omarchycn/bin/omarchy-capture-screenrecording

#!/bin/bash

# omarchy:summary=Start or stop screen recording
# omarchy:group=capture
# omarchy:args=[--fullscreen] [--with-desktop-audio] [--with-microphone-audio] [--with-webcam] [--webcam-device=<device>] [--webcam-size=<small|medium|large>] [--resolution=<size>] [--stop-recording]
# omarchy:examples=omarchy screenrecord | omarchy capture screenrecording --with-desktop-audio
# omarchy:aliases=omarchy screenrecord
#
# Env: OMARCHY_SCREENRECORD_USE_PORTAL=true skips the built-in slurp picker and
# uses gpu-screen-recorder's xdg-desktop-portal capture backend instead. The
# portal backend was originally added (PR #3401) for HDR-aware capture, support
# for monitors driven by external GPUs, and window capture — enable it if any
# of those matter to you. Off by default because the portal path can fail EGL
# DMA-BUF modifier import on some configurations, leaving recording unable to
# start.
#
# Env: OMARCHY_SCREENRECORD_DEBUG=true appends gpu-screen-recorder's stderr (and
# the picker target it was launched with) to /tmp/omarchy-screenrecord.log so
# users can attach a log when reporting capture failures.

[[ -f ~/.config/user-dirs.dirs ]] && source ~/.config/user-dirs.dirs
OUTPUT_DIR="${OMARCHY_SCREENRECORD_DIR:-${XDG_VIDEOS_DIR:-$HOME/Videos}}"

if [[ ! -d $OUTPUT_DIR ]]; then
  omarchy-notification-send -u critical -t 3000 "Screen recording directory does not exist: $OUTPUT_DIR"
  exit 1
fi

DESKTOP_AUDIO="false"
MICROPHONE_AUDIO="false"
WEBCAM="false"
WEBCAM_DEVICE=""
WEBCAM_SIZE="medium"
RESOLUTION=""
FULLSCREEN="false"
STOP_RECORDING="false"
RECORDING_FILE="/tmp/omarchy-screenrecord-filename"
REGION_FILE="${XDG_RUNTIME_DIR:-/tmp}/omarchy-screenrecord-region"
LOG_FILE=$([[ ${OMARCHY_SCREENRECORD_DEBUG:-false} == "true" ]] && echo "/tmp/omarchy-screenrecord.log" || echo "/dev/null")

for arg in "$@"; do
  case "$arg" in
  --with-desktop-audio) DESKTOP_AUDIO="true" ;;
  --with-microphone-audio) MICROPHONE_AUDIO="true" ;;
  --with-webcam) WEBCAM="true" ;;
  --webcam-device=*) WEBCAM_DEVICE="${arg#*=}" ;;
  --webcam-size=*) WEBCAM_SIZE="${arg#*=}" ;;
  --resolution=*) RESOLUTION="${arg#*=}" ;;
  --fullscreen) FULLSCREEN="true" ;;
  --stop-recording) STOP_RECORDING="true" ;;
  esac
done

case $WEBCAM_SIZE in
small | medium | large) ;;
*)
  echo "Invalid webcam size: $WEBCAM_SIZE (expected small, medium, or large)" >&2
  exit 1
  ;;
esac

start_webcam_overlay() {
  cleanup_webcam

  # Auto-detect first available webcam if none specified
  if [[ -z $WEBCAM_DEVICE ]]; then
    WEBCAM_DEVICE=$(omarchy-capture-webcam-list | sed -n '1s/[[:space:]].*//p')
    if [[ -z $WEBCAM_DEVICE ]]; then
      omarchy-notification-send -u critical -t 3000 "No webcam devices found"
      return 1
    fi
  fi

  # Try preferred 16:9 resolutions in order, use first available
  local preferred_resolutions=("640x360" "1280x720" "1920x1080")
  local capture_options="framerate=30"
  local available_formats=$(v4l2-ctl --list-formats-ext -d "$WEBCAM_DEVICE" 2>/dev/null)

  for resolution in "${preferred_resolutions[@]}"; do
    if echo "$available_formats" | grep -q "$resolution"; then
      capture_options="video_size=$resolution,$capture_options"
      break
    fi
  done

  mpv "av://v4l2:$WEBCAM_DEVICE" \
    --profile=low-latency --untimed --no-cache \
    --demuxer-lavf-o="$capture_options" \
    '--vf=lavfi=[crop=ih*8/9:ih]' \
    --title="WebcamOverlay" --wayland-app-id="WebcamOverlay-$WEBCAM_SIZE" \
    --no-border --no-audio --no-osc --osd-level=0 \
    --really-quiet &>/dev/null &

