Egg Quilt: Color

It’s been about three dozen months since I moved to California and joined the Flying Geese Quilters Guild, where I met Bridget Paley, who gives me dozens of eggs, whose colors I matched to the circles here by carrying crushed eggshells captured in clear packing tape through local fabric stores.

So this awesome friend of mine from the Flying Geese Quilters Guild has chickens. When her chickens lay more eggs than she can use, she gives them away. In March she gave me and two other members of the guild each a dozen eggs. We three recipients conspired to make quilts based on our eggs. Our goal was to finish them by the next guild meeting. So how do I get from a dozen eggs to a quilt?
The most unique thing about the eggs were their colors. I lined up my eggs by hue and value.

The top left egg falls squarely in the orange hue. It has a high value and a low chroma. Working with GIMP, I found this egg to have a hue of 41 which is really pretty much dead center in the orange spectrum, a saturation (vaguely translatable to chroma) of 62 on a scale where 0 is neutral gray and 100 is as full of color as color can be, and a value of 64 on a scale where 0 is black and 100 is as light as orange can be and still be orange of this hue and saturation.

The second egg from the left is ever so slightly more yellow than the first egg. Again, it has a high value and a low chroma. In HSV language that's 43, 66, 67 - so more yellow, higher saturation (less gray), and higher value (lighter).

The third egg from the left in the top row is more yellow than the second, with a chroma between the first and the second, and a value lower than either of the first two. In HSV: 49, 64, 59. This egg is an excellent example of how yellow at a low value is often misleadingly named "army green."

The egg in the second row of the third column is actually pretty spot on in terms of its placement in the correct hue column. According to GIMP, the hues of this egg and the egg above it are exactly the same. How awesome am I? OK, it's a weird talent to be awesome at, but still. This is even more impressive given the drastically lower chroma and significantly higher value. In HSV: 49, 42, 71

The egg in the third row of the third column is also pretty amazingly accurately located. It's ever so slightly more yellow than the two eggs above it. Again, it has much lower saturation than the eggs above and a higher value. In HSV: 50, 32, 77.

The egg in the fourth row of the third column is exactly the same hue as the egg above it (so ever so slightly more yellow than the top two eggs in this column). It is crazy low chroma and high value. In HSV: 50, 19, 87.

The fourth egg from the left is a bit more yellow than third column of eggs. We're definitely at the yellow end of the orange spectrum. It is very close in saturation and value to the third egg from the left in the top row. In HSV: 53, 64, 60.

The fifth egg from the left is slightly more yellow than the egg to its left. It has a slightly lower value than the egg to its left. The big difference is saturation - this egg is more gray than its neighbor to the left. In HSV: 55, 56, 59.

The egg in the second row of the fifth column is exactly the same hue as the egg above it. Dude, I rock hues! Like the trend in the column of eggs to the left, this egg is lower in saturation and higher in value than the egg above it, though that difference isn't as drastic as it was in the second row of the other column. In HSV: 55, 51, 62.

The egg in the third row of the fifth column is exactly the same hue as the other eggs in this column. It continues the lower saturation, higher value trend. It's pretty close to the saturation and value of the second egg in the other column. In HSV: 55, 42, 67.

The egg in the top row of the sixth column (also known as the top egg on the right) is, apparently, my downfall. It should be in the fifth column. It's the same hue as all the eggs in the fifth column. More specifically, it should be between the top egg in the fifth column and the second egg in the fifth column. In HSV: 55, 53, 60.

The egg in the second row of the sixth column is the yellowest of all the eggs here. It's not smack dab in the middle of the yellow spectrum, but it's a good deal closer to the center of the yellow spectrum than it is to the center of the orange spectrum. It's saturation and value are closest to the egg in the second row of the fifth column (i.e., the egg to the left of it - so that's a good arrangement). It's value In HSV: 58, 47, 61.

