All language subtitles for 03_colour-models-hsv.en

A second color model is hue, saturation, and value.

So, it's not only that it's a different color model,

it's a different way of thinking about it.

Instead of looking at it as a queue of red, green and blue,

now we're going to look at a cone of hue,

saturation, and value, or HSV.

So, this is supposed to be more intuitive than red, green and blue.

I mean red, green and blue was really designed in the early days of

computers based on what was best for a computer monitor.

Hue, saturation and value is supposed to be a

more intuitive and more human-oriented if you want.

I don't know, I guess it's because I learned RGB first.

I'm not sure if I think in hue,

saturation and value, but I can see the idea and certainly for cartography.

It can be useful to think about things in

terms of how can I change the hue, or the saturation,

or the value in terms of say upgradation of values as we'll

see from a dark red to a light red,

and doing that by changing the saturation and so on.

But, I'm getting ahead of myself,

let's just have a look at it first.

So, the idea is that hue is described based

on where a color is located around the circle at the top of the cone.

This is done based on degrees around the circle.

Saturation is going from a complete lack of color,

which would be white in the center to the maximum amount of color.

Let's go over here, let's say,

over here to a completely saturated version of that hue.

Okay. So, think of it as going from,

or if you want to take it the other way around,

going from a maximum saturated blue or red or whatever the color is.

Then as you move inwards to the center of the cone,

hue is getting more and more white attitude or you can

think of it as being more washed out or less saturated.

That's the way that they would think of it in terms of the HSV cone.

Then value is going from white at the centre here to black at the bottom of the cone.

So, you're going from white to black,

or if you want to think of it like you're adding more

black to the color so that you're going

from maximum amount of saturation on the edge to less saturation towards the middle.

If you're adding black to it at the same time,

then you're actually moving your way down the cone as well to the bottom.

So, it's shrinking down to the bottom because as you add black to it,

all of the colors tend to look more and more similar to one

another until they end up with just a pure black at the bottom.

So, as I said, the hue is defined based

on where it's positioned at the top of the circle.

So, within the ArcMap color selector,

this is based on using degrees here.

So, zero degrees is pure red,

and 120 degrees around the circle from that pure red would be pure green.

Then if we go another 120 degrees around the circle,

so this will actually be at 240 degrees,

that's going to be a pure blue.

So, what we're doing is working our way around the circle.

Of course you can do any degree increment along the way around that.

I'm just matching it up or showing you how this relates to red,

green and blue in terms of the RGB color model.

But of course, the idea is you're getting any kind of range

of colors around the top of that circle,

based on what degree they are from zero being a red all the way around 360 degrees,

which will come back to the same red.

Saturation is from the center to the outside as I said,

and so you can see here for example that we have the same red.

So, this is still zero degrees just like it is here,

so we're defining the same hue still,

but now we have 50 percent saturation instead of 100 percent saturation,

and you'll notice that it's now halfway between the middle and the outside.

So, we've gone from the same hue,

but just less saturation.

So, now we're at 50 percent, and you can see that you get this

more washed out looking red.

Then we can modify the value.

So, here again, we're using the same red hue,

we're using 100 percent saturation but now we're at 50 percent value,

and so you end up with this kind of brick red.

So again, same hue and we're back to 100 percent saturation,

but we've modified the value.

So, you've added black to it and so now we've got this darker looking red,

and so this is a nice way I think any way of being able to see like "Oh!"

So, when I'm actually modifying these things inside this little dialog box,

you can visualize in your mind what's going on with this cone and how you're

positioning that color in relation to the overall HSV color model.

This is just a nice little comparison between the RGB cube and the HSV cone,

and again I'm really just trying to emphasize this idea

that you're just using two different ways of defining the same color.

So, you can have that same red,

whether it's the full red or the brick red or whatever,

you're just using two different ways of specifying that for the software.

So, just some definitions,

we think of hue as the dominant wavelength,

that's actually what most people normally think of as a color as you

have like a green or a teal or orange or whatever,

but really the better way of thinking of that is hue,

that's the actual wavelength.

If you want to think of it as parts of

the electromagnetic spectrum or part of the colors of the rainbow,

that's what we're talking about when we say hue.

So, we can have different hues for different categories of data,

if you are using say nominal data,

something where you want to tell things apart like say land use,

you have industrial areas versus commercial areas,

you could use a different hue in order to be able to tell those apart.

Saturation is arranged from white to pure color.

So, I'm just summarizing the definitions here.

