CMYK colour chart

[timd]

Following Eben’s assumption – a book makes sense in terms of longevity/user friendliness, a poster would be more affected by environment, sunlight, dust etc. A user friendly addition (essential, really) is to include a grey (or neutral – possibly light on one side and dark on the other) frame that can be used to block out neighbouring colours.

Tim

[ill sans]

Even in case of a 10% range it would turn out to be quite the wallfiller which would look nice, but indeed in time would be suffer from sunlight & dust. Basically I'm hoping that this kind of diagram can find its use in other fields & there is an orginasation in Belgium that subsidies scientific research (or so I've been told, I haven't checked it out yet). I'll wait for SABAM's reply first & later on might contact them to finance a potential patent.

[ebensorkin]

Who were you think would use your diagramatic model beyond presspeople & designers? What qualities other than CMYK are you thinking of showing?

[ill sans]

It might prove to be usefull in mathmatics or sciences. I'm not a scientist, but if it can be used to compress a CMYK-chart to one diagram, I'm pretty sure it could be usefull to combine other diagrams to a single one as well. I have no idea where else you would have to input 4 different values for each element, but maybe some scientist does?

[wormwood]

I havn't given this much thought so I'm probably talking nonsense, but could the Lukac Quartogram work for RGB with Brightness as the 4th variable?

I'm really looking forward to seeing it.

Have you thought about calling it the Lukac Quadragram? Just a suggestion because dictionary definitions say quart is a unit of capacity and quarto is a page size.

[ill sans]

I haven't decided on the name yet except for the Lukacs part ;-)
I have actually also considered quadri-/quadro- or quadragram. I'm working on a RGB-model right now, but it's a lot trickier. I'm a bit confused as to what you define as brightness... Isn't that just the percentages of RGB combined (the higher the percentages, the brighter)? The difficulty that I'm having with the RGB-model is that I don't have a 4th variable & 100 (%) can not be divided by 3 (colours). I'll probably end up with a few doubles in the model if I end up with anything at all. The only way I see it working now is by making a different model for each given percentage range, but again I can't seem to add the numbers up correctly so far. The only hope I can give you for now is that I'm a very stubborn person & I reàlly want to make an RGB-model as well ;-)

[wormwood]

Sorry, too much wine. I said I was probably talking nonsense.

Obviously a percentage of R, G or B is the amount of brightness. Doh! (Slaps hand on forehead)

I think the Photoshop colour picker holds a clue. Spot the odd one out...

HSB
Lab
RGB
CMYK

[ill sans]

Maybe wine is the answer to my problems? ;-p

[wormwood]

After a couple of bottles you could try an Elemental Quartogram...

Earth / Air / Fire / Water

75% Air + 25% Water = Cloud

50% Earth + 50 % Water = Mud

50% Fire + 25% Earth + 25% Air = ????

%^}

[wormwood]

How about a Typographic Quartogram?

The categories could be Serif, Sans, Script and I'm not sure about the 4th.

It would be quite personal in choosing a font to represent a 100% purity of its category and what the scale of interpolation would be.

My mental type catalogue is not big enough to know if it would be possible to fill in the transitionary stages between the 4 categories but there do seem to be a lot of hybrid typefaces these days.

It might be a fun Typophile challenge.

[ill sans]

I'll pass on that challenge ;-) The RGB chart is enough for now, but once I've posted an example here and an explanation of how to put 4 different values into one quadrigam (or whatever you want to call it), anyone is free to do with it what they want, but I don't think it will work as a font schedule though. I think it can only work with quantities of the same category and not for 4 different subdivisions... Either way, the alcohol is doing its work now & giving my mind a rest ;-p

[Nick Job]

10% increments (from 0 - 100%) of all four process colours are going to give you 14,641 different coloured squares (a lot of them very black, which is why the tattoo guy only did up to 80% black). How you lay them out is going to hurt so I hope it is worth the pain.

[ill sans]

Well, like stated before, it's about more than just a colour chart now. I have to admit that it might not proof very practical as a CMYK chart (unless with bigger leaps as in my prototype), but it was really about the challenge of being able to do it. And now, I'm just hoping somehow it can be useful as a way of compressing several diagrams into one.