Now, knowing my proclivity for The New Munsell(R) Student Color Set, it comes as no surprise that I tried to match the eggs to swatches from the color charts. What did come as a surprise is the total impossibility of doing so. The 5YR chart is WAY too red to approximate even the farthest left egg.

By way of comparison (and an interesting demonstration of how GIMP's HSV is NOT equivalent to Munsell's HVC not controlling for photographic distortion), the closest HSV hue number to the most orange egg is 28 (which was only the hue number for the two farthest right chips in the second column from the top - if HSV = Munsell, all the hue numbers should be the same). This is twelve steps more red than the HSV hue number of the least yellow, most orange egg (41).

I had better luck with the Hue 5Y chart. The chip third from the left in the second row had the highest HSV hue number. It had an HSV of 52, 61, 73, which is closest to the fourth egg from the left (53, 64, 60). All of the eggs to the right of that egg were too green for any of the chips on the Hue 5Y chart.

The eggs in the third column had the same HSV hue number as the second chip from the right in the top row, but they don't look very similar because the saturation and value of the chip are 74 and 82. The range of HSV saturation numbers of that column of eggs was 64 to 19 from top to bottom row. The range of the HSV value numbers was 59 to 87 from top to bottom.

The lowest HSV hue number I could find on the chips was 44 on the farthest right chip on the third row from the bottom. The second egg from the left had an HSV hue number of 43. However, like the comparison in the previous paragraph, the egg color doesn't resemble the chip because the saturation and value of the chip are 73 and 44 whereas the saturation and value of the egg are 66 and 67. In this case the saturation of the chip and the egg are closer. The top egg on the left is too orange for the 5Y chart and not nearly orange enough for the 5YR chart.

Not capable of leaving well enough alone, I had to test the two rightmost columns of eggs against the Hue 5GY chart. The lowest HSV hue number I could wangle was the chip in second column of the top row: 72, way more green than the eggs in the fifth and six columns, which all had HSV hue numbers of 55.

"Why go through all this?" I hear you cry.

We perceive colors differently based on context. If we really want to identify the colors of things, we have to do some work. While running a picture of these eggs through Big Huge Labs' Palette Generator will generate uniform swatches, it's not the end of the color identification process. A bunch of HTML color codes and color swatches with names like "sycamore" don't really get you anywhere. But analyzing the hue, value, and chroma can tell you a lot more.

For example, these eggs, which I thought were orange and green, are actually yellow. This is HUGE! Imagine if I selected a color scheme for a quilt based on my inaccurate assessment the the eggs were orange and green. I'd have to go with a triad of orange, green, and violet, or a tetrad of orange, green, red, and blue. The actual color of the eggs isn't even in there. I'd go shopping or dyeing and lay out my fabric next to my egg pictures and be radically disappointed. Or even worse, I make the whole quilt in that color scheme and then when I set it next to my egg pictures I realize it's just totally wrong. Those two scenarios sound a lot less artistically satisfying to me than spending some time up front to get to know the colors before I buy or dye a thing.

I thought they were all very high values, but actually they run from 59 to 87. Value is the workhorse of quilt design. Knowing the value range of my inspiration eggs can help me incorporate a broader range of values in my quilt than I might based on my inaccurate perception that the eggs were all high values.

I thought they were all very low chroma, but actually their HSV saturation numbers run from 19 to 66. Again, doing a little color work offers me a much wider range of saturation possibilities than my inaccurate initial assessment.

I could take this numerical analysis even further - using the distribution of hue, value, or saturation numbers to help me decide the amount of each fabric to include in my quilt. For example, only one egg has a saturation anywhere near as low as 19, 32. Then there's a cluster of three in the forties, another cluster of three in the fifties, and a cluster of four in the low sixties.

If the most interesting thing about these eggs is their color, then it's worth taking some time and exerting some effort to get to know their colors. Maybe knowing the top left egg's HSV numbers are 41, 62, 64 isn't intrinsically useful. But spending time really getting inside those colors can help me make the best use of them.

Click here to see the rest of this post...

