LazPaint LZP Format: what changed

Old	New		Line
1	1		## LZP File format specification
2	2
3	3		In this document, [[Hexadecimal\|hexadecimal]] numbers follow Pascal convention, i.e. they are prefixed with a [[Dollar sign\|dollar sign]]. For example \$10 means 16. As an exception, sequences of bytes are not prefixed.
4	4
5	5		The default encoding for numbers is little endian. So the hexadecimal number \$12345678 would be stored as the sequence of bytes 78 56 34 12.
6	6
7	7		Numbers are 32-bit signed integers, unless stated otherwise. So -1 is stored like \$FFFFFFFF.
8	8
9	9		### Global structure
10	10
11	11		The file is composed of:
12	12
13	13		`Offset Content`\
14	14		`------ --------------------`\
15	15		`0 File header`\
16	16		`? Thumbnail (optional)`\
17	17		`? `[[LazPaint LZP Format Single Image\|Flat image]]\
18	18		`? Layer structure (optional)`
19	19
20	20		Note: some old formats do not have a header and start directly with the flat image. Image readers may discard those files. However for your information, in this case, images are always stored with ZStream compression and the size of the image is deduced from the size of the flat image.
21	21
22	22		### File header
23	23
24	24		`Offset Size Name Content`\
25	25		`------ ---- ------------------ ------------------------`\
26	26		`0 8 ``magic`` Always contains the string 'LazPaint'`\
27	27		`8 4 ``zero1`` Contains the number 0`\
28	28		`12 4 ``headerSize`` Size of this header`\
29		-	\
	29	+
30	30		`16 4 ``width`` Width of the image canvas`\
31	31		`20 4 ``height`` Height of the image canvas`\
32	32		`24 4 ``nbLayers`` Number of layers`\
33	33		`28 4 ``previewOffset`` Offset (from the beginning of the file) of the `[[LazPaint LZP Format Single Image\|flattened image]]\
34		-	\
	34	+
35	35		`32 4 ``zero2`` Contains the number 0`\
36	36		`36 4 ``compressionMode`` Contains flags indicating the compression and the presence of a thumbnail`\
37	37		`40 4 ``reserved1`` Reserved for future use. Set to 0.`\
38	38		`44 4 ``layersOffset`` Offset (from the beginning of the file) of the layer structure`
39	39
40	40		headerSize contains the size of this header. An image reader must read this to know which fields are available. It must be greater or equal to 48.
41	41
42	42		width and height are the dimension of the canvas. There is no real limitation for these numbers however it is recommended to keep it under or equal to 8192.
43	43
44	44		nbLayers is the number of layers. The minimum is 1. There is no real limitation however it is recommended to keep it under or equal to 99.
45	45
46	46		layersOffset can be 0, in which case the image is flat and that there is no layer description.
47	47
48	48		#### Compression
49	49
50	50		compressionMode contains flags indicating the compression and the presence of a thumbnail.
51	51
52	52		The low order byte indicates the compression of the flattened image. It is the value (compressionMode and \$ff):
53	53
54	54		\- \$01 indicates ZStream compression
55	55
56	56		\- \$02 indicates RLE compression
57	57
58	58		\- any other value should be considered as an indication that the image reader cannot interpret the content of the file
59	59
60	60		The following flags can be defined:
61	61
62	62		\- \$100 indicates that there is thumbnail in PNG format; check with (compressionMode and \$100) \<\> 0
63	63
64	64		If there is a thumbnail, it is directly after the header. The reader should read headerSize to have the correct location.
65	65
66	66		The compression of the layers is defined in the layer structure.
67	67
68	68		#### Thumbnail
69	69
70	70		Its presence is recommended however optional. Its presence is specified by the flag \$100 of compressionMode.
71	71
72	72		Currently, it is a PNG file embedded directly after the header. It is provided as a way to facilitate quick preview.
73	73
74	74		PNG format is widespread so it is easy to implement a browser that would display the thumbnail. To do so, check that the header contains magic and that the flag for the thumbnail is set in compressionMode. Then jump to the end of the header using headerSize. Then read the PNG file.
75	75
