simples of questions

[jamesb]

How do I create a new image, not from a file, using Magick Core? It needs to be 50k x 50k. I then need to set the pixels, in blocks of 16x16, and save the file. I have searched this forum, and the website, and not found anything that involves creating an image from scratch - only loading images and modifying them. Be grateful for some simple example code. Thanks.
James

[anthony]

Create an image with...

Code: Select all

  convert  -size 50000x50000 xc:black   save_image.png

to also set teh top left 16x16 area to red do...

Code: Select all

  convert  -size 50000x50000 xc:black   \
               -draw 'fill red -rectangle 0,0 15,15' \
               save_image.png

However if ALL 16x16 blocks are solid colors. a better way may be to create your image as a single pixel, and -scale it by 1600%

for example generating a 160x160 image with just the top left corner red
can be create dby creating a 10x10 pixel image with just one pixel red!

create a TXT image file (See http://www.imagemagick.org/Usage/files/#txt )
and call it "pixel_image.txt"

# ImageMagick pixel enumeration: 10,10,255,rgb
0,0: (255,0,0)

Now scale it

Code: Select all

  convert pixel_image.txt -scale 1600%  save_image.png

There are hundreds of ways of generating images.
See more in IM Examples, Canvases
http://www.imagemagick.org/Usage/canvas/

[jamesb]

I meant the C Magick Core interface. I just found the Architecture page and read that. What's missing is how to specify what format to save the file in, or rather, what specific version of the format. I want to save 64bit PNG, (RGBA 16bpc).

Thanks
Jamie

[magick]

To create a 50k x 50k image canvas use these MagickCore API calls:

Code: Select all

  Image *image = AcquireImage( image_info );
  if (SetImageExtent( image, 50*1024, 50*1024 ) == MagickFalse )
    perror( "somethings wrong" );

[jamesb]

How to specify saving 16bpc PNG vs 8bpc? Or float TIFF vs 8bpc TIFF? Or RGBA vs RGB?

[magick]

Use SetImageDepth() to specify the image depth:

• SetImageDepth( image, 8 );

To produce float TIFF images, try:

• SetImageProperty( image, "quantum:format", "floating-point");

For RGBA images, use:

• SetImageAlphaChannel( image, ActivateAlphaChannel );