Katie Pasquini Masopust on Design

A couple of Mondays ago I took an amazing workshop with Katie Pasquini Masopust, a world renowned quilt artist who wins top awards from every quilt show that hands out ribbons. And if that wasn't enough, one of her quilts was named one of the 100 quilts of the 20th Century - there was this whole PBS documentary. And despite all that, she's laid back and cool. Her class totally changed my entire approach to design.

The workshop was put on by the Flying Geese Quilters Guild. Katie's name on their schedule of presenters (along with Joe the Quilter's) was one of the reasons I joined this particular guild. And not only were the guild members I hadn't met super nice at the class, people I had met previously actually remembered me . . . and seemingly not in a bad way. How nice is that?

In my review of Joe's class, I said,

The workshop took place at Material Possessions Quilt Shop, which is HUGE. I suspect it has as good if not better selection of quilting fabric as G Street Fabrics in Rockville, Maryland.

Apparently, I totally jinxed them, or they were hit by economic bad times like everyone else, because they have now reduced the size of their store by at least a third. Think global, sew local: support your local quilt shop! Nonetheless the classrooms remained spacious, well-lit, well-equipped, and well-organized.

On to the workshop itself.

In some ways, this workshop was like a super condensed version of the six-week class I took with Jeanne Benson at the Smithsonian. For example, rather than laboring over color for weeks on end, Katie passed out a one page handout with half the page dedicated to nine common color schemes and a color wheel. Additionally, she had prepared sets of solid color fabrics for fusing and separated the fabric into plastic sleeves by color scheme. So at the step in each exercise when we moved from line drawing to color, each student grabbed a set of fabric and tried to use all the fabrics in that set. No decision making, just grab a set and go. For each exercise we had to use a different set.

But it should be no surprise that our color work was elided. The name of the workshop was "Composition" - color was just a means to an end. Similarly, we didn't ponder the details of line, shape, value, movement, size, and pattern, like I did with Jeanne, but all of those subjects were touched upon along the way. So it makes sense that the other half of the handout was dedicated to nine common compositions. Katie defined "composition" as the way you arrange the shapes that make up your quilt and the way your eye travels around the surface of the completed quilt. Her primary goal was to give us tools to work through our quilt designs before we start the quilts so we don't finish our quilts and say, "Oh, now I see, I should have done . . ." Additionally, she wanted us to see the potential for abstraction from naturalistic inspiration. This will all make more sense if I walk you through the exercises.

Exercise 1
In the first exercise, we selected a photograph from a pile which Katie had ripped out of magazines and developed a design attempting to take a small step from "reality" towards abstraction.

First, we determined which of the nine most common compositions matched the composition of our photograph. I determined my photo's composition was symmetrical. Specifically, I think it's glide symmetry, but this isn't a Ruth B. McDowell workshop. Then we made a line drawing of the image emphasizing the composition. I drew every piece of the sculpture including the spheres. At that point we paused to share our drawings with the group. Katie took one look at mine and said, "Simplify!" After our constructive criticism session, we went back to our tables and redrew our composition, then copied it. One copy was for the base to which we would fuse fabric. The other copy was to make patterns to cut out the fabric. I'm not a big pattern fan, so I didn't make two copies. Then we picked our fabric set. I picked a set of violet and yellow fabrics in a complementary color scheme. There were seven values of each hue in the packet. I used every one.

In our critique session at the end of the day Katie noted that with "weird" color schemes, like complementary and split complementary, your can make the contrast less jarring by using the full range of values in each hue. I think that's certainly true in this example. If I only used the bright yellow that makes up the rectangles of the door with the dark violet of the second or third to bottom lozenge, it would be harder to look at. Along the bottom where there are just big chunks of each color you can see how contrasty it is.

Katie noted that the high value of the yellow rectangles in the door made her want to move inside and see what was there. It gave the image a sense of depth.

I had the opportunity to do this exercise twice. I picked the following photo and determined its composition was horizontal.

