PDA

View Full Version : Map making



Makuzi
October 24th, 2010, 03:08
Question about making a map... I have the pictures from the PDF file from WOTC to for the maps in "Sleeper in the Tomb of Dreams" However the grid thats on them seems to be real small and I cant get it to blow up to make it 50x50 grid size.

Im using Paint.net to crop and fix anything in the maps. Does anyone have a tip I can use to make this work for me? I just dont like the way the maps are looking after I get them into FG2 and place the grid on it.. Either their off just a bit or the grids are so small you can really see the tokens.


Thanks so much for the help.

Makuzi

Fot5
October 24th, 2010, 15:25
I'm not sure whether you mean that the problem is grid alignment or image quality. If it's grid alignment, you can just run the encounter without the grid on. That's what I do with maps that already have grids on them. You just lose the distance calculation and square-snapping functionality. If it's image quality, you can play with the resolution of the saved image.

Makuzi
October 24th, 2010, 16:40
It was getting it all to line up.... I actually found the video that shows how to do it and am slowly getting it down.

Thanks for the reply

Griogre
October 24th, 2010, 17:17
Bear in mind some processed maps will *never* have the grid line up. If a map is re-sized enough the process of interpolation causes pixel drift so the grid squares are no longer square. Like Fot5, if a map image has a grid on it most of the time, I not using the FG grid is just easier.

Fot5
October 24th, 2010, 23:54
I've noticed that many of the WOTC maps have inconsistent gridding (i.e., they change spacing from one line to the next). These are the some of the types of maps that you're never going to get to align correctly. If you have a map that lays a consistent grid across it, I've found it a lot easier to match the FG and map grid by making the map a lot larger when I lay the FG grid on top.

Zeus
October 25th, 2010, 00:34
I've noticed that many of the WOTC maps have inconsistent gridding (i.e., they change spacing from one line to the next). These are the some of the types of maps that you're never going to get to align correctly. If you have a map that lays a consistent grid across it, I've found it a lot easier to match the FG and map grid by making the map a lot larger when I lay the FG grid on top.


Thats because Mike Schley (one of Wizard's main cartographists) often uses a beveled edge to his grids and as Griogre has stated the resizing of the map causes the grid lines to shift off-centre.

EugeneZ
October 25th, 2010, 17:28
Listen, guys. I'm not doubting what you're saying because you have no reason to lie and are smart enough to know better than me about maps. :)

But... I've normalized over a HUNDRED maps over the course of playing Scales of War to Level 28 (we finish in Jan!) and War of the Burning Sky to Level 16. Many of the Scales maps were done by Schley, if I'm not mistaken (in fact, after Dungeon started to heavily use Dungeon Tiles, Scales adeventures were one of the only adventures to still use handmade maps). And I have yet to see a single map where the grid didn't align perfectly. Watch the screencast below for an example of a quick, perfectly-fitting Schley map.

That said, I don't use Xorn's technique, it takes way too much time to get it right. The following works with a 99%+ accuracy every time. I've made a screencast of my technique here:

https://screencast.com/t/6AYPRIXFdPyL

But if you need it in text, here's a copy from my WotBS-related post from EnWorld:

Step 1) Use any software with a ruler to measure an arbitrary number of squares in either direction. The more the better, but they must be full squares. So, for example, measure the exact number of pixels from the bottom of the lowest possible square to the top of the highest possible square. Ignore squares that are cut off. Only count pixels in one dimension, either x or y.

Step 2) Count the number of physical squares that you just measured the length of. At this point, you should have something like this: 943 pixels = 35 squares. This number gives you the current "zoom level" of the map. Divide the number of pixels by the number of squares to give you the zoom level: 943/35 ~= 26.943. The more decimal places you keep, the more accurate the final result is, so don't just round to the nearest integer.

Step 3) Using the same dimension you used for the length count, get the image's TOTAL length in that dimension. In the example above, I measured the y (top-to-bottom) dimension of most of the squares on the map, but I didn't count a few obscured squares at the top because I couldn't tell where one square ended and the next began. The bottom-most square was also not counted because it was cut off. So the height of this image was 1200.

Step 4) Divide the total height by the zoom level you obtained earlier. This gives you the total number of squares the image REALLY has. In my example, I would do: 1200/26.943 = 44.538. My image has 44 and a half real squares.

