Manpages

NAME

video capturing program for X windows

SYNOPSIS

xvidcap|gvidcap [-v] [--fps <FPS>] [--time <seconds>] [--frames <frames>] [--continue [yes|no]] [--cap_geometry <WxH+X+Y>] [--start_no <start_number>] [--compress <level>] [--use_clone] [--expand_to_24bit] [--quality <quality>] [--source <x11|shm|v4l>] [--file <file>] [--gui [yes|no]] [--audio [yes|no]] [--audio_in <device|->] [--audio_bits <bit_rate>] [--audio_rate <sample_rate>] [--audio_channels <channels>]

INTRODUCTION

xvidcap is a tool that captures movement on a selected area of an X11 screen to files. These files may be a number of individual image files (one for each frame captured) or video files encoded on-the-fly through FFMPEG’s libavcodec. You select between the two alternatives by specifying a corresponding output filename, ref. "--file" in the OPTIONS paragraph. On-the-fly encoding will also allow you to record audio along with the video. Recording to individual frames may be more convenient if you wish to preprocess the images before encoding, or if you need other video codecs xvidcap does not support. Individual images can later be encoded with tools like ffmpeg, mencoder, or transcode.

Please note that FFMPEG follows it’s own license, see: http://www.ffmpeg.org

The utility comes with two alternative GUIs: An Xt-based front-end (xvidcap) and a GTK2-based one (gvidcap). gvidcap is maintained somewhat better, esp. in terms of user-friendliness. However, xvidcap should always provide the same basic functionality. Keep in mind that GUI layouts and behaviours will differ.

The colormap which the program uses depends on the way you select an area to capture (see GUI CONTROLS). The default colormap is the colormap of the root window. If you select a window or just parts of a window with the select button the colormap of this window is used. Not all bit depths and visuals are supported, which means you might not get the correct colors.

For help, RFEs, or bug-reports please turn to http://www.sourceforge.net/projects/xvidcap/ .

OPTIONS

-v
Runs xvidcap in verbose mode. This will provide more information about user settings, input and output codecs, performance and so forth. Messages like ’missing XX milli secs ..’ while capturing mean you are missing frames due to lack of ressources (dropped frames). This results in a video that will play back too quickly. Note though, that verbose mode slows down xvidcap and may actually be causing frame drops. Try running without verbose mode and watch the "frame drop monitor" in gvidcap, ref. GUI CONTROLS below.

--fps <fps>
Specifies the number of frames to be captured per second. This parameter accepts floating point values, which allows for very low capture rates like e.g. ’--fps 0.5’ to record one frame every 2 seconds. Note that at the moment xvidcap does not force valid frame rates for the given output format, hence your playback timing may be messed up if you don’t select valid frame rates. For MPEG1 valid frame rates are: 23.976 24 25 29.97 30 50 59.94 60. MPEG4 can use frame rates from 7.5 to 30.

--time <seconds>
Specifies the maximum time to capture, if 0 xvidcap will capture until stopped interactively. For any other value any recording session will stop automatically after the specified number of seconds. This parameter conflicts with ’--frames’ (see below). If both are given on the command line, the one given last beats the first. The program accepts floating point values, e.g. ’2.5’.

--frames <number of frames>
Stops capturing after the specified number of frames are captured. If zero xvidcap will capture until stopped interactively, ref. ’--time’ above.

--continue [yes|no]
When on-the-fly encoding is selected, maximum recording time or frames are specified, and the output filename is incrementable (see ’--file’ below), xvidcap can automatically continue to capture to a new file when one exceeds the maximum recording time or number of frames. This will change the behaviour of ’--time’ and ’--frames’ because recording will not stop at the specified limits, but rather start a new video file. Again, this is disabled for capturing to individual files. If no argument to ’--continue’ is given, xvidcap assumes ’yes’.

--cap_geometry <WIDTHxHEIGHT+XPOSITION+YPOSITION>
Override the default width and height of the capture frame. Use the ’Select’ button to adjust the frame size to an given window. The minimum size is 20 x 20.

--start_no <start_number>
Defines the start point which should be used for numbering files. The default value is ’0’.This parameter is used for capturing individual frames only. The numbering of movie files atm. always starts at zero.

--compress <compression level>
Define the compression level used for storing individual frames. This is used as the level of native compression with image formats that support it (PNG). Image formats that do not support native compression will be compressed using zlib, if the specified compression level is higher than zero. In this case a suffix of ’.gz’ is automatically appended to all file names. Valid values are 0 - 9 where zero disables the feature. WARNING: This will slow down capturing and not unlikely cause frame drops.

--use_clone
Use the CLON chunk for MNG files if needed. This will slow down capturing, but saves bytes if two or more identicaly images are saved. This is not recommended if you capture a TV/Video clip, because there will never be two consecutive images which have the same CRC32 checksum.