[elliot100]

This project reminds me of Visibone's colour charts which unwrap the RGB 3D colour cube in various ways:

http://www.visibone.com/

I am very intrigued how you've unwrapped the 4D hypercube... if you want to find out if anything similar exists I would perhaps look in the directions of maths projections of this hypercube

http://en.wikipedia.org/wiki/Tesseract

[ill sans]

Wow, that stuff is wày over my head ;-p
Actually, the idea behind my quadrigram is quite simple (which is probably why it took me a while to figure it out in the first place) & -unlike I myself wrongly stated above- it càn actually be used as a schedule for 4 subdivisions as well. It can be used in any way to give 4 different "values" to each element (or categorise an element into 4 different subcategories), but its practical use will most likely lie in maths, scientific or maybe economics as a way of compressing several diagrams into one.

[ill sans]

I figured out a way to make an RGB-model as well, in steps of 17. It may seem a bit of an odd number for leaps, but this adds up to 4096 variations which can be translated to a 64x64 scheme. I was off on the wrong foot with this at first because I thought in measurements of 100 instead of 255 for each colour. BTW Why didn't anybody correct me on this? ;-p
I'm not really happy about using the same skeleton as for my CMYK-model & leaving one variable out, so I'm still going to try & put it in a triangular model which obviously makes more sense.
Meanwhile, I still haven't heard from SABAM yet, but I did contact the organisation that subsidies innovations... More news soon.

[peterj]

As for naming, you could always switch to Greek and make it some variation of "tetragram," which the internets tell me is a word of four letters, or an abbreviation for the Hebrew god.

[ill sans]

The name will be a last minute decision since I'm still waiting to see if this kind of thing already exists or not. So far, so good though & I have contacted several organisations to try and get this thing "patented". Knowk on wood!

[yahhoo]

well you might like my cmyk chart hehe...each page increments of 5% = 400 swatches per page
1. the 6 pages that are only two color
2. the 21 pages of CMY only
3. THEN the same base 21 pages of CMY with 10% K
4 THEN " " " with 30 % K
thats roughly 27,200 swatches....now if a guy wnated some really dark pages....you just go into "channels in illustrator with the 21 pages of CMY on your screen and add
whatever percent K (black that you like)
http://motionpicturesigns.com i made three charts for my 54" $40k printer....a CMYK,a RGB(2800swatches), pantone bridge (I corrected pantones pdf chart which was in fact wrong)...all 3 yours for $30!!..no lie

[ill sans]

It definitely looks great (especially the RGB-model)! My intention however was more "scientific" rather than practical... I wanted to put it all in one oversized chart WITH all percentages of black & WITHOUT any doubles (except for every combination with 100% K or the ones where all 3 C, M & Y percentages are 100%). I haven't really been doing anything with it lately, but I still have plans to post my idea in a iDepot online to assure my rights & then I'll post it here. But just to be on the safe side, don't hold your breath ;-) I've been working on too many things lately to even be bothered to pick this one up. More news will follow some day... probably.
BTW, I still haven't had any chance on the RGB-model, but again, I haven't been trying lately ;-)

[ill sans]

I have to be honest & admit that I got bored with trying to find where to go to with this idea & kind of forgot about the whole thing until I was recently going through my archives. Since I don't really see myself picking up on that search I figured I might as well let you all in on my 'secret' (you've been patient enough ;-)).

Here's what the Quadrigram looks like with percentages of 25:

As you can (hopefully) see in the quadrigram the values for black, cyan, magenta & yellow are read respectively on the left, top, bottom & right side.

Here's how it's done:

To put 4 variables into one chart you need to create 4 axes to read your values from. Average diagrams use only one X-axis & one Y-axis making no use of the top & right space as far as value input goes.
Having 2 horizontal (X1 & X2) & 2 vertical (Y1 & Y2) axes creates the need for different calibrations. This can be achieved by making use of small & big 'squares' for values for both axes (horizontally as well as vertically). The first thing you need to do is figure out the top value of one of your axes (for percentages it's obvious, but in this example I will use a top value of 4).

For the second axis we need to divide each seperate value into these 4 steps; in this case creating 5 blocks (0,1,2,3 & 4) of paired values 0-4.

Now we do the same with the vertical axes to create a raster of big squares & one of small squares merged together & we'll thus be able to read 4 different values for each point (in the example the values for the coloured square are 3 for X1, 1 for X2, 3 for Y1 & 2 for Y2).

This method can be used for any kind of chart that needs to combine 4 different variables although for other charts it will lose the 'flowing' line that gives a first good impression of what to expect when reading the chart. Still it can be used to compress all information into one userfriendly chart.

