Editing Talk:Graphic Image Map (GIM)

=Strings from GimConv EXE and DLL=

'''image_format'''
 RGBA8888....RGBA4444....RGBA5551....RGBA5650
 INDEX32.INDEX16.INDEX8..INDEX4
 DXT5EXT.DXT3EXT.DXT1EXT.DXT5....DXT3....DXT1

'''pixel_order'''
 NORMAL
 PSPIMAGE
 FASTER
 GENERIC

'''level_type'''
 MIPMAP
 MIPMAP2
*MIPMAP3, MIPMAP4, MIPMAP5, etc... doesnt works (not valid)

'''Some hints about hierarchy''' at the most deeper levels
 IMAGE_INDEX
 IMAGE_PLANE
 IMAGE_LEVEL
 IMAGE_FRAME

 PALETTE_INDEX
 PALETTE_LEVEL
 PALETTE_FRAME

'''frame_type'''
 SEQUENCE

'''???'''
 HOLD
 CYCLE
 DISSOLVE
 EVENT
 CONSTANT
 LINEAR
*'''frame_type''' with this values doesnt works


=Frankensperiments=

==Block header size==
This was made with the goal of identifying the last value at the block headers, the file is well formed but the value in red has been modifyed, originally was a 0x10 and has been changed to 0x20

 Offset(h) 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F
 
 00000000  2E 47 49 4D 31 2E 30 30 00 50 53 50 00 00 00 00  .GIM1.00.PSP....
 00000010  00 02 00 00 00 00 00 E0 00 00 00 10 00 00 00 10  .......à........
 00000020  00 03 00 00 00 00 00 80 00 00 00 10 00 00 00 10  .......€........
 00000030  00 04 00 00 00 00 00 70 00 00 00 70 00 00 00 10  .......p...p....
 00000040  00 30 00 00 00 03 00 00 00 04 00 02 00 20 00 10  .0........... ..
 00000050  00 01 00 02 00 00 00 00 00 00 00 3C 00 00 00 40  ...........<...@
 00000060  00 00 00 60 00 00 00 00 00 01 00 01 00 03 00 01  ...`............
 00000070  55 53 45 52 44 41 54 41 41 52 45 41 00 00 00 40  USERDATAAREA...@
 00000080  CA 5E 11 00 CA 5E 12 00 CA 5E 13 00 CA 5E 14 00  Ê^..Ê^..Ê^..Ê^..
 00000090  CA 5E 21 00 CA 5E 22 00 CA 5E 23 00 CA 5E 24 00  Ê^!.Ê^".Ê^#.Ê^$.
 000000A0  00 FF 00 00 00 00 00 50 00 00 00 50 <span style="background:#ff0000">00 00 00 20</span>  .ÿ.....P...P... 
 000000B0  4E 4F 50 41 44 44 2E 42 4D 50 00 41 64 6D 69 6E  NOPADD.BMP.Admin
 000000C0  69 73 74 72 61 74 6F 72 00 54 68 75 20 41 70 72  istrator.Thu Apr
 000000D0  20 31 39 20 30 39 3A 32 32 3A 35 31 20 32 30 31   19 09:22:51 201
 000000E0  38 00 47 69 6D 43 6F 6E 76 20 31 2E 32 30 68 00  8.GimConv 1.20h.

At the time the area at 0xA0 is loaded (block id 0x00FF intended to store the '''file_info''') the real header size is 0x10 but the value changed takes the next 0x10 bytes as part of the header... so it starts reading the '''file_info''' at offset 0xC0

This is how it looks when this file is converted GIM to GIS, note the strings at the bottom how has been cropped
 .GIS 1.00
 
 Picture "picture-0" {
 	Image "image-0" {
 		Format RGBA8888
 		Order NORMAL
 		Width 4
 		Height 2
 		PitchAlign 16
 		HeightAlign 1
 		PlaneMask 0x00000000
 		LevelType MIPMAP
 		LevelCount 1
 		FrameType SEQUENCE
 		FrameCount 1
 		UserData "userdata-0" {
 			0x52455355 0x41544144 0x41455241
 		}
 		Pixels "pixels-0" RGBA8888 NORMAL 4 2 16 1 {
 			0x00115ECA 0x00125ECA 0x00135ECA 0x00145ECA
 			0x00215ECA 0x00225ECA 0x00235ECA 0x00245ECA
 		}
 	}
 }
 
 FileInfo "file_info-0" {
 	ProjectName "istrator"
 	UserName "Thu Apr 19 09:22:51 2018"
 	SavedDate "GimConv 1.20h"
 	Originator "GimConv 1.20h"
 }

Also, the length of the value marked in red is confirmed to be 4 bytes, it can be seen how the bytes are swapped when using the setting '''format_endian''' = little/big

=Links=
*https://en.wikipedia.org/wiki/PSGL
*https://research.ncl.ac.uk/game/mastersdegree/workshops/ps3introductiontogcm/tutorial6.pdf
*https://research.ncl.ac.uk/game/mastersdegree/workshops/ps3introductiontogcm/GCM%20Texturing.pdf (page 2 about pitch/height memory alignments)