It would be ideal to have the ability to modify dot size when using the Dot Density symbology. It looks great when viewing the map at a smaller scale, but when I zoom in, the dots practically disappear. The "Vary dot value by scale" feature is disabled, and I've tweaked the coloring and basemap to make the dots stand out, but the dots are just too small when zoomed in.
Is this functionality, or something similar, planned for a future release?
Solved! Go to Solution.
Yes! We would love to do this. Currently the reason why we are able to quickly draw dots is that we aren't really drawing dots at all, we have a texture (basically large image where each pixel has a random value) and fill in pixels ("dots") according to a threshold such that we wind up with approximately the desired dots within a given region, so unfortunately giving explicit control over the dot size is a little bit more difficult than it might seem. To add this, we need to add a geometry extrusion step (create real dots where we are currently just filling in the pixel) which will allow us to scale the dots arbitrarily.
Thanks for the through feedback! I appreciate you explaining some of the intricacies of dot density symbology. I'm using a 4K monitor as well, which explains why seeing the dots might prove even more difficult.
Great work so far on the new viewer! It already seems easier to use than the existing web map interface.
I also would like to be able to change the size of the dots when using dot density. I'm producing a web application that currently could use dot density, so I started researching the beta map viewer and actually trying it out. Got it to produce the dot density really well, however the dots are so tiny. Would really like to increase them. It's great this is already planned for a future release, but I'd like to see it with a higher priority as well. Changing the size of a symbol is such an important part of mapping, how can this be left off in the first place?
Hey Richard Littlefield,
Yeah. I agree that the dot size does feel quite small at times. Can you provide a few more details on where increasing the dot size will help in your case? We have technical reasons (described by Matt above) for not allowing this for now. Because it's not a true symbol (hence the term 'dot') increasing the size in the current implementation yields renderers that don't look that great.
However, I find that I do want to increase the size if I'm zoomed to a scale where dots are more dispersed. Take for example this image where I have a dot size of 1 pixel.
The dots are harder to see, especially with high dpi displays. Increasing the size helps me see the dots more clearly.
Ok. They're more visible here. But from a cartographic standpoint, the problem with changing the size of the dots at larger scales gives the map reader a false impression that these are not random dots...they tend to look like actual points that would hold some kind of geographic meaning, when in fact this is not the case. Even when it is understood that these are random dots, they are meant to convey a pattern of density that doesn't work at this scale. I can't tell where more people live or don't live in that image. So this scale isn't suitable for this map. When I constrain the map to the scales where the patterns of density are actually visible, then the 1 pixel dots are actually the best size. Two-pixel dots are too large in that range.
So dot density gets really tricky in this sense. It only really works at just a few scales, which depends on the extent of your layer and the size of each polygon. And making the dot size too big can alter the map reader's perspective that these are points, not random dots.
That being said, I agree with you. We need to provide an ability to change the size of the dots, but anything larger than 2 pixels per dot probably won't work well.
We just have a static PDF map that uses dot density. It shows population diversity across a three county area. The map allows you to change the dot size, and the dots are shown larger in our static PDF map. I was trying to recreate this in a web application. When doing this I just felt that the dots were too small, like you show in your first example. I completely understand though, that the different scales of a web application make this difficult and will make the symbols not proportionate at certain scales.
Maybe an idea would be to create recommendations/tasks/workflow/blog on how to set this up at multiple scales, but using multiple versions of the same layer with different symbol sizes and using scale dependency? Similar to how basemaps are set up. Use the zoom levels groups of a web map (esri, google, bing, etc.) all with scale dependency, so they'll change to a better suitable symbol when you switch to a different scale.
That sounds like a great idea. I'll add that to my plan. Just note there's a lot going on right now, so this may not make it in our next release. We'll discuss more on our end.