76	76		There is no real limitation but the recommended size for the thumbnail is 128 along the largest dimension. The other dimension may be smaller if the image is not a square. Of course if the image is smaller than 128 then the thumbnail will be smaller than 128.
77	77
78	78		### Flat image
79	79
80	80		The file always include a flattened image to make it easier for image viewers. It is in the [[LazPaint LZP Format Single Image\|LZP single image format]].
81	81
82	82		If there is no layer structure, the caption of this image is the name of the unique layer. If there is a layer structure, then the caption of this image is discarded.
83	83
84	84		### Layer structure
85	85
86	86		`Offset Size Content`\
87	87		`------ ---- -------------------------------------------------------`\
88	88		`0 20 Contains the string 'TBGRALayeredBitmap' followed by the bytes 1A 00`\
89	89		`20 4 Size S of the additional header after this field (must be at least 12)`\
90	90		`24 4 Number of layers N (should be identical to the number in the file header)`\
91	91		`28 4 Selected layer index. Equals to -1 if no layer is selected, otherwise positive from 0 to N-1`\
92	92		`32 4 Stack flags`\
93	93		`36 4 Canvas width (optional)`\
94	94		`40 4 Canvas height (optional)`\
95	95		`44 8 64-bit offset to the directory structure relative to the start of this header (optional)`\
96	96		`24+S First layer (bottom one)`
97	97
98	98		The following flags are defined:
99	99
100	100		- \$01: the layers are blended without gamma correction
101	101		- \$02: the layers are compressed with RLE compression (otherwise they are compressed with ZStream compression)
102	102
103	103		If the canvas size is not supplied (when the total header size is less than 44), then the canvas size is the size of the first non-empty layer bitmap.
104	104
105	105		Similarly, the [[LazPaint LZP directory structure\|directory structure]] offset may not to be supplied or set to zero. In this case, there is none. It is optional and only necessary for vector shapes.
106	106
107	107		The total header size of the layer structure is equal to 24 + S. Then for each layer there is a layer entry.
108	108
109	109		The first layer entry is the bottom most layer. Then each layer is blended over the previous ones. The last layer is on top and is not hidden by any other layer.
110	110
111	111		#### Layer entry
112	112
113	113		Header of the layer entry:
114	114
115	115		`Offset Size Content`\
116	116		`------ ---- -------------------------------------------------------`\
117	117		`0 4 Size S1 of the optional fields that follows`\
118	118		`4 4 Layer flags (lowest bit means that it is visible)`\
119	119		`8 4 Layer blend operation`\
120	120		`12 4 Layer horizontal offset in pixels`\
121	121		`16 4 Layer vertical offset in pixels`\
122	122		`20 4 Layer ID (if used, each layer must have a different ID in a given file)`\
123	123		`24 4 Layer opacity from 0 (transparent) to 65535 (opaque)`\
124	124		`28 4 Lower bits of size S2 of image data`\
125	125		`32 4 Upper bits of size S2 (optional, forming 64-bit integer)`\
126	126		`36 16 GUID of the original used to render the layer (optional)`\
127	127		`52 24 Affine matrix of the transform applied to the original`\
128	128		`4+S1 S2 `[[LazPaint LZP Format Single Image\|Image data]]\
129	129		`4+S1+S2 Next entry`
130	130
131	131		All fields are optional so in theory, S1 can be set to zero and the image data can directly follow. However it is recommended to write the whole header.
132	132
133	133		The default values are: visible, blend operation boTransparent, offset (0,0), undefined ID (maximum integer value), opaque, size S2 = 0, GUID = NULL. If an original is specified the affine matrix transform is mandatory.
134	134
135	135		The size of the image data is a 64-bit integer, though the upper value is optional and by default zero. When the image size S2 \<\> 0, the offset to the next entry is computed using the formula 4+S1+S2. Otherwise, the data is expected to continue after the end of the image data.
136	136
137	137		The GUID is stored here as binary (in little endian): one DWord, two Words and 8 bytes.
138	138
139	139		The affine transform is stored as binary (in little endian): 6 single precision floats ux, uy, vx, vy, tx, ty (axes u and v and translation t).