Step 5) Decide what zoom level you want to normalize the map to. I always use 50px. Multiply the result of Step 4 by this number. So 44.538 x 50 = 2226.9235.

Step 6) Resize the image. Make sure your editor's "Maintain aspect ratio" option is enabled. Change the value of whatever dimension you were using to the result of step 5. I was using y, so I'm changing the height of the image to 2227. This time you have to round to the nearest integer. Hit OK, and the software will resize the image. Every square will be exactly 50px, or close enough that maptools won't be able to tell the difference.

Fot5
October 25th, 2010, 17:40
Nice explanation. I haven't tried to normalize my map images, so I'll try to do that to see whether your process works with the maps that I've been having trouble with. Thanks.

Zeus
October 25th, 2010, 18:21
Oops maybe I started something unintended.

To be clear I'm not slating Mike Schley's maps or his grid style, in fact I like his style so much I have painstakingly developed my own repertoire of art skills and assets for recreating his map styles for my own 4E campaigns.

However, the point I was driving at is that because he uses a beveled edge (more notable in the Tile based versions of his maps) and because you generally want to increase the resolution of WotC maps for 50px play, the process of the enlargement can often throw one or more of the grid horizontal/vertical lines off. I believe this is what Griogre is refereing too when he quotes interpolation.

This adverse effect is generally not as evident for smaller tactical maps as per the SoW and WotBS published material (and your example) but is very evident when you try follow the same process for larger sized maps e.g. the large Level 1 map in KotS.

So in summary, whilst I agree your process does work, at some point despite following this you will encounter a larger WotC published map that will NOT align in FGII 100% properly.

When I get some time I'll post up some visual examples.

EugeneZ
October 25th, 2010, 23:28
So in summary, whilst I agree your process does work, at some point despite following this you will encounter a larger WotC published map that will NOT align in FGII 100% properly.

I will take that as a challenge! Please present me with such a map. :) I've converted many huge dungeon maps from SoW and other adventures with no problems. In fact, the bigger the map, the better the result because there are more pixels, leading to less error.

My method seems mathematically sound. I compute the number of "virtual" squares between the top and bottom squares on the map. So even if the squares are slightly "off" or something along those lines, the resulting resize will fix them -- there are N number of squares on the map, and my method figures out exactly how many. I don't see much room for error, but it's always a possibility. Show me an example and either you or I will learn something, and the community will be better for it. :)

Of course, it's also possible you're talking about subpixel differences which are only noticable when you're zoomed to 1x or greater... and even then, only if you're looking... if that's the problem you're referring to then I personally wouldn't worry about it. :p

Makuzi
October 26th, 2010, 15:05
Wow with all the replies I hope I can get this to work right. I found the link to the FG2 tutorial on map making, however my grids still dont match up perfectly. I had just decided they would be off.. But Im now going to try Eugenz tips for making the map and see if I can figure this out and make it work.

Thanks again

Griogre
October 28th, 2010, 04:08
EugeneZ, as you noted, rescaling is just simple algebra. If you count a large number of squares and note the total number of pixels in the squares then divide the total number pixels by the number of squares you get the average size of each square in pixels (your zoom level in your explanation). Once you have that you can simply solve to adjust the grid to any size IE Y pixels = (AverageGridSize) X - set Y, solve for X and multiply that times the number of squares to get the scaling factor.

The key to why this doesn't always work in the real world is that we are talking the *average* grid size and there is an unspoken assumption that the grid we are converting has 90 degree angles at each corner and each line connecting the corners is straight - the grid is "true". If the grid on the map is not "true", the FG grid won't ever perfectly match the grid on the map since the FG grid *is* true.

Before getting into when scaling isn't going to work, it is good to emphasize when the theory *is* going to work - when you have a good digital image in a lossless format and you resize down rather than up. That's not an uncommon occurrence in the digital age.

So what causes a grid not to true? Often a non-digital representation of the original which has be converted to digital does not have a grid that is "true." A scanned image is a good example of this for two reasons: 1) The scanned image is lined up such that the grid is *not* perfectly horizontal and vertical in the scanned image. 2) The image being scanned was not pressed equally tight against the glass through out the image so you have curves at one end of the image (typically this happens in a by the spine of a book).

