X Video Extension Protocol Description Version 2 25-JUL-91 David Carver Digital Equipment Corporation Workstation Engineering/Project Athena - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Copyright 1991 by Digital Equipment Corporation, Maynard, Massachusetts, and the Massachusetts Institute of Technology, Cambridge, Massachusetts. All Rights Reserved Permission to use, copy, modify, and distribute this software and its documentation for any purpose and without fee is hereby granted, provided that the above copyright notice appear in all copies and that both that copyright notice and this permission notice appear in supporting documentation, and that the names of Digital or MIT not be used in advertising or publicity pertaining to distribution of the software without specific, written prior permission. DIGITAL DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO EVENT SHALL DIGITAL BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Preface ------- The following is an outline for an X video extension protocol. It is preliminary and subject to change. My goal in writing this was to fix some the shortcomings of existing overly simplistic extensions while avoiding pitfalls in an overly complex extension. Your feedback is desired, and since the major design directions have been stable for some time, feel free to hammer on the details of the protocol. When you receive a revision of the document, refer to the changes and issues sections to guide your review and analysis. Acknowledgements --------------- The following people have made major contributions to the design of the Xv protocol: Branko Gerovac (DEC/Corporate Research) Russ Sasnett (GTE/Project Athena) Ralph Swick (DEC/Project Athena) Many ideas and approaches in Xv were the product of discussions with several groups, including Project Athena's Visual Computing Group The MIT X Consortium The MIT Media Lab's Interactive Cinema Group Changes ------- From version 1.3 to 2.0 -- Changed SetPortControl and GetPortControl to GetPortAttribute and SetPortAttribute. -- Changed QueryBestSize -- Simplified SelectVideoNotify and SelectPortNotify requests. -- Changed the way SetPortControl and GetPortControl works. -- Added a QueryExtension request to return the version and revision information of the extension. -- Changed the name of the QueryVideo request to QueryAdaptors; Removed the list of encodings from QueryVideo and added a QueryEncodings request. -- Added a PortNotify event that notifies interested clients that a port control has been changed. -- Added SelectPortNotify request to select for PortNotify events. -- The XvInterruped reason has been replaced by two new reasons: one for when video is preempted by another video request and one for when video is terminated because of hard transmission or reception errors. -- Changed the wording of the QueryBestSize request. Added issue about whether or not returned sizes should maintain the requested aspect ratio. Introduction ------------ Video technology is moving very quickly. Standards for processing high resolution video are currently a hot topic of discussion internationally, and it will soon be possible to process video entirely within the digital domain. The Xv extension, however, does not attempt to address issues of digital video. Its purpose is to provide a mechanism for support of current and near term interactive video technology. It is somewhat ironic that Xv contains nothing particularly innovative. It takes a minimalistic approach, and without a doubt it could have been defined years ago, and with several revisions. So, the life expectancy of Xv is not long. Nevertheless, it may undergo further revision and experimentation that will help our progress towards digital video systems. One premise of the Xv extension is that the X server is not alone. A separate video server is often used to manage other aspects of video processing, though the partition between what the X server does and what a video server does is a matter of great debate. Model ----- This extension models video monitor capabilities in the X Window System. Some advanced monitors support the simultaneous display of multiple video signals (into separate windows), and that is prepresented here through the ability to display video from multiple video input adaptors into X drawables. Some monitors support multiple video encodings (mostly for internationalization purposes) either through switches or automatic detection, thus each video adaptor specifies the set of encodings it supports. The requests to display video from an adaptor into a drawable are modeled after the core PutImage request, though extended to support scaling and source clipping. Video output is also supported and is symmetric with the video input function, though fewer GC components are used. Mechanism --------- The Xv extension does the following: -- lists available video adaptors -- identifies the number of ports each adaptor supports -- describes what drawable formats each adaptor supports -- describes what video encodings each adaptor supports -- displays video from a port to a drawable -- captures video from a drawable to a port -- grabs and ungrabs ports -- sets and gets port attributes -- delivers event notification Adaptors -------- A display may have multiple video input and output adaptors. An adaptor may support multiple simultaneously active ports, and in some cases the number of ports has no fixed limit. An input port receives encoded video data and converts it to a stream of data used to update a drawable. An output port samples data from a drawable and produces a stream of encoded video data. The ADAPTORINFO structure is used to describe a video adaptor. ADAPTORINFO: [base-id: PORT num-ports: CARD16 type: SETofADAPTORTYPE formats: LISTofFORMAT name: STRING] ADAPTORTYPE: {Input, Output} FORMAT: [depth: CARD8 visual: VISUALID] The