--expand_to_24bit
This will work only for PPM files. It expands 15 and 16bpp images to 24bpp. It should also work for TrueColor/8bpp images, but this is still not tested. Use this option if you want to make input files for the Berkeley MPEG encoder.

--quality <quality>
Specify a value between 0 and 100 to define the quality setting for JPEG output or online encoding. The default value is 75.

--source <x11|shm>
Enable or disable the usage of the X11 shared memory extension. For shm support both client and server have to run on the same host.

--file <file>
File name or pattern to use for saving the captured frames. A file name pattern contains printf()-like formatting information to make the file name incrementable, e.g. --file "out-%02d.xwd". This is necessary for capturing to individual frames. xvidcap will replace the first printf() format string with the number of the frame currently captured. Thus, in the example given, it will write to files out-00.xwd, out-01.xwd, out02.xwd, etc. Here, the frame number to start with can be specified with ’--start_no’, see above.
On-the-fly encoding does not need a file pattern. If you specify one nonetheless like --file "mymovie-%d.mpeg", xvidcap will replace the format string with the number of the current recording session (always starting the count from zero). This will enable you to manually in-/decrement that number, or use the autocontinue feature (ref. ’--continue’ above) effectively creating a series of video files.
Output file types are selected based on the extension specified for the filename. The default file type is "X Window Dump". Valid extensions are:
Portable Anymap: .pnm .pgm .ppm
Multiple Network Graphics: .mng
Portable Network Graphics: .png
Joint Picture Expert Group: .jpg .jpeg
X Window Dump: .xwd
AVI video: .avi .mpeg .mpg
ASF video stream: .asf
Flash Video File: .flv .flv1
Support for PNM and PNG is experimental and only Direct- and TrueColor 15bpp, 16bpp, and 24bpp, or PseudoColor 8bpp seems to work today. Selecting either of AVI, ASF or Flash Video enables on-the-fly encoding through FFMPEG. You can specify the codec to use through a ’codec’ parameter in the config file, cf. below.

--gui [yes|no]
Start xvidcap with or without control GUI overriding what’s in the preferences file. Running without GUI will automatically start a single capture session. It can be stopped by either specifying one of --time or --frames, or by hitting CTRL-C.

AUDIO SUPPORT

On systems where FFMPEG supports audio capture, you will be able to add an audio stream to an encoded video (i.e. using an *.mpeg extension). Audio streams can either be captured from a compatible audio device (e.g. /dev/dsp ) or from stdin (cf. --audio_in below). Audio can be enabled by specifying any of the audio parameters below on the command line or enabling audio support in gvidcap’s preferences dialog. Once audio support has been saved to the preferences file, it will be activated automatically next time xvidcap starts, unless overridden.

--audio [yes|no]
Enable or disable audio capture using default parameters or those saved to the preferences file.

--audio_in <device|->
Capture audio from the specified device or from stdin. The latter allows for dubbing a captured video using a command line like the following. The default is /dev/dsp.

cat some.mp3 | xvidcap --audio_in -

--audio_rate <sample_rate>
Set the desired sample rate. The default is 22050 Hz. Note that when using stdin input the input file will be resampled as needed.

--audio_bits <bit_rate>
Set the desired bit rate. The default is 32000 bit. Note that when using stdin input the input file will be resampled as needed.

--audio_channels <number_of_channels>
Set the desired number of channels. The default is 1 for mono. Any value above 2 is probably only useful with stdin input and a 5-channel AC audio input file.

GUI CONTROLS

´File’ menu: This menu contains entries for quitting, help and options. If you quit the program all open files will be closed. (gvidcap pops up the menu at a right click on the ’Reset Frame Counter’ button. It has an additional menu-item to save your preferences and an autocontinue toggle. Also, the ’Make Animated GIF’ function has been moved to this menu in gvidcap.)

´Make Video’ menu entry: Select this to execute a program specified in the ’mkVideoCommand’ resource. It could be used to start e.g. a MPEG encoder or similar program. The following parameters are passed to the executed application:
- %s: file name of the frames, e.g. ’cap-%04d.xwd’
- %d: number of the first saved frame
- %d: number of the last saved frame
- %d: width of the frame
- %d: height of the frame
- %f: frames per second
- %d: time per frame (ms, TPF = 1000 / FPS)
You can omit parameters from the right to the left but not in the other direction. This will work ’ppm2mpeg %s %d %d’ but this will not ’ppm2mpeg %d %d %f’. Note that you might not need the ppm2mpeg or in fact the mkVideo feature at all, if your version has on-line MPEG encoding built in (again: Try a file extension of *.mpeg). Also, gvidcap does not honour X resources. Thus the ’mkVideoCommand’ is not user-definable for gvidcap in this release.

