moviePtr = Screen(‘CreateMovie’, windowPtr, movieFile [, width][, height][, frameRate=30][, movieOptions][, numChannels=4][, bitdepth=8]);
Create a new movie file with filename ‘movieFile’ and according to given
The function returns a handle ‘moviePtr’ to the file.
Currently only single-track video encoding is supported.
See ‘Screen AddAudioBufferToMovie?’ on how to add audio tracks to movies.
Movie creation is a 3 step procedure:
- Create a movie and define encoding options via ‘CreateMovie’.
- Add video and audio data to the movie via calls to ‘AddFrameToMovie’ et al.
- Finalize and close the movie via a call to ‘FinalizeMovie’.
All following parameters are optional and have reasonable defaults:
‘width’ Width of movie video frames in pixels. Defaults to width of window
‘height’ Height of movie video frames in pixels. Defaults to height of window
‘frameRate’ Playback framerate of movie. Defaults to 30 fps. Technically this is
not the playback framerate but the granularity in 1/frameRate seconds with which
the duration of a single movie frame can be specified. When you call
‘AddFrameToMovie’, there’s an optional parameter ‘frameDuration’ which defaults
to one. The parameter defines the display duration of that frame as the fraction
‘frameDuration’ / ‘frameRate’ seconds, so ‘frameRate’ defines the denominator of
that term. However, for a default ‘frameDuration’ of one, this is equivalent to
the ‘frameRate’ of the movie, at least if you leave everything at defaults.
‘movieoptions’ a textstring which allows to define additional parameters via
keyword=parm pairs. For [GStreamer](GStreamer) movie writing, you can provide the same
options as for [GStreamer](GStreamer) video recording. See ‘help VideoRecording’ for
supported options and tips.
Keywords unknown to a certain implementation or codec will be silently ignored:
EncodingQuality=x Set encoding quality to value x, in the range 0.0 for lowest
movie quality to 1.0 for highest quality. Default is 0.5 = normal quality. 1.0
often provides near-lossless encoding.
‘numChannels’ Optional number of image channels to encode: Can be 1, 3 or 4 on
OpenGL graphics hardware, and 3 or 4 on OpenGL-ES hardware. 1 = Red/Grayscale
channel only, 3 = RGB, 4 = RGBA. Please note that not all video codecs can
encode pure 1 channel data or RGBA data, ie. an alpha channel. If an unsuitable
codec is selected, movie writing may fail, or unsupported channels (e.g., the
alpha channel) may get silently discarded. It could also happen that a codec
which doesn’t support 1 channel storage will replicate the Red/Grayscale data
into all three RGB channels, leading to no data loss but increased movie file
size. Default is to request RGBA 4 channel data from the system, encoding to
RGBA or RGB, depending on codec.
‘bitdepth’ Optional color/intensity resolution of each channel: Default is 8
bpc, for 8 bit per component storage. OpenGL graphics hardware, but not
OpenGL-ES, also supports 16 bpc image readback. However, not all codecs can
encode with > 8 bpc color/luminance precision, so encoding with 16 bpc may fail
or silently reduce precision to less bits, possibly 8 bpc or less. If you
specify the special keyword UsePTB16BPC in ‘movieoptions’, then PTB will use its
own proprietary 16 bpc format for 1 or 3 channel mode. This format can only be
read by PTB’s own movie playback functions, not by other software.
In general, embedded OpenGL-ES graphics hardware is more restricted in the type
of image data it can return. Most video codecs are lossy codecs. They will
intentionally throw away color or spatial precision of encoded video to reduce
video file size or network bandwidth, often in a way that is not easily
perceptible to the naked eye. If you require high fidelity, make sure to
double-check your results for a given codec + parameter setup, e.g., via
encoding + decoding the movie and checking the original data against decoded