[oldnick]

FWIW, your charts ought to exclude any combination of colors which exceeds 260% (e.g., 100C 20Y 100M 40K) because, if you run higher percentages through an offset press, the ink will never dry...

[JamesM]

I have a reference book that has all printable CYMK combinations in 5% increments. It must be tens of thousands of swatches. If you tore out the pages you could wallpaper a wall. Can't conceive of fitting it into a poster.

[oprion]

@oldnick

Funny enough, I've actually made a similar chart with the 260% range marked out in the past!

[Theunis de Jong]

Interesting, so I wrote an Illustrator CS4 Javascript (save as "CmykSpread.jsx") that draws the above :-)
Select any four-pointed object (not limited to rectangles!) and run the script. It should create something like the above, or -- it took me two days to get the linear interpolation working -- something like this from a free-form four-point object:

I added Oprion's ink limit warning because that's a real good idea.

---- (snip script below) ----
//DESCRIPTION:Convert Rectangle to CMYK Spread
// Jongware, 15-Jul-2010
if (app.documents.length > 0 && app.selection.length == 1 && app.selection[0].constructor.name=="PathItem" && app.selection[0].pathPoints.length == 4)
{
sel = app.selection[0];
steps = prompt ("Number of steps (1..100)", 5);
if (steps != null)
{
inklimit = prompt ("Ink limit (percentage)", 260);
if (inklimit != null)
{
inklimit = Number(inklimit);
steps = Number(steps);
if (steps >= 1 && steps < 100)
{
doc = app.activeDocument;
p0 = app.selection[0].pathPoints[0].anchor;
p1 = app.selection[0].pathPoints[1].anchor;
p2 = app.selection[0].pathPoints[2].anchor;
p3 = app.selection[0].pathPoints[3].anchor;
dx = (p1[0] - p0[0])/(steps*steps);
dy = (p1[1] - p0[1])/(steps*steps);
dlx = (p3[0] - p0[0])/(steps*steps);
dly = (p3[1] - p0[1])/(steps*steps);
drx = (p2[0] - p1[0])/(steps*steps);
dry = (p2[1] - p1[1])/(steps*steps);
xl = p0[0];
yl = p0[1];
xr = p1[0];
yr = p1[1];
white = new GrayColor();
white.gray = 0;
allpts = new Array();
for (v=0; v<=steps*steps; v++)
{
xp = xl;
yp = yl;
dx = (xr-xl)/(steps*steps);
dy = (yr-yl)/(steps*steps);
ln = new Array();
for (h=0; h<=steps*steps; h++)
{
ln.push (Array(xp,yp));
xp += dx;
yp += dy;
}
allpts.push (ln);
xl += dlx;
yl += dly;
xr += drx;
yr += dry;
}
c = 0;
m = 0;
y = 0;
k = 0;
for (v=0; v<steps*steps; v++)
{
c = 100*(v % steps)/(steps-1);
m = 100*Math.floor(v / steps)/(steps-1);
for (h=0; h<steps*steps; h++)
{
y = 100*(h % steps)/(steps-1);
k = 100*Math.floor(h / steps)/(steps-1);
myPath = Array (
allpts[v][h],
allpts[v][h+1],
allpts[v+1][h+1],
allpts[v+1][h]
);
p = doc.pathItems.add();
p.setEntirePath (myPath);
p.closed = true;
p.filled = true;
p.stroked = false;
color = new CMYKColor();
color.cyan = c;
color.magenta = m;
color.yellow = y;
color.black = k;
p.fillColor = color;
if (c + m + y + k > inklimit)
{
p.stroked = true;
p.strokeColor = white;
p.strokeWidth = 0.5;
p = doc.pathItems.add();
p.setEntirePath (Array(allpts[v][h], allpts[v+1][h+1]));
p.closed = false;
p.filled = false;
p.stroked = true;
p.strokeColor = white;
p.strokeWidth = 0.5;
p = doc.pathItems.add();
p.setEntirePath (Array(allpts[v+1][h], allpts[v][h+1]));
p.closed = false;
p.filled = false;
p.stroked = true;
p.strokeColor = white;
p.strokeWidth = 0.5;
}
}
}
} else
alert ("Invalid number of steps -- please use between 1 (few) and 100 (lots)");
}
}
} else
{
alert ("Please select a template rectangle");
}
}