I picked a monochromatic color scheme in seven values of green. I used all of the fabrics.

During the critique session at the end of the day I wondered whether the green was too naturalistic of a choice. For example, Olle Baertling, a Swedish modern artist didn't like to use blue and green in his paintings because no matter what your intent the viewer would always see earth and sky or earth and water. Katie felt my pueblo design did not seem too naturalistic in part because she lives in New Mexico, the indigenous habitat of pueblos, and they sure ain't green. But also, looking at the design apart from the photo, Katie said it looks like an abstract collection of squares. Katie noted that monochromatic color schemes are cohesive and peaceful, and if you use the full spectrum of values you get an elegant result.

When I showed my designs to my husband when I got home he immediately asked if he could take this design and put it up on his office wall at school. He has NEVER had that kind of reaction to anything I've made quite possibly ever. So I might have to make this into a quilt for him.

Exercise 2
The second exercise required us to select another photograph and develop a design a few steps further toward abstraction.

In step 1 we made four small line drawings, one only using rectangles, one using only circles, one using only triangles, and one using rectangles, circles, and triangles.

In step 2 we picked our favorite from the four drawings in step 1 and determined to which of the nine most common compositions it conformed. I chose the combination of rectangles, circles, and triangles and determined that it most conformed to a diagonal composition. Then I drew my selected design again and tried to emphasize the diagonal composition.

In step 3 we redrew the drawing from step 2 with an eye towards improving it. I simplified my rectangles to three squares and drew them with a ruler. I laid my paper over the previous drawing to align my triangles along the curve without a continuous line.

In step 4 we redrew the drawing from step 3 with an eye towards simplifying it. I reduced the number of triangles, removed the skinny vertical rectangles, and reduced the number of circles.

At this point we were to pick our favorite design from steps 2, 3, and 4 and execute it in color. I picked the design from step 4 but redrew it to reduce the circles to one large circle (I traced the color wheel) and to divide the background into three triangles. I selected a split complementary color scheme with seven values of red, one value of blue-green, and one value of yellow-green. I used all of the fabrics.

Reworking the image basically nine times really changed my perception of it. I definitely felt like I worked out a lot of design issues by processing it so relentlessly.

During the end of the day critique I wondered whether it was derivative of Jane Sassaman. Katie reassured me that she knows Jane Sassaman and my design is no Jane Sassaman. But she didn't mean that in a Dan Quayle/JFK way. She really loved my design and suggested I should concentrate on it by working in a series. She thought I handled the weird split complementary colors very well. Yay me! This design in my favorite of those I generated at the workshop, right down to the crazy colors.

Exercise 3
For our final exercise, we selected one of the designs we made for exercises 1 and 2 and cropped to focus on one part of it, which would then be the subject of a new design. I picked the first design I executed for exercise 1 and cropped it like the photo below.

I made a line drawing at a larger scale of the cropped portion of my design. I selected an analogous color scheme of five values of yellow, one value of yellow-orange, and one value of orange. It's a little hard to see in the picture, but I divided the background with the orange in the middle and the yellow-orange on either side.

I didn't have much time to complete this exercise (lots of folks didn't get to this exercise at all), but the time pressure forced me to work faster which, I felt, generated simplicity and elegance. I do think I had this Josef Albers work in my head when I threw mine together.
Sorry for getting my head in the way. I took these late an night. But you get the idea.

In discussing the analogous color scheme Katie suggested using up to four adjacent hues. Analogous colors closely relate to one another and are really just a bare step from monochromatic, so try to use the full value range, particularly of the outside hues.

Katie had some general words of wisdom. She suggested using an unnatural or unexpected color scheme. In competitions it makes the judges take notice because it is different from all the rest, it's unpredictable, it makes people ask questions.

She warned that if you make something like this based on someone else's photo, you may run into copyright issues. So make your own photos the basis of your work. If you are inspired by the subject matter, like ferns, go take your own pictures of the subject matter.

