Going from a larger to smaller area

[Hikki-Jin]

'Kay, so, I'm working on some maps featuring a setting based on a fictional version of Japan. One map will depict the whole archipelago (map A). Eventually, though, I want to make other maps focusing on individual regions within islands (ex. map B). I'm using Photoshop CS5.

Map A has a scale of 9px = 1mi². As it is, map B is just a crop of map A, and depicts approx. 220,086px/24,454mi² of land.

What I'm wondering is the best way to go about making map B larger, allowing greater detailing of the region, but keeping the area and shape and everything consistent with map A.

[Azélor]

I believe there is a problem with you numbers, unless it's intentional.
In A, the size of each pixel is 0,111... mi2
In B, each pixel is 0,102...mi2

That's almost exactly the same scale. Actually, when I look at you maps here, the scale of B looks larger than A, but when I import them in my PS, the scales are identical.

The best way to enlarge the map would be to use a vector program like Illustrator or Inkscape. Using vectors, you can enlarge without loosing any details although I don't recommend making a map only with a vector program especially if you plan having a lot of elements or using textures. If you enlarge it in PS, it's going to look blurry, unless you intend to remake the map using the original only as a model.

How big do you want B to be in pixels x pixels?

[johnvanvliet]

photoshop like gimp,krita,Cinepaint,nip2,and so on are all RASTER image editors and have NO !!! 0!!! ZERO!!! concept of scale or pixels per degree or pixels per Meter

100 pixels is 100 pixels no matter what
( printing on paper -- ok that dose need pixels per inch or CM)

even most vector SVG will be the same on every scale while zoomning


now to add MORE information in a area you have to well ADD information to it

for a raster editor that is painting in that new more information

basically this is NOT easy . it is MUCH easier to remove information than it is to add information that is not there

there are a few "tricks"

for a DEM ( digital Elevation Model ) one light erode cycle in Wilbur on a X2 or X3 ( 16 bit or 32 bit float) map works rather well .

for SOME landscape just adding 0..2 to 0.06 % Gaussian noise works fairly well -- SOMETIMES

but for the most part you have to paint by hand the things that are now visible if you move form 10 px = 1000 Meters to 10 px = 100 meter to 10 px = 10 meters

it is MUCH easier to go the other way

that is why on TV when you see the " image enhancement "( HA-HA-HA-HA) they start with a GOOD sharp image than make it look bad and say the bad one was the original ( the TV lie )

[Gamerprinter]

My solution to your problem probably won't work for you. Instead of using an image editor like Photoshop or GIMP, I use a vector drawing application like Illustrator, though I use Xara Designer Pro x11. In vector pixels mean absolutely nothing, until you export your file to image format (.JPG, .PNG, etc.) In vector a shape is determined by the points (and arcs) on the outline determining the shape. If you rescale a shape of points the integrity of the shape does not change. So your coastal outlines remain the same no matter how far you zoom closer.

In fact when I create a map using vectors, when I do export the file, I can export the same map to 50 ppi or 300 ppi, depending if it's going used in a virtual tabletop application or if it's going to printed in large format (such as 24 x 36 inches in size). I can also select the entire shape and physically rescale smaller or larger, if I need to (I usually don't). Now if you use photo image texture fills to fill the shapes, if you rescale larger JPG artifacts can show up (like when trying to rescale a bitmap image, like what the other posters are talking about). I generally design my maps at large format scales such as 72 x 144 inches in size, but usually at 18 x 24 or 24 x 36. I am especially conscious of designing encounter scale maps (1 inch = 5 feet scale) at that scale. Again when working in vector I work with real measurements (inches, centimeters, etc.) which are completely independent of pixels (which again means nothing in vector).

One thing though, your existing map sample could easily be replicated in vector, though most people think vector can only be solid color type maps, this is not true. All my photo-realistic maps are vector creations, though most people assume I created such in an image editor like Photoshop - I use vector (Xara) only for all my map work.

As an aside, that 5th map posted below, is my fictional Japanese archipelago analog setting, which I've already published - that map will be used in my upcoming Gamemaster's Guide to Kaidan due to be printed within the year. It is for my Kaidan setting of Japanese Horror (PFRPG).

GP

Here are some sample maps of mine done in vector...

skate-class-raider-trader.jpgentire-village-11C.jpgcastaway_final.jpgcastle-hill-done.jpgkaidan-final-map-tweaked.jpg

[Redrobes]

He said B was just a crop of A so I would imagine that the map scale in meters per pixel stays the same.

Whilst 100 pixels is 100 pixels, if you apply that to a map which represents real world scale then it now has a pixels per meter scale. It is true that if you add more pixels for a given area then you have to add detail to the image. If you want to add random detail to it then you can add noise and then clamp the colours back to some default and some of the noise makes for a new more crinkly coastline. If your working with an area inland then all the layers and all of the detail has to either be manually added or you have to have something fancy like a terrain simulation such as Wilburs erode going on to get it to do something more meaningful than just random.

In terms of using raster or vector, if you scale it up then you have the same identical problem in that both systems are lacking information that has not been described. If you scale up vector then sure you can get it to produce a raster image of any DPI but it still contains a low detail map. I think Xara is a hybrid of vector and raster. I dont think those trees and base terrain is all vector shapes. I too use a different hybrid system for my maps which I think is the best of both worlds.

Its not a good answer but its a problem that is very hard to solve in a meaningful way. I wrote my terrain erosion simulator and programmable texture engine so that I could try to solve this problem. We used this on MeDem so that we only had to describe middle earth in so much detail and then let the simulation take it up a few notches.

[Hikki-Jin]

Well, the consensus seems to be that I should look into using a vector-based program alongside Photoshop.

I believe there is a problem with you numbers, unless it's intentional.
In A, the size of each pixel is 0,111... mi2
In B, each pixel is 0,102...mi2

Oops. I made a typo in my post. B shows 24,454mi², not 22,454mi². Haha, sorry about that.

How big do you want B to be in pixels x pixels?

Hmm, I don't quite know yet... Probably similar to A, though keeping its own aspect ratio, of course... also, preferably something that would allow me to use clean, whole numbers for the scale (i.e. 64px = 1mi², as opposed to 64.xxxxxx), if that's even possible (math is not my strong suit).

that is why on TV when you see the " image enhancement "( HA-HA-HA-HA) they start with a GOOD sharp image than make it look bad and say the bad one was the original ( the TV lie )

Hah! Oh, yes. The good ol' magic enhance button. Especially amusing are the scenes where they use some manner of witchcraft to rotate the image, revealing more information.

As an aside, that 5th map posted below, is my fictional Japanese archipelago analog setting, which I've already published - that map will be used in my upcoming Gamemaster's Guide to Kaidan due to be printed within the year. It is for my Kaidan setting of Japanese Horror (PFRPG).

I've not played Pathfinder, but I've seen your Kaidan map. It's pretty cool (as are the other maps you posted).