  # The move has to settle before gpu-screen-recorder starts, or the camera is
  # recorded sliding into its corner. Waiting for the map is what the blind
  # second was partly guessing at, so the remainder is trimmed to hold the
  # pre-capture delay where it was: starting later costs the first words spoken.
  local waited=0
  while ((waited < 40)) && ! hyprctl clients -j | jq -e 'any(.[]; .title == "WebcamOverlay")' >/dev/null 2>&1; do
    sleep 0.05
    ((waited++))
  done

  [[ ${1:-} == region:* ]] && echo "${1#region:}" >"$REGION_FILE"
  omarchy-capture-webcam-resize "$WEBCAM_SIZE"

  sleep 0.6
}

cleanup_webcam() {
  pkill -f "WebcamOverlay" 2>/dev/null
  rm -f "$REGION_FILE"
}

default_resolution() {
  local width height
  read -r width height < <(hyprctl monitors -j | jq -r '.[] | select(.focused == true) | "\(.width) \(.height)"')
  if ((width > 3840 || height > 2160)); then
    echo "3840x2160"
  else
    echo "0x0"
  fi
}

# Echoes "monitor:NAME" when the selection matches an entire monitor (prefer
# -w <monitor> over a region capture — same kms backend, but no scaling math
# and full native res), otherwise "region:WxH+X+Y". Returns non-zero if the
# user cancelled the picker.
select_capture_target() {
  local target
  target=$(omarchy-capture-region smart --match-monitor) || return 1

  if [[ $target == monitor:* ]]; then
    echo "$target"
    return
  fi

  [[ $target =~ ^(-?[0-9]+),(-?[0-9]+)[[:space:]]([0-9]+)x([0-9]+)$ ]] || return 1

  # gpu-screen-recorder wants region geometry in the compositor's logical
  # coordinate space — same space slurp returns — so pass the values through
  # untouched. (gsr scales to physical pixels itself based on the monitor.)
  echo "region:${BASH_REMATCH[3]}x${BASH_REMATCH[4]}+${BASH_REMATCH[1]}+${BASH_REMATCH[2]}"
}

start_screenrecording() {
  local capture_args=()
  local target

  # Opt-in path for HDR, external-GPU monitors, and window capture (all things
  # the portal backend supports and the kms backend doesn't). Default flow uses
  # slurp + the kms backend, which avoids the EGL DMA-BUF modifier import
  # failures the portal path can hit on some configurations.
  if [[ $FULLSCREEN == "true" ]]; then
    target="monitor:$(omarchy-hyprland-monitor-focused)"
    capture_args=(-w "${target#monitor:}" -s "${RESOLUTION:-$(default_resolution)}")
  elif [[ ${OMARCHY_SCREENRECORD_USE_PORTAL:-false} == "true" ]]; then
    target="portal"
    capture_args=(-w portal -s "${RESOLUTION:-$(default_resolution)}")
  else
    target=$(select_capture_target) || return 1

    case $target in
    monitor:*)
      capture_args=(-w "${target#monitor:}" -s "${RESOLUTION:-$(default_resolution)}")
      ;;
    region:*)
      capture_args=(-w "${target#region:}")
      [[ -n $RESOLUTION ]] && capture_args+=(-s "$RESOLUTION")
      ;;
    esac
  fi

  [[ $WEBCAM == "true" ]] && start_webcam_overlay "$target"

  local filename="$OUTPUT_DIR/screenrecording-$(date +'%Y-%m-%d_%H-%M-%S').mp4"
  local audio_devices=""
  local audio_args=()

  [[ $DESKTOP_AUDIO == "true" ]] && audio_devices+="default_output"

  if [[ $MICROPHONE_AUDIO == "true" ]]; then
    # Merge audio tracks into one - separate tracks only play one at a time in most players
    [[ -n $audio_devices ]] && audio_devices+="|"
    audio_devices+="default_input"
  fi