If you're having trouble moving from realistic to abstraction try turning your inspiration photo upside down. Or when you've finished a drawing, turn it - you have four possible tops. Find a focal point, redo it in a different way, and inset it into the quilt.

A concept she referred to without directly instructing was transparency, which appears frequently in her work. The few pointers she shared were that one aspect of successful transparency is continuous line. The continuity of the line is more important than accuracy in color mixing.

Later that day Katie presented her Quilts and Quilt Trips lecture to the guild. She was the first guild lecturer I've seen who showed other quilter's work. It was like she is a quilt fan and wanted to share all of her favorites. It just felt incredibly generous. Here's a list of some of the quilters she mentioned:

• Judy Warren Blaydon
  
• Linda MacDonald
  
• Fran Skiles
  
• Nancy Halpern
  
• Therese May
  
• Michael James

Click here to see the rest of this post...

Flags by Colour

Via Dear Ada I found Flags by Colours by Shahee Ilyas. The page describes the work:

Using a list of countries generated by The World Factbook database, flags of countries fetched from Wikipedia (as of 26th May 2007) are analysed by a custom made python script to calculate the proportions of colours on each of them. That is then translated on to a piechart using another python script. The proportions of colours on all unique flags are used to finally generate a piechart of proportions of colours for all the flags combined.

Value: Markers

Since I'm not much of a painter, I've continued my value studies with the Munsell Student Color Set using my Prismacolor markers.
Prismacolor makes three different gradient grey color marker sets: french grey, cool grey, and warm grey. I have the french grey set. The cool greys clearly have a blue tint. Looking at the samples on my monitor I don't see as much of a specific color tint in the warm greys - maybe red, but if so it's very subtle. The french grey set clearly has a yellow tint - so at this low chroma it appears most like taupe. Though without any other color in context, the yellow is barely noticeable.

Like the value scale Deb Menz uses in her book, the Prismacolor grey scale is based on the percentage of black that is mixed with white. So the lightest marker is 10% and the darkest non-black marker is 90% and there are nine markers in the grey scale, plus black. As you might be able to tell, my 50% marker has seen better days. I should really buy a replacement.

One difference between opaque paint and the ink in Prismacolor markers is that you can build up layers of ink in such a way that the value and chroma change. These last two images are samples I made layering the 10% French grey and 20% French grey respectively. I found that by the fourth layer of the 10% French grey, the value matches the 20% French grey. So you can layer the 10% French grey to create two values between 10% and 20%.
I found that by the third layer of the 20% French grey, the value matches the 30% French grey. So you can layer the 20% French grey to create one value between 20% and 30%. Alternatively, if you want to make sure an area of grey is all one value, if you layer it at least four times it will be uniformly one value. This corrects the striped appearance I often get filling in blocks of color. I suppose I should do this test on every marker to see how dark it gets and how many layers I need to do to even out the color.
I also need to line up my other value indicies to see what Munsell values apply to which Prismacolor values and what Menz values apply to which Prismacolor values.

Click here to see the rest of this post...

Seeing Red

This value scale is from Art History by Marilyn Stokstad, which I'm reading with a couple of friends to make up for our failure to take art history in college. This value scale and some other basic tools to describe art are included in the Starter Kit section of the introductory material. For no particular reason that I can discern, values are divided into seven gradations including white and black at either end.


Stokstad's value variation in red uses the highest chroma colors from almost every value row on the 5R chart. For some reason it skips value 7/. From right to left: 5R 8/4, 5R 6/12, 5R 5/14, 5R 4/14, 5R 3/4.

Click here to see the rest of this post...

