SCRYDOSBox-XReferenceCapturesrecordAlso in MAME
record
Record an exact number of frames of the screen to an AVI video, then stop the program again.
record(frames, dest="", seconds=60.0)
Try it
Record the next 110 frames of the screen to a video named second-brick.avi.
record(frames=110, dest="./tmp/demos/tour/files/second-brick.avi", seconds=60){
"summary": "record 110 frames: ELAPSED, 109 frames at ./tmp/demos/tour/files/second-brick.avi",
"verdict": "ELAPSED",
"frame_before": 36,
"frame_after": 146,
"kind": "avi",
"frames": 110,
"began": true,
"ended": true,
"avi_frames": 109
}ELAPSED means the 110 frames ran, from frame 36 to 146, and the capture was turned off (ended). The file holds 109 frames: the first frame of this run wrote none.
The whole reply
{
"summary": "record 110 frames: ELAPSED, 109 frames at ./tmp/demos/tour/files/second-brick.avi",
"status": "ok",
"verdict": "ELAPSED",
"polls": 110,
"pc": "<volatile>",
"cpu_mode": "real",
"frame": 146,
"regs": "<volatile>",
"watches": [
{
"ref": "ball_x",
"by": "symbol",
"mode": "show",
"length": 8,
"fmt": "{x:i16,y:i16,dx:i16,dy:i16}",
"state": "evaluated",
"value": {"x": 182, "y": 88, "dx": -2, "dy": 2},
"bytes": "<volatile>"
},
{
"ref": "score",
"by": "symbol",
"mode": "count",
"length": 2,
"fmt": "u16",
"tally_id": "1",
"addr": "08DF:00000615",
"installed": true,
"state": "tallied",
"total": 1,
"since_last_stop": 1,
"overflow_values": 0,
"overflow_pcs": 0
},
{
"ref": "paddle_x",
"by": "symbol",
"mode": "capture",
"length": 1,
"fmt": "u8",
"tally_id": "2",
"addr": "08DF:0000061F",
"trigger_type": "r",
"installed": true,
"state": "captured",
"total": 110,
"since_last_stop": 110,
"rows_path": "<volatile>",
"rows_kept": 110,
"overflow_rows": 0,
"rows_max": 100000
}
],
"cause": "<volatile>",
"e9": {"lines_total": 1, "truncated": false, "lines": ["Bochs port E9h: HIT 5 6"], "mark": "<volatile>"},
"stop_seq": 8,
"instruction": "<volatile>",
"emulated_before_s": "<volatile>",
"emulated_after_s": "<volatile>",
"frame_before": 36,
"frame_after": 146,
"cycles": 3595712,
"kind": "avi",
"frames": 110,
"began": true,
"ended": true,
"path": "./tmp/demos/tour/files/second-brick.avi",
"bytes": "<volatile>",
"avi_frames": 109
}Arguments
- frames integer, required
- How many frames of the emulated display to record: 2 or more.
- dest string, default
"" - An absolute
.avipath that does not exist yet. Left empty, the file goes to the session's temporary folder, deleted atclose. - seconds number (> 0, <= 600.0), default
60.0 - The time limit of the whole call, in seconds of your clock. Left out, 60.
What comes back
- path
- The AVI, once the reply has read it back as a finished file.
- avi_frames
- How many frames the file's own header says it holds.
- ended
trueonce the capture is turned off.falsewhen it could not be: there is nopath, and it records untilclose.- verdict
- How the run ended, as
run_forreports it:ELAPSEDwhen every frame ran,TIMEOUTwhensecondsran out first.
All 26 reply fields
cause(string)regpoint_hit(object)cpu_mode(string or object)cycles(integer or object)e9(object)emulated_after_s(number or object)emulated_before_s(number)enrichment(object)frame(integer or object)frame_after(integer or object)frame_before(integer or object)instruction(DosboxInstruction)pc(string or object)polls(integer)regs(object)retrace_pending(boolean)stop_seq(integer)verdict(string)watches(array)avi_frames(integer)began(boolean)bytes(integer)ended(boolean)frames(integer)kind(string)path(string or object)
What it refuses
framesunder 2, or adestthat is not a new.avipath (bad_argument).- A machine that is running (
not_stopped). - A video capture already on, as in a session started with
capture_video(no_channel). - A build whose debugger has no
CAPTUREcommand (no_channel). - A file that does not read back as a finished AVI with a frame in it (
readback_refused).
Why it works this way
The recording is counted in frames of the emulated display, one per vertical retrace, so no time of your clock goes into the file. The debugger's CAPTURE command turns the video on at this stop and off at the stop the run reaches.
No window or menu is involved, so a session with no window records as one with a window does. A video already on is refused, never ended by a second switch.
One frame is written per frame run, but a run can write none for its first. The run from the stop start leaves is one: the display has drawn nothing yet.
More on why
The capture is turned off whatever the run did, a cancelled call included. One that could not be turned off answers ended: false.
The reply is also a stop: it carries verdict and the fields run_for returns, with the readings taken at each stop.
Related tools
The contract your client receives
Record exactly `frames` retraces of the screen to an AVI, then stop. Needs the machine
STOPPED and a build whose debugger has `CAPTURE`; headless or windowed.
`dest`: an absolute `.avi` path not there yet; empty lands in the session's scratch, gone at `close`, and the reply names it either way. `frames` is 2 or more. `seconds` bounds the wait for the recording. The AVI is read before its path is given.
One frame is written per retrace, but none for the first from the stop `start` leaves.
A run cut at that bound is released before the reply: when the release answers, the reply says `ended: true` and names the AVI it read back, and a file that does not read as finished is INVALID. A capture that could not be turned off is `ended: false`, no `path`, and
records until `close`.
Refused: `frames` under 2, a video capture already on (`capture_video`), a build with no
`CAPTURE`, a running machine.
Reply, once answered: `verdict`, `frames`, `path`, `began`, `ended`, `bytes`, `avi_frames`, `kind`, `note`, the run's fields as `run_for` gives
them, the stop: `pc` (`SSSS:OOOOOOOO`), `cpu_mode`, `stop_seq`, `frame`, `instruction`, `regs`, `watches`, `cause`, `regpoint_hit` (a registerpoint stopped it) and `e9`.
Reference: REFERENCE.md, Input and the screen.