captureSystemAudio | Capture system audio

captureSystemAudio is a JavaScript library. captureSystemAudio has no bugs, it has no vulnerabilities and it has low support. You can download it from GitHub.

Capture system audio ("What-U-Hear"). To be able to record from a monitor source (a.k.a. "What-U-Hear", "Stereo Mix"), use pactl list to find out the name of the source in PulseAudio (e.g. alsa_output.pci-0000_00_1b.0.analog-stereo.monitor). Based on the results of testing default implementation and experiments with different approaches to get access to the device within the scope of API's shipped with the browser it is not possible to select Monitor of at Chromium at Linux, which is not exposed at getUserMedia() UI prompt or at enumerateDevices() after permission to capture audio is granted, without manually setting the device to Monitor of during recording a MediaStream from getUserMedia({audio: true}) at PulseAudio sound settings GUI Recording tab. Once that user action is performed outside of the browser at the OS the setting becomes persistent where subsequent calls to getUserMedia({audio: true}). To capture microphone input after manually setting the Monitor of at PulseAudio sound settings GUI the user must perform the procedure in reverse by recording a MediaStream and setting the device back to the default Built-in during capture of a MediaStream from getUserMedia({audio: true}). Firefox supports selection of Monitor of at getUserMedia() at Linux at the UI prompt by selecting the device from enumerateDevices() after permission is granted for media capture at first getUserMedia() and getUserMedia() is executed a second time with the deviceId of Monitor of from MediaDeviceInfo object constraint set {audio:{deviceId:{exact:device.deviceId}}}. Firefox and Chromium do not support system or application capture of system audio at getDisplayMedia({video: true, audio: true}) at Linux. Chrome on Windows evidently does to support the user selecting audio capture at getDisplayMedia({video: true, audio: true}) UI prompt. getUserMedia() and getDisplayMedia() specifications do not explicitly state the user agent "MUST" provide the user with the option to capture application or system audio. From Screen Capture In the case of audio, the user agent MAY present the end-user with audio sources to share. Which choices are available to choose from is up to the user agent, and the audio source(s) are not necessarily the same as the video source(s). An audio source may be a particular application, window, browser, the entire system audio or any combination thereof. Unlike mediadevices.getUserMedia() with regards to audio+video, the user agent is allowed not to return audio even if the audio constraint is present. If the user agent knows no audio will be shared for the lifetime of the stream it MUST NOT include an audio track in the resulting stream. The user agent MAY accept a request for audio and video by only returning a video track in the resulting stream, or it MAY accept the request by returning both an audio track and a video track in the resulting stream. The user agent MUST reject audio-only requests. "MAY" being the key term in the language at "the user agent MAY", indicating that implementation of capturing audio from "a particular application, window, browser, the entire system audio or any combination thereof" is solely an individual choice of the "user agent" to implement or not and thus can be considered null and void as to being a requirement for conformance with the specification if the "user agent" decides to omit audio capture from the implementation of the specification. Audio capture is described in broad context as to potential applicable coverage in general in the Screen Capture specification where that same description of potential coverage can be narrowly interpreted by the term "MAY" to mean not required to implement for conformance and thus not applicable solely at the "user agent" discretion. Specify and implement web compatible system audio capture. The origin of and primary requirement is to capture output of window.speechSynthesis.speak(). The code can also be used to capture playback of media at native applications where the container and codec being played are not be supported at the browser by default, not supported as a video document when directly navigated to, or output from a native application supporting features not implemented at the browser, for example, mpv output.amr sound.caf, ffplay blade_runner.mkv, paplay output.wav, espeak-ng -m 'A paragraph.A sentence'. Open local files watched by inotifywait from inotify-tools to capture system audio monitor device at Linux, write output to a local file, stop system audio capture, get the resulting local file in the browser. opus-tools, mkvtoolnix, both used by default to convert WAV to Opus and write Opus to WebM container to decrease resulting file size and encoded and write track to Matroska, WebM, or other media container supported at the system. opus-tools, mkvtoolnix are included in the code by default to reduce resulting file size of captured stream by converting to Opus codec from audio