140	140
141	141		#### Blend operations
142	142
143	143		- Under is the value of the channels of the layer underneath
144	144		- Over is the value of the channels of the layer above
145	145		- Mix is an intermediate value that is then blended over the previous layers, and that, unless specified, is computed for each RGB channel. The alpha channel is computed differently (see below table).
146	146		- Final is the resulting value
147	147
148	148		Note:
149	149
150	150		- values (Red, Green, Blue, Alpha, Mix, Under, Over, Final) are considered to be between 0 and 255
151	151		- GE means gamma expansion. GE(x) = Power(x/255, 2.1)*65535
152	152		- GC means gamma compression. GC(x) = Power(x/65535, 1/2.1)*255
153	153		- ByteSqrt means the square root. ByteSqrt(x) = Power(X/255, 0.5)*255
154	154		- div stands for integer division (rounded down)
155	155		- HSL is the standard HSL colorspace, without gamma correction
156	156		- GSB is the corrected HSL colorspace of LazPaint. It is computed with gamma correction and the B (brightness) component correspond to the perceived lightness:
157	157
158	158		`B = (306GE(Red) + 601GE(Green) + 117*GE(Blue))/1024`\
159	159		`B between 0 and 65535`
160	160
161	161		Here is the list of operations supported in LazPaint:
162	162
163	163		`Value Name Description`\
164	164		`----- ----------------- ----------------------------------------------------`\
165	165		`0 LinearBlend Blend over the previous layers without gamma correction`\
166	166		`1 Transparent Blend over the previous layers `\
167	167		` (with gamma correction unless it is specified otherwise in the stack flags)`\
168		-	\
	168	+
169	169		`2 Lighten Mix := Max(Over, Under)`\
170	170		`3 Screen Mix := 255 - ((255 - Under) * (255 - Over) ``shr`` 8)`\
171	171		`4 Additive Mix := GC(GE(Under)+GE(Over))`\
172	172		`5 LinearAdd Mix := Under + Over`\
173	173		`6 ColorDodge Mix := ``if`` Over = 255 ``then`` 255`\
174	174		` ``else`` Min(255, (Under ``shl`` 8) ``div`` (255 - Over))`\
175	175		`7 Divide Mix := ``if`` Over = 0 ``then`` 255`\
176	176		` ``else`` Min(255, (Under ``shl`` 8) ``div`` Over)`\
177	177		`8 NiceGlow Mix := ``if`` Under = 255 ``then`` 255`\
178	178		` ``else`` Min(255, OverOver ``div`*` (255 - Under))`\
179	179		`9 SoftLight Mix := ((255 - Under)Over shr 7 + Under)Under div 255`\
180	180		`10 HardLight Mix := ``if`` Over <= 128 ``then`` UnderOver ``shr`*` 7`\
181	181		` ``else`` 255 - ((255 - Under)(255 - Over) ``shr`*` 7)`\
182		-	\
	182	+
183	183		`11 Glow Mix := ``if`` Under = 255 ``then`` 255`\
184	184		` ``else`` Min(255, OverOver ``div`*` (255 - Under))`\
185	185		`12 Reflect Mix := ``if`` Over = 255 ``then`` 255`\
186	186		` ``else`` Min(255, UnderUnder ``div`*` (255 - Over))`\
187	187		`13 Overlay Mix := ``if`` Under <= 128 ``then`` UnderOver ``shr`*` 7`\
188	188		` ``else`` 255 - ((255 - Under)(255 - Over) ``shr`*` 7)`\
189	189		`14 DarkOverlay Mix := ``if`` GE(Under) <= 32768 ``then`` UnderOver ``shr`*` 7`\
190	190		` ``else`` GC(65535 - ((65535 - GE(Under))(65535 - GE(Over)) ``shr`*` 15)`\
191	191		`15 Darken Mix := Min(Under, Over)`\
192	192		`16 Multiply Mix := ``if`` Over >= 128 ``then`` Under(Over+1) ``shr`` 8 ``else`` UnderOver ``shr`` 8`\
193	193		`17 ColorBurn Mix := ``if`` Over = 0 ``then`` 0`\
194	194		` ``else`` Max(0, 255 - (Under ``shl`` 8) ``div`` Over)`\
195		-	\
	195	+
196	196		`18 Difference Mix := GC(Abs(GE(Under) - GE(Over)))`\
197	197		`19 LinearDifference Mix := Abs(Under - Over)`\
198	198		`20 Exclusion Mix := GC(GE(Under) + GE(Over) - (GE(Under)GE(Over) ``shr`*` 15)`\
199	199		`21 LinearExclusion Mix := Under + Over - (UnderOver ``shr`*` 7)`\
200	200		`22 Subtract Mix := GC(Max(0, GE(Under) - GE(Over)))`\
201	201		`23 LinearSubtract Mix := Max(0, Under - Over)`\
202	202		`24 SubtractInverse Mix := GC(Max(0, GE(Under) - (65535 - GE(Over))))`\
203	203		`25 LinearSubtractInv Mix := Max(0, Under - (255 - Over))`\
204		-	\
	204	+
205	205		`26 Negation Mix := ``if`` GE(Under) + GE(Over) <= 65535 ``then`` GC(GE(Under) + GE(Over))`\
206	206		` ``else`` GC(131071 - (GE(Under) + GE(Over)))`\