Discussions of Value: A Comparison of Resources

To further my remedial review of value after my first failed exercise from The Munsell Student Color Set, I reviewed the sections on value in all of the other books I have that address color: Interaction of Color by Josef Albers; Your First Quilt Book by Carol Doak; Color for the Terrified Quilter: Plain Talk, Simple Lessons, 11 Projects by Ionne McCauley and Sharon Pederson; and Deb Menz's Color Works: The Crafter's Guide to Color.
The above picture shows one value exercise in Interaction of Color by Josef Albers. Interaction of Color is like The Munsell Student Color Set in that it is intended as a companion to class exercises. Interaction of Color is probably an advanced seminar on color where The Munsell Student Color Set is an introductory course. If I could ever get through The Munsell Student Color Set, I would love to work my way through Interaction of Color.
The above picture shows the entire value discussion in Your First Quilt Book by Carol Doak. Considering the broad scope of this book, that's actually a pretty hefty word count for a discussion of value. The book dedicates all of eight pages to color-related issues. So for one of those pages to be solely dedicated to value is pretty good. The patterns in the book are also fairly easy to break down into at least light, medium, and dark values, which for beginning quilters is fairly sufficient. Most of the diagrams in the book are black and white, which probably doesn't appeal to buyers like full color graphics might, but the benefit of grayscale diagrams is that they give you a lot more sense of value placement than potentially confusing pictures using actual fabric. If you're looking for an introductory quilt book, I highly recommend Your First Quilt Book by Carol Doak. If you're looking for an in depth discussion of value, this is probably not the one. But it certainly introduces value in terms of fabric choices clearly. It also describes the universal quilters' value tool: The Ruby Beholder. Essentially an easily marketable substitute for a piece of red cellophane, a Ruby Beholder is a transparent piece of red plastic quilters look through to filter out the colors of fabric in order to determine color. Of course, if you love red fabric, the Ruby Beholder is of limited, ahem, value. They have green versions out there as well, just for red fabric lovers.
The above picture shows almost the entire value discussion in Color for the Terrified Quilter: Plain Talk, Simple Lessons, 11 Projects by Ionne McCauley and Sharon Pederson. Considering the relatively narrow scope of this book, I am disappointed in the lack of attention paid to value. Essentially it includes two paragraphs and two exercises, one organizing fabrics into a value scale and another addressing relative value. All of the pictures are full color, which I think might confuse folks. Also, the exercises only use one hue. Arranging a value scale within one hue isn't really the problem. Trying to compare values among varied hues is the real challenge most quilters face. This book doesn't even broach the subject.
This is one of my favorite depictions of value from Deb Menz's Color Works: The Crafter's Guide to Color. Not only does Menz dedicate part of her chapter on describing color to the subject of value, she dedicates a whole chapter to understanding value. Most importantly, she uses color and grayscale images to illustrate value differences among different colors. She even has a diagram depicting the twelves hues on her color wheel with their values. Genius!

Click here to see the rest of this post...

Color Theory: Value

After my abysmal showing at Exercise 1.3, I thought I should back up and take a closer look at value.

If you click on any of these images you should be able to see a larger version of the image in which the values for each Munsell chip are labeled in hot pink. The first value scale accompanied by Munsell chips is from Color Works: The Crafter's Guide to Color by Deb Menz, which was recommended in a quilt guild lecture on color theory by Robin Edmundson.

The first thing I notice when I look at this picture is that the grays of the Munsell chips seem to have a yellow cast to them compared to the color of the Menz grays. The Munsell chips are actual chips of opaque paint which has a different color gamut than the printers ink used to make the Menz book.

This scale is from the same book as the first example, Color Works: The Crafter's Guide to Color by Deb Menz. The back of the book has a number of handy pull out color tools including a gray scale. It's printed on glossy card stock, which might explain the differences between the arrangement of the Munsell chips on this scale and the first scale.

Note the unequal distribution of Munsell chips along the Menz value scale. Particularly at the high end of the value scale there seem to be a lot more grays in the Menz scale than the Munsell chips. Menz creates her value scale according to the percentage of black that has been mixed with white, from 0% (white) to 100% (black). In the value scale at left you can see the numbers along the left side of the value scale which indicate the percentage of black that has been mixed with white to form that particular gray. This gives us 11 different values.

