#N canvas 480 0 620 717 10; #X obj 0 0 doc_demo; #X obj 0 817 doc_f; #X text 20 51 the default format of an image in GridFlow is a grid that has three dimensions \, in which the first one is for the rows \, the second one is for the columns \, and the third one is for channels. Thus they are often numbered like this (starting at 0) :; #X text 20 170 However \, they don't need to be represented like that \, so \, you can change the order of dimensions \, but this changes the order of the data in memory \, and most of all \, it won't work with any classes that assume the default numbering of the dimensions. In several cases \, though \, the class doesn't care about the numbering \, and it is up to the user to know which dimension is which \, and it is in this context that it is possible to use alternate representations. ; #X text 279 310 Channels for the YUV color model are :; #X text 14 309 Channels for the RGB color model are :; #X text 280 371 Channels for the Y color model are :; #X text 382 383 (only one channel !); #X text 14 371 Channels for the RGBA color model are :; #X text 280 406 Channels for the YA color model are :; #X obj 3 797 doc_also; #X obj 3 27 doc_section About Images; #X text 38 123 0: rows; #X text 38 136 1: columns; #X text 38 150 2: channels; #X obj 3 289 doc_section Channels and Color Models; #X text 39 322 0: red; #X text 39 334 1: green; #X text 39 347 2: blue; #X text 39 384 0: red; #X text 39 396 1: green; #X text 39 409 2: blue; #X text 39 421 3: opacity; #X text 305 322 0: grey; #X text 305 335 1: blue/yellow (blue minus grey); #X text 305 348 2: red/cyan (red minus grey); #X text 311 382 0: grey; #X text 311 417 0: grey; #X text 311 430 1: opacity; #X obj 3 580 doc_section Number Range; #X obj 103 797 doc_link grid 1; #X text 17 450 You can also invent your own color models \, with as many channels as you want.; #X text 18 479 Many classes only support RGB \, or RGB and RGBA \, or a small set of colour models. Many others accept anything but will assume that it's working on RGB (thus with other colour models it will do something wrong \, which might be weird \, confusing \, or really really interesting). Many others don't even know what a channel is and so you're supposed to know what you're doing.; #X text 18 602 Because Grids are made of 32-bit integers \, a three-channel picture uses 96 bpp (bits per pixel) \, and have to be downscaled to 24 bpp (or 16 bpp) for display. That huge amount of slack is there because when you create your own effects you often have intermediate results that need to be of higher precision than a normal picture. Especially \, results of multiplications are big and should not overflow before you divide them back to normal \; and similarly \, you can have negative values all over \, as long as you take care of them before they get to the display. In the final conversion \, high bits are just ignored. This means: black is 0 \, maximum is 255 \, and values wrap like with [# % 256] or [# & 255]. If you want to clip them \, you may use [# max 0] and [# min 255] objects.; #X obj 157 797 doc_link indexmap 1; #X connect 10 1 30 0; #X connect 10 1 34 0;