  [[ -n $audio_devices ]] && audio_args+=(-a "$audio_devices" -ac aac)

  echo "===== $(date '+%F %T') args: $* target: $target =====" >>"$LOG_FILE"
  gpu-screen-recorder "${capture_args[@]}" -k auto -f 60 -fm cfr -fallback-cpu-encoding yes -o "$filename" "${audio_args[@]}" 2>>"$LOG_FILE" &
  local pid=$!

  while kill -0 $pid 2>/dev/null && [[ ! -f $filename ]]; do
    sleep 0.2
  done

  if kill -0 $pid 2>/dev/null; then
    echo "$filename" >"$RECORDING_FILE"
    toggle_screenrecording_indicator
  fi
}

stop_screenrecording() {
  pkill -SIGINT -f "^gpu-screen-recorder" # SIGINT required to save video properly

  # Wait a maximum of 5 seconds to finish before hard killing
  local count=0
  while pgrep -f "^gpu-screen-recorder" >/dev/null && ((count < 50)); do
    sleep 0.1
    count=$((count + 1))
  done

  toggle_screenrecording_indicator
  cleanup_webcam

  if pgrep -f "^gpu-screen-recorder" >/dev/null; then
    pkill -9 -f "^gpu-screen-recorder"
    omarchy-notification-send -u critical -t 5000 "Screen recording error" "Recording process had to be force-killed. Video may be corrupted."
  else
    finalize_recording
    local filename=$(cat "$RECORDING_FILE" 2>/dev/null)
    echo "$filename"
    local preview="${filename%.mp4}-preview.png"

    # Generate a preview thumbnail from the first frame
    ffmpeg -y -i "$filename" -ss 00:00:00.1 -vframes 1 -q:v 2 "$preview" -loglevel quiet 2>/dev/null

    omarchy-notification-send "Screen recording saved" "Open with Super + Alt + , (or click this)" \
      -t 10000 --image "${preview:-$filename}" \
      --exec "$(printf 'mpv %q' "$filename")"

    # The shell loads the thumbnail into memory when the toast appears and never
    # re-reads the file, so the preview only has to outlive that load -- not the
    # toast. Clear it out of the recordings directory a moment later.
    (
      sleep 2
      rm -f "$preview"
    ) &
  fi

  rm -f "$RECORDING_FILE"
}

toggle_screenrecording_indicator() {
  omarchy-shell -q omarchy.indicators refresh
}

screenrecording_active() {
  pgrep -f "^gpu-screen-recorder" >/dev/null
}

finalize_recording() {
  local latest
  latest=$(cat "$RECORDING_FILE" 2>/dev/null)
  [[ -f $latest ]] || return

  # Re-encode only when the first GOP contains discardable warmup packets — stream copy can't
  # trim those (it rewinds to the keyframe). Clean recordings stay on the fast stream-copy path.
  local video_codec=(-c:v copy)
  if ffprobe -v error -select_streams v:0 -read_intervals %+0.2 -show_entries packet=flags -of csv=p=0 "$latest" 2>/dev/null | grep -q D; then
    video_codec=(-c:v libx264 -preset veryfast -crf 20)
  fi

  # Trim the first frame, and normalize audio to -14 LUFS if present, in a single pass
  local args=(-y -ss 0.1 -i "$latest" "${video_codec[@]}")
  if ffprobe -v error -select_streams a -show_entries stream=codec_type -of csv=p=0 "$latest" 2>/dev/null | grep -q audio; then
    # Hard-mute the first 400ms to drop the PipeWire capture-open pop (a near-clipping
    # transient around 130-200ms that a gentle fade-in can't attenuate enough), then a
    # 50ms fade avoids a click at the boundary before loudnorm normalizes the rest.
    args+=(-af "volume=enable='lt(t,0.4)':volume=0,afade=t=in:st=0.4:d=0.05,loudnorm=I=-14:TP=-1.5:LRA=11")
  fi

  local processed="${latest%.mp4}-processed.mp4"
  if ffmpeg "${args[@]}" "$processed" -loglevel quiet 2>/dev/null; then
    mv "$processed" "$latest"
  else
    rm -f "$processed"
  fi
}

if screenrecording_active; then
  stop_screenrecording
elif [[ $STOP_RECORDING == "true" ]]; then
  exit 1
else
  start_screenrecording || cleanup_webcam
fi