On the Munsell Hue/Value/Chroma chart we have 9 rectangles and only 8 repositionable chips as representative samples for N 1/ (black) cannot be achieved in matte finish. Munsell created his gray scale based on ten visually equal steps between black and white. Because human perception of lightness is not uniform, a gray reflecting 50 percent of the light falling on it does not appear halfway between black and white. On the Munsell value scale, the middle gray, N 5/, reflects only 19.77% of the light. This is because people are much more sensitive to value differences between dark colors than they are to value differences between light colors.

For this reason, the steps in the Munsell value scale have a geometric relationship to one another. Munsell would argue that Menz's gray scale does not appear uniform because the percentage of black that has been mixed with white increases in equal increments.




This value scale is from Art History by Marilyn Stokstad, which I'm reading with a couple of friends to make up for our failure to take art history in college. This value scale and some other basic tools to describe art are included in the Starter Kit section of the introductory material. For no particular reason that I can discern, values are divided into seven gradations including white and black at either end. Their correspondence, or lack there of, to Munsell value chips are apparent from the super out of focus picture.

Click here to see the rest of this post...

Munsell Exercise 1.3: Determining Values of Colors - Checking My Work

It's times like these where doing these exercises in a class would be both harrowing and helpful. I'm relieved I'm not being graded on my ability to determine the values of colors. But I would really like some feedback both on my technique and how to double check myself. Following, please find my thoughts on how to check the assessment of value of a color using an image editing program.

In my image editor of choice, GIMP, if I select Curves from the Color menu and look at the Value graph it appears that I have one big spike in value right around the middle and a tiny curved bump in the top quarter, which I'll blame on the non-green or gray portions of the image. This seems to confirm my guess as to the Munsell value of this green.

I have a few different options for turning color images into greyscale images. One option is to desaturate the image. There are three ways to desaturate an image: lightness, luminosity, and average. Desaturating by lightness has the exact same effect as selecting the Hue-Saturation option on the Color Menu and sliding the Saturation toggle to the bottom.

Desaturation by luminosity appears to confirm that the chip and the background are the same value.





Desaturation by the average (I'm guessing the average of lightness and luminosity), as one might expect, results in an image about halfway between the relatively high contrast of the image desaturated by lightness and the very low contrast of the image desaturated by luminosity. Not sure it actually tells me much.


I started with the second example first because this first example is so clearly wrong. When I pull up the Value graph, I get at least two distinct bumps, implying I've got at least two distinct values in this image. If I were correct, there should only be one bump because the value of the blue background and the value of the gray chip should be the same. At least that's my layperson's understanding of the Curves/Value graph in GIMP.
I didn't even bother to run the average desaturation considering even the luminosity desaturation didn't make these two look like the same value.



The original sample for this image included a lot of other colors than the one with which I was trying to work, so I trimmed out the other colors to check it. Just like the previous example, when I pull up the Value graph, I get two distinct spikes, implying I've got at least two distinct values in this image. If I were correct, there should only be one spike because the value of the orange background and the value of the gray chip should be the same.



But, unlike the blue example, this desaturation by luminosity actually looks pretty good. How can my Curve have two spikes, but my desaturation by luminosity looks almost as good as the first example?




Granted, the average looks a lot like the desaturation by lightness, implying that it's more different than it is the same, but it still seems like a closer match that the blue example.

Click here to see the rest of this post...

Munsell Exercise 1.3: Determining Values of Colors

In my ongoing exploration of color theory centered on The New Munsell(tm) Student Color Set, I've completed another exercise. This time the goal was to determine the Munsell value for a number of colors. Judging values of colors is always a challenge for me. First of all, I thought this was a grayed color. When compared to a chip of actual gray, it doesn't look grayed at all. The chip is Munsell(tm) value 3/, which was my best guess for the value of this color.


I did this exercise in one lighting situation and then when I went to photograph it in another lighting situation I realized my answers were WAY off for a few of them.

The green in this image - which is a tiny segment of a magazine reproduction of what I suspect is an acrylic painting - I suspect might be Munsell value 5/.