base-id field specifies the XID of the first port of the adaptor. The `num-ports' field specifies how many ports the adaptor supports. The ports of the adaptor have XIDs in the range [base-id..base-id + num-ports - 1] The type attribute determines if the adaptor can process video input, output, or input and output. The if the adaptor can process input then Input is asserted, if the adaptor can process output then Output is asserted. The drawable depths and visual types supported by the adaptor are listed in `formats'. Note: that when video is being processed for pixmaps the visual format is taken to be the visual of the first pair that matches the depth of the pixmap. The name field contains an a vendor specific string that identifies the adaptor. It should be noted that the existence of separate adaptors doesn't necessarily imply that simultaneous operation is supported. Errors ------ Port A Port error is returned if any request names a PORT that does not exist. Encoding An Encoding error is returned if any request names an ENCODINGID that does not exist. Query Requests ------------------- QueryExtension ==> version: CARD16 revision: CARD16 The QueryExtension request returns the extension version and revision numbers. QueryAdaptors win: WINDOW ==> adaptors: LISTofADAPTORINFO The QueryAdaptors request returns the video adaptor information for the screen of the specified window. Errors: {Window} QueryEncodings port: PORT ==> encodings: LISTofENCODINGINFO The QueryEncodings request returns the list of encodings supported by the port adaptor. Use the SetPortAttribute request to set which encoding a port is to process. The ENCODINGINFO record describes an encoding: ENCODINGINFO: [encoding: ENCODINGID name: STRING width, height: CARD16 rate: FRACTION] The `encoding' field identifies an encoding supported by a port. Its value is unique for a screen. Width and height specify the size of the video image and rate specifies the rate at which fields of image information are encoded. An encoding is identified by a string that names the encoding. Encoding naming conventions need to be established (i.e., something along the lines of font naming, but simpler) FRACTION [numerator, denominator: INT32] The FRACTION structure is used to specify a fractional number. Errors: {Port} Put Video Requests ------------------ PutVideo port: PORT drawable: DRAWABLE gc: GCONTEXT vid-x, vid-y: INT16 vid-w, vid-h: CARD16 drw-x, drw-y: INT16 drw-w, drw-h: CARD16 The PutVideo request writes video into a drawable. The position and size of the source rectangle is specified by vid-x, vid-y, vid-w, and vid-h. The position and size of the destination rectangle is specified by drw-x, drw-y, drw-w, drw-h. Video data is clipped to the bounds of the video encoding, scaled to the requested drawable region size (or the closest size supported), and clipped to the bounds of the drawable. If video is successfully initiated, a VideoNotify event with detail Started is generated for the drawable. If the port is already in use, its video is preempted, and if the new drawable is different than the old, a VideoNotify event with detail Preempted is generated for the old drawable. If the port is grabbed by another client, this request is ignored, and a VideoNotify event with detail Busy is generated for the drawable. If the port is not receiving a valid video signal or if the video signal is interrupted while video is active a VideoNotify event with detail HardError is generated for the drawable. GC components: subwindow-mode, clip-x-origin, clip-y-origin, clip-mask. Errors: {Match, Value, GContext, Port, Alloc} PutStill port: PORT drawable: DRAWABLE gc: GCONTEXT vid-x, vid-y: INT16 vid-w, vid-h: CARD16 drw-x, drw-y: INT16 drw-w, drw-h: CARD16 The PutStill request writes a single frame of video into a drawable. The position and size of the source rectangle is specified by vid-x, vid-y, vid-w, and vid-h. The position and size of the destination rectangle is specified by drw-x, drw-y, drw-w, drw-h. Video data is clipped to the bounds of the video encoding, scaled to the requested drawable region size (or the closest size supported) and clipped to the bounds of the drawable. If the port is grabbed by another client, this request is ignored, and a VideoNotify event with detail Busy is generated for the drawable. If the port is not receiving a valid video signal a VideoNotify event with detail HardError is generated for the drawable. GC components: subwindow-mode, clip-x-origin, clip-y-origin, clip-mask. Errors: {Match, Value, GContext, Port, Alloc} Get Video Requests ------------------ GetVideo port: PORT drawable: DRAWABLE gc: GCONTEXT vid-x, vid-y: INT16 vid-w, vid-h: CARD16 drw-x, drw-y: INT16 drw-w, drw-h: CARD16 The GetVideo request outputs video from a drawable. The position and size of the destination rectangle is specified by vid-x, vid-y, vid-w, and vid-h. The position and size of the source rectangle is specified by drw-x, drw-y, drw-w, and drw-h. Drawable data is clipped to the bounds of the drawable, scaled to the requested video region size (or the closest size supported) and clipped to the bounds of the video encoding. The contents of any region not updated with drawable data is undefined. If video is successfully initiated, a VideoNotify event with detail Started is generated for the drawable. If the port is already in use, its video is preempted, and if the new drawable is different than the old, a VideoNotify event with detail Preempted is generated for the old drawable. If the port is grabbed by another client, this request is ignored, and a VideoNotify event with detail Busy is generated for the drawable. GC components: subwindow-mode, clip-x-origin, clip-y-origin, clip-mask. Errors: {Match, Value, GContext, Port, Alloc} GetStill port: PORT drawable: DRAWABLE gc: GCONTEXT vid-x, vid-y: INT16 vid-w, vid-h: CARD16 drw-x, drw-y: INT16 