I think I cited this measure before...
Quote:
Originally Posted by
Blood_Hawk23
I've found that it depends on what its for. AC DXT1 works for alpha channeled parts. Where DXT3 can cause issues with shadows. Mostly holes in the shadows. All other parts the Extended Bit 565 works best.
Same for ground objects. DXT1 works best. DXT3 might not show compounded objects. Other objects the ACs shadow might not show if you have a large alpha channel on the object. I discovered this while doing the euro buildings and trees.
Now this is on my computer. Others might not have the same issue.
I noticed this a long time ago... the increasing deterioration of extended bmp bitmaps with successive "edits" of a texture... so called "moireau pattern" effect on color shades and loss of finer detail... The way I got this over... was to:
1. make a PSP or PSD copy template of the intact texture... in high-def colors and do the editing on that...
2. then replace the aircraft texture with an extended bmp copy via Wright's program.
3. Then, every time I need to add or change the texture... I open the template... change it in PSP or Photoshop an replace the new version on the aircraft texture folder... in place of the one that needed editing...
With this method, there is no an iota of change in color or shape fidelity... I NEVER OPEN, EDIT AND REPLACE THE OLD ONE! Hope this is of help...
The one thing that keeps me puzzled is that CFS 2 aircraft textures as well as FS9 - highly detailed originals - when edited never show again (at least in my computers -different ones as I change to newer and better rigs) unless they are put on 256 colors... I mean... the original is 8 bit color, and not extended, but its edited copy does not show if put back after edit... the plane shows blank textures just plain gray... then when I reduce the color ... it shows... but not as nicely as the high-def edit....
so... I started the method described above... and replace them as extended bmp files... and this resolves the problem for me... I think I basically do what Kelticheart describes here...
G.