Anything that will save me time without sacrificing quality piques my interest. But I'm not clear on the advantages of this over the grid method I used to use. Or how it requires less clean-up than BMP2BIC. I'd like to hear more about the details of the process. The base image must require some prep work before rescaling. Can just imagine the mess things like place names & dark outlines on source maps could present. On the reduced scale how do rivers show?
I'll preface this by saying neither option is wrong or right, just a matter of preference and skill (or in my case, a lack thereof).
I used BMP2BIC many years ago, when it was new and it's original thread was active. I attempted to make quite a few maps with it, some successful, many not, some were pure ocean. Some were just plain silly. I am by no means a Graphic Artist. Until very recently my work has always consisted of Cut N Paste, or pixel-by-pixel work, or later a combination of both (fancy!) - I never used palettes or shading or models or anything of the sort. It is only with my recent work (past 9 months or so) that I have done any work at all with pallettes on leaderheads for my Trinity of mods. Considering I've been modding Civ3 since the day it was released this is surprising even to me.
Point being, without knowing how to work a palette at all at the time, a direct image resizing and transmogrification through BMP2BIC almost never came out how I wanted, so what I started doing was using the colors in the palette image that comes with the utility and painting over every pixel, one at a time, so that when B2B processed the image what came out was what I wanted, sans volcanoes and marshes and all the overlays. I was always a little frustrated by the cookie cutter sizes I was restricted to, but that was tolerable.
Then It occurred to me that if I am having to repaint the pixels on the image, why not just use the image pixels to determine the placement of the tile on the map, that way I am only painting the pixels\tiles once, without having to circle back and fill any gaps and missing terrain.
So this saves time I guess only if you have a severe lack of graphics knowledge
I should say that it also makes it very clear what you have done and what is left to do. Using the B2B method, even if done correctly, you have a chance of missing some tile corrections, whereas with my pixel-to-tile method, going 1 column at a time, top to bottom, I know exactly where I left off.
For me with either method, once the base map is done, I visually compare it with the original, larger image, and adjust accordingly. Writing on the map I usually correct by replacing it with terrain, based on the surrounding. Rivers I mostly place manually after, though with maps that have large rivers, some or all of the river may end up being Coast tiles.
This method is how I've done almost all of my from-scratch maps, including Orange County and Hawaii 362*362. For those I used an image I captured from Google Earth, with SCCOOS Bathymetry overlay. I also used GE to accurately place the freeway system and city placement in OC.
I try to find images without thick borders or writing, but I know in one map I did of the earth after a massive raise in water elevation due to Global Warming, the only good map out there had thick black outlines on not only the new coast but the old. That image was like 4096*4096 or something equally daunting, and I started to go through through it all, pixel by pixel, to remove the borders and writing on the map so the shrunk version would be clean, but I gave about a third of the way through as that was going to take me eons.
Long story short, I prefer this for the following:
- Finish base terrain in a single pass
- No cleanup
- More control over the map size
- Easy to know where you left off
- I am a stubborn old goat
