1.. _imwri: 2 3ImageMagick Writer-Reader 4========================= 5 6ImageMagick Writer-Reader (IMWRI) is a plugin that can read and write many image formats. 7 8.. function:: Write(clip clip, string imgformat, string filename[, int firstnum=0, int quality=75, bint dither=True, string compression_type, bint overwrite=False, clip alpha]) 9 :module: imwri 10 11 Supported input formats for writing: 12 ImageMagick with Quantum Depth 16 and HDRI: 8-16 bit integer, 32 bit float 13 14 Write will write each frame to disk as it's requested. If a frame is never requested it's also never written to disk. 15 16 Parameters: 17 clip 18 Input clip. RGB and Gray supported. See list for accepted inputs. 19 20 imgformat 21 The name of the output format. Examples of supported format strings are "JPEG", "PNG", and "DPX". Visit the ImageMagick website for a full list. 22 23 filename 24 The filename string must have one or more frame number substitutions. The syntax is printf style. For example "image%06d.png" or "/images/%d.jpg" is common usage. 25 26 firstnum 27 The first image number in the sequence to write. 28 29 quality 30 Quality adjustment for formats where it's applicable. Range is 0 to 100. 31 32 dither 33 Use Floyd–Steinberg dithering if the input needs to be reduced in depth. 34 35 compression_type 36 Select the specific compression type for *imgformats* that have more than one possible compression method. Recognized constants are: 37 Undefined, None, BZip, DXT1, DXT3, DXT5, Fax, Group4, JPEG, JPEG2000, LosslessJPEG, LZW, RLE, Zip, ZipS, Piz, Pxr24, B44, B44A, LZMA, JBIG1, JBIG2 38 39 overwrite 40 Overwrite already existing files. This option also disables the requirement that output filenames contain a number. 41 42 alpha 43 A grayscale clip containing the alpha channel for the image to write. Apart from being grayscale, its properties must be identical to the main *clip*. 44 45 46.. function:: Read(string[] filename[, int firstnum=0, bint mismatch=False, bint alpha=False, bint float_output = False]) 47 :module: imwri 48 49 Possible output formats when reading: 8-16 bit integer and 32 bit float 50 51 Note that by default 8-16 bit images are returned as integer and 32 bit images as float. When reading half precision float images you have to manually set *float_output* to have the unmodified floating point range returned. 52 53 Read is a simple function for reading single or series of images and returning them as a clip. 54 55 Parameters: 56 filename 57 The filename argument has two main modes. Either it takes a list of 1 or more files to open in the given order, or it takes a single filename string with one or more frame number substitutions. The syntax is printf style. For example "image%06d.png" or "/images/%d.jpg" is common usage. 58 59 firstnum 60 The first image number to start reading from when reading a sequence. 61 62 mismatch 63 Allow reading of multiple images with different resolutions. If required and not set, an error will be generated. 64 65 alpha 66 Return the alpha channel from the read images as a separate grayscale clip. Note that an alpha channel clip is always returned when this parameter is set, even for image formats without support for it. 67 68 float_output 69 Always return the read image in a float format. Due to the output format guessing this option can be useful when reading half precision float images.