´Reset frame counter’ button: When capturing to individual frames press this button to reset the internal picture number to the start number (default is ’0’). This will make xvidcap overwrite previously captured individual frames. The file name which is displayed in the label is the file name of the next frame which will be saved. While capturing the label is only updated if there is enough time to do it. When capturing to movie files, this button resets the count of the recording sessions, not the frames. (gvidcap: right-click to popup the file menu.)

´Stop recording’ button: Stop a running recording session.

´Pause recording’ button: Use this to halt a recording session until you release manually release this button again.

´Start recording’ button: Start to capture from the selected area.

´Capture single frame’ button: During a paused recording session this button captures a single frame per mouse click.

´Move cursor back one frame’ and ’Move cursor to next frame’ buttons: The cursor is the picture number of the next frame to capture.

´Attach/Detach selection frame’ button: This toggle controls whether or not the capture selection frame moves with the control panel.

´Select rectangular area’ button: Click this icon to get a cross-hair cursor for selecting the area you want to capture. Clicking the cursor on an application will fit the selection frame to that application window’s size and position. You can also click and drag the cursor to select a custom rectangle.

´Make animated GIF’ button: This executes an animation program as defined in the X resources and plays the captured frames. Note this is not configurable in gvidcap. (gvidcap: This can be found in the file menu, right-click on the ’Reset Frame Count’ button.)

´Edit current frame’ button: Executes an image edit program as defined in the X resouces. Note this is not configurable in gvidcap.

X RESOURCES

You can set some default values with the X resources (for example in your private ’.Xdefaults’ file). The syntax is ’xvidcap*<resource>: <value>’. Note that gvidcap as a GTK2 application completely ignores X resources. Some configuration options can also be saved to a private options file $HOME/.xvidcap.scf.

fps
Float, see OPTIONS
options file example -> fps: 15.000000

file
String, see OPTIONS
options file example -> file: test.mpeg

capWidth and capHeight
Integers, see the --cap_geometry option

time
Float, see OPTIONS
options file example -> max_time: 0.000000

frames
Integer, see OPTIONS
options file example -> max_frames: 0

compress
Integer, see OPTIONS
options file example -> compression: 0

startNo
Integer, see OPTIONS

source
String, from where to read the data: x11, shm; see also OPTIONS

quality
Integer, see OPTIONS
options file example -> quality: 75

animateCommand
String, animation command which should be executed if the ’Animation’ button is pressed. The default value is ’animate "%s" &’. The ’%s’ is substituted by a file pattern, e.g. if the file name is ’cap-%03d.xwd’ it is called with ’cap-???.xwd’. You can also add a ’%d’, but this must be after(!) the ’%s’. The ’%d’ will be filled with the seconds per frame in milli seconds.

editCommand
String, edit command, which should be executed if the ’Edit’ button is pressed. The default value is ’display "%s" %’. If you don’t have the ’display’ program from the ImageMagick package you can also use ’xv’ or ’gimp’ or what ever you want.

mkVideoCommand
String, see ’mkVideo’ button, default: ’ppm2mpeg %s %d %d %d %d %f %d &’.

helpCommand
String, help command which should be executed if the ’Help’ button is pressed. The default value is ’MANPATH=/usr/local/X11/man rxvt -e man xvidcap &’.

useShm
Integer, use shared memory, don’t change manually (not yet implemented for Xresources)
options file example -> shm: 16

codec
String, video codec used by ffmpeg, (not yet implemented for Xresources). Supported values atm are ’MPEG1’, ’MJPEG’, ’MS_DIV2’, ’MS_DIV3’, ’MPEG4’, and ’FLASH_VIDEO’. Selecting MS_DIV2 will write an AVI file compatible with Windows Media Player, MS_DIV3 an ASF, FLASH_VIDEO will NOT write an *.swf file but rather a streamable *.flv file that can be imported into an *.swf file with Macromedia tools.
options file example -> codec: MPEG4

mouse_wanted
Integer, defines whether or not to capture mouse-pointer movement using a pseudo mouse-pointer. Supported values are ’0’ to disable the feature, ’1’ to capture a white mouse-pointer, or ’2’ to capture a black mouse-pointer.
options file example -> mouse_wanted: 2 (not yet implemented for Xresources)

audio
Integer (0|1), toggles audio capture off|on.

audio_in
String, see AUDIO SUPPORT above.

audio_rate
Integer, see AUDIO SUPPORT above.

audio_bits
Integer, see AUDIO SUPPORT above.

audio_channels
Integer, see AUDIO SUPPORT above.

continue
Integer (0|1), toggles autocontinue off|on.

AUTHOR

Rasca Gmelch, Karl H. Beckers
http://www.sourceforge.net/projects/xvidcap/
XVidCap is published under the GNU General Public License.