207	207		`27 LinearNegation Mix := ``if`` Under + Over <= 255 ``then`` Under + Over`\
208	208		` ``else`` 511 - (Under + Over)`\
209	209		`28 Xor Mix := Under ``xor`` Over`\
210		-	\
	210	+
211	211		`29 SvgSoftLight Mix := ``if`` Over <= 128 ``then`` Under - (((256 - 2Over) Under ``shr`` 8)(255 - Under) ``shr`*` 8)`\
212	212		` ``else`` ``if`` Under <= 64 ``then`` Under + ((2Over - 256) (Under7 - (4Under(4Under + 256)(256 - Under) ``shr`` 16)) ``shr`*` 8)`\
213	213		` ``else`` Under + ((2Over - 255) (ByteSqrt(Under)-Under) ``shr`` 8)`\
214	214		`30 Mask ApplyOpacity := ((2Over.Red + 4Over.Green + Over.Blue)Over.Alpha + 2557(255-Over.Alpha)) / (2557)`\
215	215		`31 LinearMultiplySat L := (min(Under.Red, Under.Green, Under.Blue) + max(Under.Red, Under.Green, Under.Blue)) / 2`\
216	216		` Mix := ``if`` Over = 0 ``then`` L`\
217	217		` ``else`` (Under - L)*Over/255 + L`\
218	218		`32 LinearHue Mix := HSL(Over.H, Under.S, Under.L)`\
219	219		`33 LinearColor Mix := HSL(Over.H, Over.S, Under.L)`\
220	220		`34 LinearLightness Mix := HSL(Under.H, Under.S, Over.L)`\
221	221		`35 LinearSaturation Mix := HSL(Under.H, ``if`` Under.S <> 0 ``then`` Over.S ``else`` 0, Under.L)`
222	222
223	223		`36 CorrectedHue Mix := GSB(Over.H, Under.S, Under.L)`\
224	224		`37 CorrectedColor Mix := GSB(Over.H, Over.S, Under.L)`\
225	225		`38 CorrectedLightness Mix := GSB(Under.H, Under.S, Over.L)`\
226	226		`39 CorrectedSat Mix := GSB(Under.H, ``if`` Under.S <> 0 ``then`` Over.S ``else`` 0, Under.L)`
227	227
228	228		For the mask, the color channels are unchanged and the ApplyOpacity value is used to multiply the destination alpha value:
229	229
230	230		`Final.Alpha := Under.Alpha * ApplyOpacity / 255`
231	231
232	232		For other blend modes, first the RGB channels are modified according to the underlying alpha value:
233	233
234	234		`Mix := (Mix * Under.Alpha + Over * (255 - Under.Alpha)) / 255`\
235	235		`Mix.Alpha := Over.Alpha`
236	236
237	237		In other words, if the layer underneath is transparent, the merged color is simply the color of the layer above.
238	238
239	239		Then the Mix color is drawn over the underlying layer. If Mix.Alpha = 255 then the color is replace:
240	240
241	241		` Final = Mix`
242	242
243	243		If Mix.Alpha = 0 then the color is unchanged:
244	244
245	245		` Final = Under`
246	246
247	247		Otherwise, the color is blended using the regular drawing function (with or without gamma correction depending on the stack option).
248	248
249	249		#### Computation of blend-over
250	250
251	251		Linear blend-over:
252	252
253	253		`var`` `\
254	254		` a1f, a2f, a12, a12m: DWord; //unsigned`\
255	255		`begin`\
256	256		` a12 := 65025 - (255 - UnderAlpha) * (255 - OverAlpha);`\
257	257		` a1f := UnderAlpha * (255 - OverAlpha);`\
258	258		` a2f := OverAlpha * 255;`\
259	259		` Under := (Under * a1f + Over * a2f + (a12 ``shr`` 1)) ``div`` a12; //for each channel`\
260	260		` UnderAlpha := (a12 + a12 ``shr`` 7) ``shr`` 8; `\
261	261		`end``;`
262	262
263	263		Blend-over with gamma correction:
264	264
265	265		`var`` `\
266	266		` a1f, a2f, a12, a12m: DWord; //unsigned`\
267	267		`begin`\
268	268		` a12 := 65025 - (255 - UnderAlpha) * (255 - OverAlpha);`\
269	269		` a1f := UnderAlpha * (255 - OverAlpha);`\
270	270		` a2f := OverAlpha * 255;`\
271	271		` Under := GC( (GE(Under) * a1f + GE(Over) * a2f + (a12 ``shr`` 1)) ``div`` a12 ); //for each channel`\
272	272		` UnderAlpha := (a12 + a12 ``shr`` 7) ``shr`` 8; `\
273	273		`end``;`
274	274
275	275		### Directory structure
276	276
277	277		The [[LazPaint LZP directory structure\|directory structure]] is optional. It can contain additional information about the layered image, in particular originals used to render the layers. A reader that does not handle vectorial information can discard it. Its offset is determined by the layer stack header.
278	278
279	279		---
280	280		From the Free Pascal wiki, page "LazPaint LZP Format", as it stood on 2021-03-01 after a change by Circular, out of the wiki's dump of 2024-05-26. Copyright stays with the author who wrote it, according to the old wiki's [[Licensing]] page. If you have concerns about this material, [message the admin](/wiki-contact).
281	281		Category: LazPaint