drw-w, drw-h: CARD16 The GetStill request outputs video from a drawable. The position and size of the destination rectangle is specified by vid-x, vid-y, vid-w, and vid-h. The position and size of the source rectangle is specified by drw-x, drw-y, drw-w, and drw-h. Drawable data is clipped to the bounds of the drawable, scaled to the requested video region size (or the closest size supported) and clipped to the bounds of the video encoding. The contents of any region not updated with drawable data is undefined. If the still is successfully captured a VideoNotify event with detail Still is generated for the drawable. If the port is grabbed by another client, this request is ignored, and a VideoNotify event with detail Busy is generated for the drawable. GC components: subwindow-mode, clip-x-origin, clip-y-origin, clip-mask. Errors: {Match, Value, GContext, Port, Alloc} Grab Requests ------------- GrabPort port: PORT timestamp: {TIMESTAMP, CurrentTime} ==> status: {Success, AlreadyGrabbed, InvalidTime} The GrabPort request grabs a port. While a port is grabbed, only video requests from the grabbing client are permitted. If timestamp specifies a time older than the current port time, a status of InvalidTime is returned. If the port is already grabbed by another client, a status of AlreadyGrabbed is returned. Otherwise a status of Success is returned. The port time is updated when the following requests are processed: GrabPort, UngrabPort, PutVideo, PutStill, GetVideo, GetStill If the port is actively processing video for another client, the video is preempted, and an VideoNotify event with detail Preempted is generated for its drawable. Errors: {Port} UngrabPort port: PORT timestamp: {TIMESTAMP, CurrentTime} The UngrabPort request ungrabs a port. If timestamp specifies a time before the last connection request time of this port, the request is ignored. Errors: {Port} Other Requests -------------- StopVideo port: PORT drawable: DRAWABLE The StopVideo request stops active video for the specified port and drawable. If the port isn't processing video, or if it is processing video in a different drawable, the request is ignored. When video is stopped a VideoNotify event with detail Stopped is generated for the associated drawable. Errors: {Drawable, Port} SelectVideoNotify drawable: DRAWABLE onoff: BOOL The SelectVideoNotify request enables or disables VideoNotify event delivery to the requesting client. VideoNotify events are generated when video starts and stops. Errors: {Drawable} SelectPortNotify port: PORT onoff: BOOL The SelectPortNotify request enables or disables PortNotify event delivery to the requesting client. PortNotify events are generated when port attributes are changed using SetPortAttribute. Errors: {Port} QueryBestSize port: PORT motion: BOOL vid-w, vid-h: CARD16 drw-w, drw-h: CARD16 ==> actual-width, actual-height: CARD16 The QueryBestSize request returns, for the given source size and desired destination size, the closest destination size that the port adaptor supports. The returned size will be equal or smaller than the requested size if one is supported. If motion is True then the requested size is intended for use with full motion video. If motion is False, the requested size is intended for use with stills only. The retuned size is also chosen to maintain the requested aspect ratio if possible. Errors: {Port} SetPortAttribute port: PORT attribute: ATOM value: INT32 The SetPortAttribute request sets the value of a port attribute. The port attribute is identified by the attribute atom. The following strings are guaranteed to generate valid atoms using the InternAtom request. String Type ----------------------------------------------------------------- "XV_ENCODING" ENCODINGID "XV_HUE" [-1000..1000] "XV_SATURATION" [-1000..1000] "XV_BRIGHTNESS" [-1000..1000] "XV_CONTRAST" [-1000..1000] If the given attribute doesn't match an attribute supported by the port adaptor a Match error is generated. The supplied encoding must be one of the encodings listed for the adaptor, otherwise an Encoding error is generated. If the adaptor doesn't support the exact hue, saturation, brightness, and contrast levels supplied, the closest levels supported are assumed. The GetPortAttribute request can be used to query the resulting levels. When a SetPortAttribute request is processed a PortNotify event is generated for all clients that have requested port change notification using SelectPortNotify. Errors: {Port, Match, Value} GetPortAttribute port: PORT attribute: ATOM ==> value: INT32 The GetPortAttribute request returns the current value of the attribute identified by the given atom. If the given atom doesn't match an attribute supported by the adaptor a Match error is generated. Errors: {Port, Match} Events ------ VideoNotify drawable: DRAWABLE port: PORT reason: REASON time: TIMESTAMP REASON: {Started, Still, Stopped, Busy, Preempted, HardError} A VideoNotify event is generated when video activity is started, stopped, or unable to proceed in a drawable. A Started reason is generated when video starts in a drawable. A Stopped reason is generated when video is stopped in a drawable upon request. A Busy reason is generated when a put or get request cannot proceed because the port is grabbed by another client. A Preempted reason is generated when video is stopped by a conflicting request. A HardError reason is generated when the video port cannot initiate or continue processing a video request because of an underlying transmission or reception error. PortNotify port: PORT attribute: ATOM value: INT32 time: TIMESTAMP The PortNotify event is generated when a SetPortAttribute request is processed. The event is delivered to all clients that have performed a SelectPortNotify request for the port. The event contains the atom identifying the attribute that changed, and the new value of that attribute.