoE "Genesis" collection. I was VERY fond of that video. And, repetitively, I'm amazed at your generosity and graciousness in answering questions from "newbies" such as I. Truly there are angels in this arena, and I'm lucky enough to to be able to communicate with two of them (the other being Barb Tewksbury, who entertained and quickly responded to my emailed taunt about the Integrity astronauts discovering that there might actually BE brown on the lunar surface.

[BobF]

.
There's even a single take, a shot from Downton Abbey, where one character, dressed in a tweed shooting outfit, walks across a room and as he does his suit changes from grey to brown.

I have a pair of Haggar corduroy pants that, when I bought them under the store lights, looked a lovely shade of forest green. Took them home and they looked brown. I never know what color they will display in any particular room. I’ve never had an item of clothing that was as absurdly dependent on the light source to determine what color it appears.