1

## LZP File format specification

2

3

In this document, [[Hexadecimal|hexadecimal]] numbers follow Pascal convention, i.e. they are prefixed with a [[Dollar sign|dollar sign]]. For example \$10 means 16. As an exception, sequences of bytes are not prefixed.

4

5

The default encoding for numbers is little endian. So the hexadecimal number \$12345678 would be stored as the sequence of bytes 78 56 34 12.

6

7

Numbers are 32-bit signed integers, unless stated otherwise. So -1 is stored like \$FFFFFFFF.

8

9

### Global structure

10

11

The file is composed of:

12

13

`Offset Content`\

14

`------ --------------------`\

15

`0 File header`\

16

`? Thumbnail (optional)`\

17

`? `[[LazPaint LZP Format Single Image|Flat image]]\

18

`? Layer structure (optional)`

19

20

Note: some old formats do not have a header and start directly with the flat image. Image readers may discard those files. However for your information, in this case, images are always stored with ZStream compression and the size of the image is deduced from the size of the flat image.

21

22

### File header

23

24

`Offset Size Name Content`\

25

`------ ---- ------------------ ------------------------`\

26

`0 8 `*`magic`*` Always contains the string 'LazPaint'`\

27

`8 4 `*`zero1`*` Contains the number 0`\

28

`12 4 `*`headerSize`*` Size of this header`\

30

`16 4 `*`width`*` Width of the image canvas`\

31

`20 4 `*`height`*` Height of the image canvas`\

32

`24 4 `*`nbLayers`*` Number of layers`\

33

`28 4 `*`previewOffset`*` Offset (from the beginning of the file) of the `[[LazPaint LZP Format Single Image|flattened image]]\

35

`32 4 `*`zero2`*` Contains the number 0`\

36

`36 4 `*`compressionMode`*` Contains flags indicating the compression and the presence of a thumbnail`\

37

`40 4 `*`reserved1`*` Reserved for future use. Set to 0.`\

38

`44 4 `*`layersOffset`*` Offset (from the beginning of the file) of the layer structure`

39

40

*headerSize* contains the size of this header. An image reader must read this to know which fields are available. It must be greater or equal to 48.

41

42

*width* and *height* are the dimension of the canvas. There is no real limitation for these numbers however it is recommended to keep it under or equal to 8192.

43

44

*nbLayers* is the number of layers. The minimum is 1. There is no real limitation however it is recommended to keep it under or equal to 99.

45

46

*layersOffset* can be 0, in which case the image is flat and that there is no layer description.

47

48

#### Compression

49

50

*compressionMode* contains flags indicating the compression and the presence of a thumbnail.

51

52

The low order byte indicates the compression of the flattened image. It is the value (*compressionMode* **and** \$ff):

53

54

\- \$01 indicates ZStream compression

55

56

\- \$02 indicates RLE compression

57

58

\- any other value should be considered as an indication that the image reader cannot interpret the content of the file

59

60

The following flags can be defined:

61

62

\- \$100 indicates that there is thumbnail in PNG format; check with (*compressionMode* **and** \$100) \<\> 0

63

64

If there is a thumbnail, it is directly after the header. The reader should read *headerSize* to have the correct location.

65

66

The compression of the layers is defined in the layer structure.

67

68

#### Thumbnail

69

70

Its presence is recommended however optional. Its presence is specified by the flag \$100 of *compressionMode*.

71

72

Currently, it is a PNG file embedded directly after the header. It is provided as a way to facilitate quick preview.

73

74

PNG format is widespread so it is easy to implement a browser that would display the thumbnail. To do so, check that the header contains *magic* and that the flag for the thumbnail is set in *compressionMode*. Then jump to the end of the header using *headerSize*. Then read the PNG file.

75

76

There is no real limitation but the recommended size for the thumbnail is 128 along the largest dimension. The other dimension may be smaller if the image is not a square. Of course if the image is smaller than 128 then the thumbnail will be smaller than 128.

77

78

### Flat image

79

80

The file always include a flattened image to make it easier for image viewers. It is in the [[LazPaint LZP Format Single Image|LZP single image format]].

