Image(ID)

Image Data

The Image data-block is a shallow wrapper around image or video file(s) (on disk, as packed data, or generated).

All actual data like the pixel buffer, size, resolution etc. is cached in an imbuf.types.ImBuf image buffer (or several buffers in some cases, like UDIM textures, multi-views, animations…).

Several properties and functions of the Image data-block are then actually using/modifying its image buffer, and not the Image data-block itself.

Warning

One key limitation is that image buffers are not shared between different Image data-blocks, and they are not duplicated when copying an image.

So until a modified image buffer is saved on disk, duplicating its Image data-block will not propagate the underlying buffer changes to the new Image.

This example script generates an Image data-block with a given size, change its first pixel, rescale it, and duplicates the image.

The duplicated image still has the same size and colors as the original image at its creation, all editing in the original image’s buffer is ‘lost’ in its copy.

import bpy

image_src = bpy.data.images.new('src', 1024, 102)
print(image_src.size)
print(image_src.pixels[0:4])

image_src.scale(1024, 720)
image_src.pixels[0:4] = (0.5, 0.5, 0.5, 0.5)
image_src.update()
print(image_src.size)
print(image_src.pixels[0:4])

image_dest = image_src.copy()
image_dest.update()
print(image_dest.size)
print(image_dest.pixels[0:4])

base classes — bpy_struct, ID

class bpy.types.Image(ID)

Image data-block referencing an external or packed image

alpha_mode

Representation of alpha in the image file, to convert to and from when saving and loading the image

STRAIGHT Straight – Store RGB and alpha channels separately with alpha acting as a mask, also known as unassociated alpha. Commonly used by image editing applications and file formats like PNG.
PREMUL Premultiplied – Store RGB channels with alpha multiplied in, also known as associated alpha. The natural format for renders and used by file formats like OpenEXR.
CHANNEL_PACKED Channel Packed – Different images are packed in the RGB and alpha channels, and they should not affect each other. Channel packing is commonly used by game engines to save memory.
NONE None – Ignore alpha channel from the file and make image fully opaque.

Type: enum in [‘STRAIGHT’, ‘PREMUL’, ‘CHANNEL_PACKED’, ‘NONE’], default ‘STRAIGHT’

bindcode

OpenGL bindcode

Type: int in [0, inf], default 0, (readonly)

channels

Number of channels in pixels buffer

Type: int in [0, inf], default 0, (readonly)

colorspace_settings

Input color space settings

Type: ColorManagedInputColorspaceSettings, (readonly)

depth

Image bit depth

Type: int in [0, inf], default 0, (readonly)

display_aspect

Display Aspect for this image, does not affect rendering

Type: float array of 2 items in [0.1, inf], default (1.0, 1.0)

file_format

Format used for re-saving this file

Type: enum in Image Type Items, default ‘TARGA’

filepath

Image/Movie file name

Type: string, default “”, (never None)

filepath_raw

Image/Movie file name (without data refreshing)

Type: string, default “”, (never None)

frame_duration

Duration (in frames) of the image (1 when not a video/sequence)

Type: int in [0, inf], default 0, (readonly)

generated_color

Fill color for the generated image

Type: float array of 4 items in [0, inf], default (0.0, 0.0, 0.0, 0.0)

generated_height

Generated image height

Type: int in [1, 65536], default 1024

generated_type

Generated image type

Type: enum in Image Generated Type Items, default ‘UV_GRID’

generated_width

Generated image width

Type: int in [1, 65536], default 1024

has_data

True if the image data is loaded into memory

Type: boolean, default False, (readonly)

is_dirty

Image has changed and is not saved

Type: boolean, default False, (readonly)

is_float

True if this image is stored in floating-point buffer

Type: boolean, default False, (readonly)

is_multiview

Image has more than one view

Type: boolean, default False, (readonly)

is_stereo_3d

Image has left and right views

Type: boolean, default False, (readonly)

packed_file

First packed file of the image

Type: PackedFile, (readonly)

packed_files

Collection of packed images

Type: bpy_prop_collection of ImagePackedFile, (readonly)

pixels

Image buffer pixels in floating-point values

Type: float in [-inf, inf], default 0.0

render_slots

Render slots of the image

Type: RenderSlots bpy_prop_collection of RenderSlot, (readonly)

resolution

X/Y pixels per meter, for the image buffer

Type: float array of 2 items in [-inf, inf], default (0.0, 0.0)

size

Width and height of the image buffer in pixels, zero when image data can’t be loaded

Type: int array of 2 items in [-inf, inf], default (0, 0), (readonly)

source

Where the image comes from

FILE Single Image – Single image file.
SEQUENCE Image Sequence – Multiple image files, as a sequence.
MOVIE Movie – Movie file.
GENERATED Generated – Generated image.
VIEWER Viewer – Compositing node viewer.
TILED UDIM Tiles – Tiled UDIM image texture.

