icon_map3.dds X,Y origin

tknarr · #1 06-16-2007, 11:09 PM

I've been trying to create a set of icons for the icon_map3 file. I've got it 95% there, but I need a quick check of something.

First, is the origin used in the skinnable_Skinnnable_Styles.xml file for the icons based at 0,0 or 1,1? That is, is 1,1 the upper-leftmost pixel in the .dds image, or is it one pixel down and to the right? Secondly, am I right in calculating that the image chunks extracted are 41 pixels by 41 pixels square with a 1-pixel gutter between boxes that isn't used?

I ask because I'd laid down a 1-pixel white grid to seperate boxes so I could see how centered I had icons, and when I left that grid in the final .dds file I ended up with 2 edges just barely visible in the displayed icons on the map. Changing the alpha channel so the grid was transparent cleared that up, but I'm wondering if I got my origin off so my boxes are 1 pixel off from where I think they are.

Thanks.

**Landiin** · #2 06-18-2007, 11:11 AM

0,0 is the upper-leftmost pixel. I don't remember the size of the rect I used for the image chunk but they are all the same.

I didn't create the image I just added the you are here cross hair image so I can't tell you how the person that made this image in the 1st place layed it out. I just stuck the image the next spot.

**gm9** · #3 06-18-2007, 02:01 PM

Doesn't the icon_map3.dds that comes with EQ2Map already contain a grid? At least the one I have in ProfitUI does and it is based on an old official one, so it's probably correct.

The image file uses 42x42 for the icons including the 1px border, leaving a net space of 40x40 for the actual icon. Whether the XML references that correctly I don't know since I don't have it at hand.

tknarr · #4 06-18-2007, 05:45 PM

Some had a grid, some didn't. I was trying to work out what exactly the rules were. I ended up with what looked like a 1-pixel error, but I wasn't sure whether I was 1 pixel off on size or origin. Normally I'd trust the XML that comes with EQ2Map, but I found things like positions apparently being reversed (the second pair was less than the first) and one box with a different size than any of the others of the same type.

Time, then, to experiment. Make myself a nice obvious white-black-white border on the boxes, throw it in and see what gets clipped where.

Thanks for the responses.