81

82

If there is no layer structure, the caption of this image is the name of the unique layer. If there is a layer structure, then the caption of this image is discarded.

83

84

### Layer structure

85

86

`Offset Size Content`\

87

`------ ---- -------------------------------------------------------`\

88

`0 20 Contains the string 'TBGRALayeredBitmap' followed by the bytes 1A 00`\

89

`20 4 Size S of the additional header after this field (must be at least 12)`\

90

`24 4 Number of layers N (should be identical to the number in the file header)`\

91

`28 4 Selected layer index. Equals to -1 if no layer is selected, otherwise positive from 0 to N-1`\

92

`32 4 Stack flags`\

93

`36 4 Canvas width (optional)`\

94

`40 4 Canvas height (optional)`\

95

`44 8 64-bit offset to the directory structure relative to the start of this header (optional)`\

96

`24+S First layer (bottom one)`

97

98

The following flags are defined:

99

100

- \$01: the layers are blended without gamma correction

101

- \$02: the layers are compressed with RLE compression (otherwise they are compressed with ZStream compression)

102

103

If the canvas size is not supplied (when the total header size is less than 44), then the canvas size is the size of the first non-empty layer bitmap.

104

105

Similarly, the [[LazPaint LZP directory structure|directory structure]] offset may not to be supplied or set to zero. In this case, there is none. It is optional and only necessary for vector shapes.

106

107

The total header size of the layer structure is equal to 24 + S. Then for each layer there is a layer entry.

108

109

The first layer entry is the bottom most layer. Then each layer is blended over the previous ones. The last layer is on top and is not hidden by any other layer.

110

111

#### Layer entry

112

113

Header of the layer entry:

114

115

`Offset Size Content`\

116

`------ ---- -------------------------------------------------------`\

117

`0 4 Size S1 of the optional fields that follows`\

118

`4 4 Layer flags (lowest bit means that it is visible)`\

119

`8 4 Layer blend operation`\

120

`12 4 Layer horizontal offset in pixels`\

121

`16 4 Layer vertical offset in pixels`\

122

`20 4 Layer ID (if used, each layer must have a different ID in a given file)`\

123

`24 4 Layer opacity from 0 (transparent) to 65535 (opaque)`\

124

`28 4 Lower bits of size S2 of image data`\

125

`32 4 Upper bits of size S2 (optional, forming 64-bit integer)`\

126

`36 16 GUID of the original used to render the layer (optional)`\

127

`52 24 Affine matrix of the transform applied to the original`\

128

`4+S1 S2 `[[LazPaint LZP Format Single Image|Image data]]\

129

`4+S1+S2 Next entry`

130

131

All fields are optional so in theory, S1 can be set to zero and the image data can directly follow. However it is recommended to write the whole header.

132

133

The default values are: visible, blend operation boTransparent, offset (0,0), undefined ID (maximum integer value), opaque, size S2 = 0, GUID = NULL. If an original is specified the affine matrix transform is mandatory.

134

135

The size of the image data is a 64-bit integer, though the upper value is optional and by default zero. When the image size S2 \<\> 0, the offset to the next entry is computed using the formula 4+S1+S2. Otherwise, the data is expected to continue after the end of the image data.

136

137

The GUID is stored here as binary (in little endian): one DWord, two Words and 8 bytes.

138

139

The affine transform is stored as binary (in little endian): 6 single precision floats ux, uy, vx, vy, tx, ty (axes u and v and translation t).

140

141

#### Blend operations

142

143

- Under is the value of the channels of the layer underneath

144

- Over is the value of the channels of the layer above

145

- Mix is an intermediate value that is then blended over the previous layers, and that, unless specified, is computed for each RGB channel. The alpha channel is computed differently (see below table).

146

- Final is the resulting value

147

148

Note:

149

150

- values (Red, Green, Blue, Alpha, Mix, Under, Over, Final) are considered to be between 0 and 255

151

- GE means gamma expansion. GE(x) = Power(x/255, 2.1)*65535

152

- GC means gamma compression. GC(x) = Power(x/65535, 1/2.1)*255

153

- ByteSqrt means the square root. ByteSqrt(x) = Power(X/255, 0.5)*255

154

- **div** stands for integer division (rounded down)

155

- HSL is the standard HSL colorspace, without gamma correction

156

- GSB is the corrected HSL colorspace of LazPaint. It is computed with gamma correction and the B (brightness) component correspond to the perceived lightness:

157

158

`B = (306*GE(Red) + 601*GE(Green) + 117*GE(Blue))/1024`\

159

`B between 0 and 65535`

160

161

Here is the list of operations supported in LazPaint:

162

163

`Value Name Description`\

164

`----- ----------------- ----------------------------------------------------`\

165

`0 LinearBlend Blend over the previous layers without gamma correction`\

166

`1 Transparent Blend over the previous layers `\

167

` (with gamma correction unless it is specified otherwise in the stack flags)`\

169

`2 Lighten Mix := Max(Over, Under)`\

170

`3 Screen Mix := 255 - ((255 - Under) * (255 - Over) `**`shr`**` 8)`\

171

`4 Additive Mix := GC(GE(Under)+GE(Over))`\

172

`5 LinearAdd Mix := Under + Over`\

173

`6 ColorDodge Mix := `**`if`**` Over = 255 `**`then`**` 255`\

174

` `**`else`**` Min(255, (Under `**`shl`**` 8) `**`div`**` (255 - Over))`\

175

`7 Divide Mix := `**`if`**` Over = 0 `**`then`**` 255`\

176

` `**`else`**` Min(255, (Under `**`shl`**` 8) `**`div`**` Over)`\

177

`8 NiceGlow Mix := `**`if`**` Under = 255 `**`then`**` 255`\

178

` `**`else`**` Min(255, Over*Over `**`div`**` (255 - Under))`\

179

`9 SoftLight Mix := ((255 - Under)*Over shr 7 + Under)*Under div 255`\

180

`10 HardLight Mix := `**`if`**` Over <= 128 `**`then`**` Under*Over `**`shr`**` 7`\

181

` `**`else`**` 255 - ((255 - Under)*(255 - Over) `**`shr`**` 7)`\

183

`11 Glow Mix := `**`if`**` Under = 255 `**`then`**` 255`\

184

` `**`else`**` Min(255, Over*Over `**`div`**` (255 - Under))`\

185

`12 Reflect Mix := `**`if`**` Over = 255 `**`then`**` 255`\

186

` `**`else`**` Min(255, Under*Under `**`div`**` (255 - Over))`\

187

`13 Overlay Mix := `**`if`**` Under <= 128 `**`then`**` Under*Over `**`shr`**` 7`\

188

` `**`else`**` 255 - ((255 - Under)*(255 - Over) `**`shr`**` 7)`\

189

`14 DarkOverlay Mix := `**`if`**` GE(Under) <= 32768 `**`then`**` Under*Over `**`shr`**` 7`\

190

` `**`else`**` GC(65535 - ((65535 - GE(Under))*(65535 - GE(Over)) `**`shr`**` 15)`\

191

`15 Darken Mix := Min(Under, Over)`\

192

`16 Multiply Mix := `**`if`**` Over >= 128 `**`then`**` Under*(Over+1) `**`shr`**` 8 `**`else`**` Under*Over `**`shr`**` 8`\

193

`17 ColorBurn Mix := `**`if`**` Over = 0 `**`then`**` 0`\

194

` `**`else`**` Max(0, 255 - (Under `**`shl`**` 8) `**`div`**` Over)`\

196

`18 Difference Mix := GC(Abs(GE(Under) - GE(Over)))`\

197

`19 LinearDifference Mix := Abs(Under - Over)`\

198

`20 Exclusion Mix := GC(GE(Under) + GE(Over) - (GE(Under)*GE(Over) `**`shr`**` 15)`\

199

`21 LinearExclusion Mix := Under + Over - (Under*Over `**`shr`**` 7)`\

200

`22 Subtract Mix := GC(Max(0, GE(Under) - GE(Over)))`\

201

`23 LinearSubtract Mix := Max(0, Under - Over)`\

202

`24 SubtractInverse Mix := GC(Max(0, GE(Under) - (65535 - GE(Over))))`\

203

`25 LinearSubtractInv Mix := Max(0, Under - (255 - Over))`\

205

`26 Negation Mix := `**`if`**` GE(Under) + GE(Over) <= 65535 `**`then`**` GC(GE(Under) + GE(Over))`\

206

` `**`else`**` GC(131071 - (GE(Under) + GE(Over)))`\

207

`27 LinearNegation Mix := `**`if`**` Under + Over <= 255 `**`then`**` Under + Over`\

208

` `**`else`**` 511 - (Under + Over)`\

209

`28 Xor Mix := Under `**`xor`**` Over`\

211

`29 SvgSoftLight Mix := `**`if`**` Over <= 128 `**`then`**` Under - (((256 - 2*Over)* Under `**`shr`**` 8)*(255 - Under) `**`shr`**` 8)`\