from WAV. ffmpeg is used to write WebM file to local filesystem piped from parec and opusenc in "real-time", where MediaSource can be used to stream the captured audio in "real-time" (ffmpeg does not write WebM to local filesystem until 669 bytes are accumulated). Create a local folder in /home/user/localscripts containing the files in this repository, run the command. to start inotifywait watching two .txt files in the directory for open events and launches Chromium. To start system audio capture at the browser open the local file captureSystemAudio.txt, to stop capture by open the local file stopSystemAudioCapture.txt, where each file contains one space character, then get the captured audio from local filesystem using or where implemented Native File System showDirectoryPicker(). Adjust shell script captureSystemAudio.sh to pipe opusenc to ffmpeg to write file while reading file at browser. at JavaScript use HTMLMediaElement, MediaSource to capture timeSlice seconds, minutes, hours, or, given unlimited computational resources, an infinite stream of system audio output. Where it is currently not possible to select "Monitor of Built-in Audio Analog Stereo" at Chromium implementation of media capture by default, launch pavucontrol Recording tab using pavucontrol -t 2 after getUserMedia({audio: true}) for capability to change the audio device being captured dynamically, e.g., from default microphone "Built-in Audio Analog Stereo" to "Monitor of Built-in Audio Analog Stereo" ("What-U-Hear"). To launch pavucontrol or pavucontrol-qt using Native Messaging open a terminal, cd to native_messaging/host folder, open launch_pavucontrol.json and substitute aboslute path to launch_pavucontrol.sh for "HOST_PATH", then run the commands. navigate to chrome://extensions, set Developer mode to on, click Load unpacked and select app folder. Pin the app badge to the extension toolbar (it might be necessary to enable Extentions Toolbar Menu at chrome://flags/#extensions-toolbar-menu). When the browser action of clicking the icon occurs pavucontrol (or, if installed and set in launch_pavucontrol.sh, pavucontrol-qt) will be launched. When no audio device is being captured the Recording tab will be empty. When navigator.getUserMedia({audio: true}) is executed a list populate the Recording tab where the user can check a device that will be dynamically set as the device being captured by getUserMedia({audio: true}), using pavucontrol-qt. Set permissions for .js, .sh files in host folder to executable. Set "HOST_PATH" in host/native_messaging_file_stream.json to absolute path to host/native_messaging_file_stream.js. Copy native_messaging_file_stream.json to ~/.config/chromium/NativeMessagingHosts. Click Load unpacked at chrome://extensions, select app folder. To set permission to communicate with Native Messaging on a web page run app/set_externally_connectable.js at console, select app directory to update app/manifest.json, then reload background.js at extensions tab GUI or using chrom.runtime.reload() at DevTools chrome-extension URL. Select app directory at Native File System prompts for read and write access to local filesystem where raw PCM of system audio output is written to a file using parec while reading the file during the write using Native File System, storing the data in shared memory, parsing input data in AudioWorklet connected to MediaStreamTrack outputting the captured system audio. This article Virtual microphone using GStreamer and PulseAudio describes a workaround Chrome and Chromium browsers' refusal to list or capture monitor devices on Linux. Remap source While the null sink automatically includes a "monitor" source, many programs know to exclude monitors when listing microphones. To work around that, the module-remap-source module lets us clone that source to another one not labeled as being a monitor: pactl load-module module-remap-source \ master=virtmic.monitor source_name=virtmic \ source_properties=device.description=Virtual_Microphone. and then at Chromium and Chrome run. to first get permission to read labels of devices, find the device we want to capture, capture the virtual microphone device, in this case a monitor device, see When no microphone input devices are connected to the machine the remapped monitor device will be the default device "Virtual_Microphone" when navigator.mediaDevices.getUserMedia({audio: true}) is executed the first time, negating the need to call MediaStreamTrack.stop() to stop capture of a microphone device just to get device access permission, then use navigator.mediaDevices.enumerateDevices() to get deviceId of monitor device, create a constraints object {deviceId: {exact: device.deviceId}} and call navigator.mediaDevices.getUserMedia({audio: constraints}) a second time. To set the default source programmatically to the virtual microphone "virtmic" set-default-source can be utilized. if running, closing then restarting Chrome, Chromium, or Firefox, the device selected by navigator,mediaDevices.getUserMedia({audio: true}), unless changed by selection or other setting, will be the remapped monitor device "Virtual_Microphone".