You can also run into the same problem with digital maps. As DrZuess mentions if you up-size a map from say a 30 pixel grid to a 50 pixel grid even if you start with a true grid on you original, your art program uses an algorithm to generate the extra pixels you told it to make and resulting image's grid is probably not a perfectly straight line though it likely is sorta close.

The interpolation issue is further magnified if the image is in a lossy format like a jpg where the original was first resized for a PDF, and then you pull it out of the pdf and resize it again.

AnAxeToGrind
October 30th, 2010, 18:19
I have a different but related issue. Due to the color formatting of some maps, the FGII grid sometimes vanishes into the flooring matching to closely the color scheme of the entire map. I have found that tweaking the color scheme or contrast of the map can make the grid show.

Now for my question: Is anyone aware if there is any way thru the program or an extension to adjust the grid color without going thru the work of trying to adjust the color scheme on the original map?

Griogre
October 30th, 2010, 23:27
It's on the wish list to have a different colored grid. I wish they would just allow you to switch the grid from black to white like the dice numbers which would solve most visibility problems. Until then the best you can do is not make/use maps with dark backgrounds. Sometimes you don't have much choice though.

Makuzi
October 31st, 2010, 14:56
Thanks all for the help.... I figured out why I was having such a hard time making my maps grids line up with using both ways shown.

The map's I copied from the PDF's had a boarder around them with no grid. So when I was resizing it using both techniques it would come out wrong. I then used a snipping tool to crop all the excess boarder off and make the grid fill the entire image.

Needless to say that fixed my problem and now the grids match damn near 100% using Euginz way of doing it..

Thanks again

EugeneZ
November 2nd, 2010, 09:59
Ah, a satisfied customer! Glad it's working Makuzi.


You can also run into the same problem with digital maps. As DrZuess mentions if you up-size a map from say a 30 pixel grid to a 50 pixel grid even if you start with a true grid on you original, your art program uses an algorithm to generate the extra pixels you told it to make and resulting image's grid is probably not a perfectly straight line though it likely is sorta close.

The interpolation issue is further magnified if the image is in a lossy format like a jpg where the original was first resized for a PDF, and then you pull it out of the pdf and resize it again.

Griogre, this is essentially the only part of your post I disagree with. Everything you were saying was solid up until this. You use the word "probably" in your second sentence there, and the ambiguity in that makes me wonder. You clearly understand the math behind resizing a map and putting a grid on top of it. So in the the same sense, you should see that that when you have a map with a true grid, as you call it, and you shrink it down, the end-users (us) have an EASIER time placing the grid perfectly than before. Why? If given the original map, there is only a single position of the grid in FG2 that will be "perfect." Directly on top of the original grid. But when shrunk and then resized back up with a decent sampling algorithm, the original 1px grid has been reconstructed to be larger than that 1px -- making it easier to place your grid in FG2, which in turn means there's less room for error when resizing, as well.

One could argue, and I assume this is what you mean, that once the image has been resampled to create a larger image, the grid will *never* fit as perfectly (for mathemtical, not visual, definitions of perfect) into the original location as it did when the grid was 1px and unsampled. If this is what you mean, then why bring it up at all? It's hard for me to believe that anyone could notice a problem with FG2's grid lines matching up against a blurry gridline in the image. Again, as long as the resizing is done mathematically, the grid is off by LESS than 1px. Which would normally be very ugly if it wasn't for my paragraph above, which explains why it's perfectly reasonable to have a grid positioned in any of a number of pixels, once you've increased the size of the image.

In other words, I still see no meaningful downside to resizing digital images, assuming the source at one point had the "true" grids you're talking about. If not, then the rest of your post applies nicely.

Teonis
November 7th, 2010, 19:06
Wow, some of these tips are amazing. It really blows my technique out of the water. What I was doing was using photoshop. Open the map and make a second layer with a 50x50p square, then scale and transform to make the 50p tile fit in the grid square.

I however, am currently working on SoW: Den of the destroyer and I noticed that the grid squares, even on the original, were not squares. (they were like 26x27) So scaling them alone was not enough. I had to distort some as well.

Math is win. lol

Makuzi
December 3rd, 2010, 21:35
Would there be a way to have a sticky of EugeneZ's tips for making a map??? I found this one to work the best for myself and keep coming back to it, and thought others might find it useful as well.

:)