212

` `**`else`**` `**`if`**` Under <= 64 `**`then`**` Under + ((2*Over - 256) * (Under*7 - (4*Under*(4*Under + 256)*(256 - Under) `**`shr`**` 16)) `**`shr`**` 8)`\

213

` `**`else`**` Under + ((2*Over - 255) * (ByteSqrt(Under)-Under) `**`shr`**` 8)`\

214

`30 Mask ApplyOpacity := ((2*Over.Red + 4*Over.Green + Over.Blue)*Over.Alpha + 255*7*(255-Over.Alpha)) / (255*7)`\

215

`31 LinearMultiplySat L := (min(Under.Red, Under.Green, Under.Blue) + max(Under.Red, Under.Green, Under.Blue)) / 2`\

216

` Mix := `**`if`**` Over = 0 `**`then`**` L`\

217

` `**`else`**` (Under - L)*Over/255 + L`\

218

`32 LinearHue Mix := HSL(Over.H, Under.S, Under.L)`\

219

`33 LinearColor Mix := HSL(Over.H, Over.S, Under.L)`\

220

`34 LinearLightness Mix := HSL(Under.H, Under.S, Over.L)`\

221

`35 LinearSaturation Mix := HSL(Under.H, `**`if`**` Under.S <> 0 `**`then`**` Over.S `**`else`**` 0, Under.L)`

222

223

`36 CorrectedHue Mix := GSB(Over.H, Under.S, Under.L)`\

224

`37 CorrectedColor Mix := GSB(Over.H, Over.S, Under.L)`\

225

`38 CorrectedLightness Mix := GSB(Under.H, Under.S, Over.L)`\

226

`39 CorrectedSat Mix := GSB(Under.H, `**`if`**` Under.S <> 0 `**`then`**` Over.S `**`else`**` 0, Under.L)`

227

228

For the mask, the color channels are unchanged and the ApplyOpacity value is used to multiply the destination alpha value:

229

230

`Final.Alpha := Under.Alpha * ApplyOpacity / 255`

231

232

For other blend modes, first the RGB channels are modified according to the underlying alpha value:

233

234

`Mix := (Mix * Under.Alpha + Over * (255 - Under.Alpha)) / 255`\

235

`Mix.Alpha := Over.Alpha`

236

237

In other words, if the layer underneath is transparent, the merged color is simply the color of the layer above.

238

239

Then the Mix color is drawn over the underlying layer. If Mix.Alpha = 255 then the color is replace:

240

241

` Final = Mix`

242

243

If Mix.Alpha = 0 then the color is unchanged:

244

245

` Final = Under`

246

247

Otherwise, the color is blended using the regular drawing function (with or without gamma correction depending on the stack option).

248

249

#### Computation of blend-over

250

251

Linear blend-over:

252

253

**`var`**` `\

254

` a1f, a2f, a12, a12m: DWord; //unsigned`\

255

**`begin`**\

256

` a12 := 65025 - (255 - UnderAlpha) * (255 - OverAlpha);`\

257

` a1f := UnderAlpha * (255 - OverAlpha);`\

258

` a2f := OverAlpha * 255;`\

259

` Under := (Under * a1f + Over * a2f + (a12 `**`shr`**` 1)) `**`div`**` a12; //for each channel`\

260

` UnderAlpha := (a12 + a12 `**`shr`**` 7) `**`shr`**` 8; `\

261

**`end`**`;`

262

263

Blend-over with gamma correction:

264

265

**`var`**` `\

266

` a1f, a2f, a12, a12m: DWord; //unsigned`\

267

**`begin`**\

268

` a12 := 65025 - (255 - UnderAlpha) * (255 - OverAlpha);`\

269

` a1f := UnderAlpha * (255 - OverAlpha);`\

270

` a2f := OverAlpha * 255;`\

271

` Under := GC( (GE(Under) * a1f + GE(Over) * a2f + (a12 `**`shr`**` 1)) `**`div`**` a12 ); //for each channel`\

272

` UnderAlpha := (a12 + a12 `**`shr`**` 7) `**`shr`**` 8; `\

273

**`end`**`;`

274

275

### Directory structure

276

277

The [[LazPaint LZP directory structure|directory structure]] is optional. It can contain additional information about the layered image, in particular originals used to render the layers. A reader that does not handle vectorial information can discard it. Its offset is determined by the layer stack header.

278

279

---

280

*From the Free Pascal wiki, page "LazPaint LZP Format", as it stood on 2021-03-01 after a change by Circular, out of the wiki's dump of 2024-05-26. Copyright stays with the author who wrote it, according to the old wiki's [[Licensing]] page. If you have concerns about this material, [message the admin](/wiki-contact).*

281

Category: LazPaint