Type: enum in [‘FILE’, ‘SEQUENCE’, ‘MOVIE’, ‘GENERATED’, ‘VIEWER’, ‘TILED’], default ‘FILE’

stereo_3d_format

Settings for stereo 3d

Type: Stereo3dFormat, (readonly, never None)

tiles

Tiles of the image

Type: UDIMTiles bpy_prop_collection of UDIMTile, (readonly)

type

How to generate the image

Type: enum in [‘IMAGE’, ‘MULTILAYER’, ‘UV_TEST’, ‘RENDER_RESULT’, ‘COMPOSITING’], default ‘IMAGE’, (readonly)

use_deinterlace

Deinterlace movie file on load

Type: boolean, default False

use_generated_float

Generate floating-point buffer

Type: boolean, default False

use_half_precision

Use 16 bits per channel to lower the memory usage during rendering

Type: boolean, default False

use_multiview

Use Multiple Views (when available)

Type: boolean, default False

use_view_as_render

Apply render part of display transformation when displaying this image on the screen

Type: boolean, default False

views_format

Mode to load image views

Type: enum in Views Format Items, default ‘INDIVIDUAL’

save_render(filepath, scene=None, quality=0)

Save image to a specific path using a scenes render settings

Parameters

filepath (string, (never None)) – Output path
scene (Scene, (optional)) – Scene to take image parameters from
quality (int in [0, 100], (optional)) – Quality, Quality for image formats that support lossy compression, uses default quality if not specified

save(filepath='', quality=0)

Save image

Parameters

filepath (string, (optional, never None)) – Output path, uses image data-block filepath if not specified
quality (int in [0, 100], (optional)) – Quality, Quality for image formats that support lossy compression, uses default quality if not specified

pack(data='', data_len=0)

Pack an image as embedded data into the .blend file

Parameters

data (string, (optional, never None)) – data, Raw data (bytes, exact content of the embedded file)
data_len (int in [0, inf], (optional)) – data_len, length of given data (mandatory if data is provided)

unpack(method='USE_LOCAL')

Save an image packed in the .blend file to disk

Parameters: method (enum in Unpack Method Items, (optional)) – method, How to unpack

reload(): Reload the image from its source path

update(): Update the display image from the floating-point buffer

scale(width, height)

Scale the buffer of the image, in pixels

Parameters

width (int in [1, inf]) – Width
height (int in [1, inf]) – Height

gl_touch(frame=0, layer_index=0, pass_index=0)

Delay the image from being cleaned from the cache due inactivity

Parameters

frame (int in [0, inf], (optional)) – Frame, Frame of image sequence or movie
layer_index (int in [0, inf], (optional)) – Layer, Index of layer that should be loaded
pass_index (int in [0, inf], (optional)) – Pass, Index of pass that should be loaded

Returns

Error, OpenGL error value

Return type

int in [-inf, inf]

gl_load(frame=0, layer_index=0, pass_index=0)

Load the image into an OpenGL texture. On success, image.bindcode will contain the OpenGL texture bindcode. Colors read from the texture will be in scene linear color space and have premultiplied or straight alpha matching the image alpha mode

Parameters

frame (int in [0, inf], (optional)) – Frame, Frame of image sequence or movie
layer_index (int in [0, inf], (optional)) – Layer, Index of layer that should be loaded
pass_index (int in [0, inf], (optional)) – Pass, Index of pass that should be loaded

Returns

Error, OpenGL error value

Return type

int in [-inf, inf]

gl_free(): Free the image from OpenGL graphics memory

filepath_from_user(image_user=None)

Return the absolute path to the filepath of an image frame specified by the image user

Parameters: image_user (ImageUser, (optional)) – Image user of the image to get filepath for
Returns: File Path, The resulting filepath from the image and its user
Return type: string, (never None)

buffers_free(): Free the image buffers from memory

classmethod bl_rna_get_subclass(id, default=None)

Parameters: id (string) – The RNA type identifier.
Returns: The RNA type or default when not found.
Return type: bpy.types.Struct subclass

classmethod bl_rna_get_subclass_py(id, default=None)

Parameters: id (string) – The RNA type identifier.
Returns: The class or default when not found.
Return type: type

Inherited Properties

Inherited Functions

References