The yellowy-orange color from another tiny segment of the same painting in a magazine might be Munsell value 6/.


Under the six chips, this is a portion of Giorgio Morandi's "Natura Morta" (1954) as reproduced in the September 22, 2008, issue of The New Yorker magazine. If you click on the picture, you can see that I've labeled each chip with its value to indicate what I think the adjacent colors' values are.

Click here to see the rest of this post...

The Color of Heartache

Susanna Clarke describes colors amazingly in Jonathan Strange & Mr Norrell. For example on page 610 she writes,

The box was small and oblong and apparently made of silver and porcelain. It was a beautiful shade of blue, but then not exactly blue, it was more like lilac. But then, not exactly lilac either, since it had a tinge of grey in it. To be more precise, it was the color of heartache.

This has, of course, caused me to spend excessive amounts of time pondering my New Munsell Student Color Set for the color of heartache.

I've narrowed it down to somewhere between 5P 6/2 and 5PB 6/2. What do you think?

Munsell Color Charts

In my hour spent creatively I finally finished Exercise 1.4 in The New Munsell(R) Student Color Set. The exercise basically asks you to complete the book by filling in the blanks on the color charts with colored chips. I've blogged about Munsell's color theory before elsewhere. I've also blogged about my specific experience assembling the color charts, back when I'd finished only the first three out of ten. Here I thought I'd share my pictures of the completed charts with a few observations.

5R with 30 chips. The chip that matches the 5R chip on the hue circle of the Hue Value/Chroma Chart is 5R 5/14. I started with the 5R chart because a previous chart illustrating the concepts of hue, value, and chroma used one row from the 5R chart to illustrate chroma. Also 5R contains the most chips, 30, so I could check all the rest of the charts against that chart if there wasn't the same value and chroma on the adjacent hue charts.

5YR with 24 chips. The chip that matches the 5YR chip on the hue circle of the Hue Value/Chroma Chart is 5YR 7/12. Then by chance I worked clockwise around the hue circle. This turned out to be a good plan because in that order the charts with the most chips come at the end. Charts with fewer chips are easier to arrange.

5Y with 21 chips. The chip that matches the 5Y chip on the hue circle of the Hue Value/Chroma Chart is 5Y 8/12. Do my charts look right to you? Any chips seem out of place? This is the moment when I wish I could work on this with a class so I could have a group of well-trained eyes to double check my charts.

5GY with 20 chips. The chip that matches the 5GY chip on the hue circle of the Hue Value/Chroma Chart is 5GY 8/10.

5G with 22 chips. The chip that matches the 5G chip on the hue circle of the Hue Value/Chroma Chart is 5G 6/10.

5BG with 20 chips. The chip that matches the 5BG chip on the hue circle of the Hue Value/Chroma Chart is 5BG 5/8.

5B with 20 chips. The chip that matches the 5B chip on the hue circle of the Hue Value/Chroma Chart is 5B 5/8. It certainly got easier the more charts I completed. By the last few I had no trouble pulling out that first column /2 with the lowest chroma.

5PB with 27 chips. The chip that matches the 5PB chip on the hue circle of the Hue Value/Chroma Chart is 5PB 5/10. Once I figured out that if I set up the chart as I thought it should be on medium gray paper then switched the position of chips I wasn't sure about those switches really pointed out mistakes better than comparing to other charts.

5P with 26 chips. The chip that matches the 5P chip on the hue circle of the Hue Value/Chroma Chart is 5P 5/10. The only bummer is I'm almost 100% sure I got two of the same chip for my 5P chart. So I have two 5P 8/4 chips and no 5P 7/4 chip. I'm going to write to the publisher, Fairchild Publications, and see if they'll send me a single 5P 7/4 chip.

Finally 5RP with 28 chips. The chip that matches the 5RP chip on the hue circle of the Hue Value/Chroma Chart is 5RP 5/12.

Click here to see the rest of this post...