So, the way that people tend to

interpret saturation is that the more saturated something is,

the more important it is or the higher the value is.

By value, I mean the number associated with that particular location,

not value in terms of the color model.

When I think of value in terms of the HSV color model,

that's the brightness or how light or dark a color is with the same hue,

and so darker is interpreted as being more important or of greater magnitude,

and that's definitely something to keep in mind.

If you're trying to show a gradation of values from say high to low,

so this can be whatever temperature,

something like that, then you could show that based

on changing the saturation or you could do it based on value,

but you have to think about.

So do this intentionally,

how are people going to interpret this and how can I modify these parts of

the color model in order to help

them interpret them in the most easy efficiently possible?

So, for relating the HSV color model to levels of measurements,

with hue we can we can map qualitative or nominal data,

like I was saying, things like land cover can be urban crop force and so on.

So, the idea here is that you want to make it

clear to your map reader that these are distinct from one another.

Saturation is good for quantitative data whether it's ordinal interval or ratio data.

Essentially, if you have a sequence or upgradation of values,

you can work with saturation to go from less

saturated to more saturated to be able to show that gradation or sequence of values.

Basically, the same thing is true for value,

again it's ordinal interval or ratio data.

But notice here that we have low numbers in our datasets,

have a higher value or they're less dark if you want,

and if you have a higher number that you're trying to show somebody, so it could be,

like I said, temperature or amount of corn harvested from a field or whatever it is,

that you're still able to show that gradation of values,

but now you're doing it in a different way than you were here with saturation.

Both of them are perfectly fine,

depends on the look you're going for or what's most effective

based on other variables or things that are going on with your design.

So, think of these as like these are options that

are available to you when you're designing your map.

So, I put together some apps just to show you what happens when

you isolate one of these parts of the color model.

So, for example here,

I've made a map strictly based on modifying saturation,

and this is population density for census tracks in Toronto.

All I did is I kept the hue the same,

so these are all zero,

so they're all the same, red,

and the value is the same,

they're all at 100 here.

The only thing that changed was the saturation,

so I went from 20 percent to 40 to 60 to 80 to 100,

and that's how I'm getting this range

of colors or amount of saturation in my legend that's being reflected on the map.

It tends to work well.

Saturation is a nice one to modify because you can get

this really nice gradation that is good for this type of map,

it's called a choropleth map,

so it's easy for people to interpret that they automatically want to think of areas with

less saturation as being a lower number

and areas with higher saturation as being a higher number.

I'm going to do the same thing again, only now,

I'm only going to modify value,

and again this is not the value as in terms of the population density.

The census track is the value in terms of the color model,

and so here, the only thing that's being changed is the value.

So, I've gone from 100 to 90 to 80 to 70 to 60.

You notice that I'm not doing it by

20s like I did in the last one because it just didn't look very good.

So, I was trying my best to be able to make it look at least decent.

I'm not thrilled with this map,

but I wanted you to see what it looks like if you just changed value.

Of course, I want to make sure it's clear,

you can mix and match these,

you can modify both the value and the saturation at the same time.

I'm just trying to isolate them so you can see what

the difference is between them and get a sense of how they work.

So, definitely here, you've got lighter areas for lower numbers,

and darker areas which are a higher value in the color,

this is confusing, isn't it?

Higher value for their color model,

and so you do get that gradation and I do believe that people would

interpret this as this is an area with higher population density,

this is an area with lower population density.

Even though I mentioned earlier that hue was often used for

nominal data to be able to tell it apart things like land use or land cover,

it can be used to show a gradation of values if it's done carefully.

Make sure you're getting the right effect that makes sense.

So, I tried it here and I think it worked pretty well.

I went from a very light yellow to a

darker red with a gradation of these warm colors from low to high.

I'm not saying it's the best way to do this or perfect,

but I was trying to show what could you use hue

in the same way as we did with saturation and value.

So, here the saturation and values are kept constant at 60 and 100 for all of them,

and the only thing that's being changed is the hue, so 60,

45, 30, 15, and zero.

I think it works pretty well.

I'm not super thrilled with it,

I think that somebody might confuse especially the reds and the orangey-brown areas,

like which one is higher.

Obviously, if you see the legend,

you'd be able to tell them apart, but ideally,

someone should be able to interpret it without the legend,

at least get a pretty good sense of what's going on.

So, like I said, I think it's pretty good, it's not fantastic,

but at least it shows you how this could work with only modifying hue.