Captures a region of a screen and returns a new BGRA32 texture for it.

This function is intended for "one-off" uses (e.g. capturing the desktop and using it as a background texture) - if you want to capture repeatedly, consider instantiating DesktopScreenshot.

var tex = DesktopScreenshot.Capture(new RectInt(200, 100, 400, 300));
GetComponent<RawImage>().texture = tex;

would capture a 400x300 region starting at x=200 and y=100 in primary display's top-left corner and assign it as a texture to caller's RawTexture.

Creates a screenshoter for capturing regions of specified size.

This will create a BGRA32 texture to it.

Creates a screenshoter for an capturing regions into the texture.

Note that the texture should have format set to TextureFormat.BGRA32.

Don't forget to call the destructor when you're done using a screenshoter to free up the associated native bitmap.

This will not do anything to your texture and can be called more than once without harm.

The screenshoter's target texture.

You can resize or swap it out as you see fit, but you should be aware that there is minimal overhead overhead related to re-creating a native bitmap prior to capturing with changed dimensions.

A shortcut for texture.width

A shortcut for texture.height

A shortcut for texture.Resize.

Clears pixels in the texture to RGBA(0,0,0,0).

This can be slightly faster than doing texture.SetPixels.

Captures a section of the screen into the current texture.

x,y point at the section's top-left corner, relative to top-left corner of the primary display.

Returns a RectInt covering bounds of the primary display.

x/y of this rect are always 0.

The following would take a screenshot of the primary display and save it to screenshot.png:

var tex = DesktopScreenshot.Capture(DesktopScreenshot.GetDisplayBounds());
System.IO.File.WriteAllBytes("screenshot.png", tex.EncodeToPNG());

Returns a RectInt covering bounds of desktop (encompassing all displays).

x/y count from top-left and can be negative
(if a secondary display is to the left/top of the primary display).

The following would take a screenshot of desktop and save it to screenshot.png:

var tex = DesktopScreenshot.Capture(DesktopScreenshot.GetDesktopBounds());
System.IO.File.WriteAllBytes("screenshot.png", tex.EncodeToPNG());

Returns an array of classes with per-display information.

Each DisplayInfo has the following fields:

The following would log information about all available displays:

foreach (var inf in DesktopScreenshot.GetDisplayInfos()) {
    Debug.Log(inf);
}

The package is implemented entirely via P/Invoke, which means a few things:

• It does not have any additional dependencies.
• You can edit/extend the implementation (e.g. capture parts of windows instead of desktop) without any additional setup.

This package uses GDI's BitBlt, which is by far the most reliable way capture pixels of the screen, but is CPU-bound - for example, capturing a 1920x1080 region takes from 10ms on higher-end systems to 30ms on lower-end systems.

In other words, it is slightly less suitable for fullscreen video capture, though some capture software (e.g. OBS) does offer this method as a compatibility option.

The package is compatible with a multitude of Unity versions - it only needs Texture2D class to exist and to have a LoadRawTextureData method, which had been around since at least 4.x.

If you for some reason need to use this in Unity 5.2 or earlier, you will need to comment a single line of code, conveniently marked with a // <- needs to be commented out in pre-5.3 comment.

Being something that uses WinAPI and GDI, this package is only compatible with Windows.

While it may be possible to have it work on Mac/Linux, that requires a complete rewrite per-platform.