Topics

discord.js - Imagine an app•5mo ago

Saving an opus stream from a voice channel to .mp3/.wav/.webm

So with the help of another thread here, I managed to save an opus stream to .ogg. Now, I need to save that same opus stream to an .mp3 file instead, and i’m even more stuck than last time. Can I somehow save the audio to a .mp3 or .wav file instead? Help would be much mich appreciated…

const { opus } = require("prism-media");

//other stuff

let opusStream = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });
    const oggStream = new opus.OggLogicalBitstream({
        opusHead: new opus.OpusHead({
            channelCount: 2,
            sampleRate: 48000
        }),
        pageSizeControl: {
            maxPackets: 10
        }
    });

    let buffer = [];
    oggStream.on("data", (chunk) => {
        buffer.push(chunk);
    });

    opusStream.on("end", async () => {
        logReliably(user.globalName + " stopped speaking. " + getTimeElapsedSinceLastCall());
        oggStream.destroy()
        fs.createWriteStream('output.ogg').write(Buffer.concat(buffer))
        buffer = []
    });

    opusStream.pipe(oggStream)

const { opus } = require("prism-media");

//other stuff

let opusStream = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });
    const oggStream = new opus.OggLogicalBitstream({
        opusHead: new opus.OpusHead({
            channelCount: 2,
            sampleRate: 48000
        }),
        pageSizeControl: {
            maxPackets: 10
        }
    });

    let buffer = [];
    oggStream.on("data", (chunk) => {
        buffer.push(chunk);
    });

    opusStream.on("end", async () => {
        logReliably(user.globalName + " stopped speaking. " + getTimeElapsedSinceLastCall());
        oggStream.destroy()
        fs.createWriteStream('output.ogg').write(Buffer.concat(buffer))
        buffer = []
    });

    opusStream.pipe(oggStream)

37 Replies

d.js toolkit•5mo ago

- What's your exact discord.js npm list discord.js and node node -v version? - Not a discord.js issue? Check out #other-js-ts. - Consider reading #how-to-get-help to improve your question! - Explain what exactly your issue is. - Post the full error stack trace, not just the top part! - Show your code! - Issue solved? Press the button! - ✅ Marked as resolved by OP

ThePedroo•5mo ago

You must use ffmpeg for that You can use prism-media's ffmpeg class, which is basically a cp.spawn Then, pipe to ffmpeg, and write the output to a file Also, if you're not going to pipe with createWriteStream, use fs.writeFile

KaiFireborn•5mo ago

async function onStartedSpeaking(userId, voiceChannel) {
    let user = await getUserById(userId);

    let listener = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });
    let audioOutput = new prism.FFmpeg({
        args: ["-f", "opus", "-ar", "48000", "-ac", "2", "-codec:a", "libmp3lame", "-i", "pipe:0"]
    });
    listener.pipe(audioOutput);

    buffer = []
    audioOutput.on("data", (data) => {
        buffer.push(data);
    });

    listener.on("end", async () => {
        audioOutput.end();

        const audioOutputFilename = "output.mp3";
        saveOutput(audioOutputFilename, buffer);

        await transcribe_and_reply(audioOutputFilename);
        setMuteSelectively(true, false, false, voiceChannel, user.id);
    });
}

async function onStartedSpeaking(userId, voiceChannel) {
    let user = await getUserById(userId);

    let listener = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });
    let audioOutput = new prism.FFmpeg({
        args: ["-f", "opus", "-ar", "48000", "-ac", "2", "-codec:a", "libmp3lame", "-i", "pipe:0"]
    });
    listener.pipe(audioOutput);

    buffer = []
    audioOutput.on("data", (data) => {
        buffer.push(data);
    });

    listener.on("end", async () => {
        audioOutput.end();

        const audioOutputFilename = "output.mp3";
        saveOutput(audioOutputFilename, buffer);

        await transcribe_and_reply(audioOutputFilename);
        setMuteSelectively(true, false, false, voiceChannel, user.id);
    });
}

To be honest, I don't feel to sure in this field at all. Is that the way I'm supposed to do it? I keep getting EOF.

ThePedroo•5mo ago

You shouldn't need to manually end audioOutput Please also provide backtrace of the error

KaiFireborn•5mo ago

Oh, sorry, completely fogot. So basically, this works perfectly: (without converting the file to mp3, of course)

async function onStartedSpeaking(userId, voiceChannel) {
    getTimeElapsedSinceLastCall();
    let user = await getUserById(userId);
    logReliably("🟢 " + user.globalName + " started speaking. " + getTimeElapsedSinceLastCall());
    setMuteSelectively(true, false, true, voiceChannel, user.id);

    // voiceReceiver.speaking.removeAllListeners();
    let opusStream = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });
    let audioOutput = new prism.FFmpeg({
        args: ["-f", "opus", "-ar", "48000", "-ac", "2", "-codec:a", "libmp3lame", "-i", "pipe:0"]
    });
    // opusStream.pipe(audioOutput);

    buffer = []
    opusStream.on("data", (chunk) => {
        buffer.push(chunk);
    });

    opusStream.on("end", async () => {
        logReliably("🟠 " + user.globalName + " stopped speaking, saving. " + getTimeElapsedSinceLastCall());
        setMuteSelectively(false, true, true, voiceChannel, user.id);

        const audioOutputFilename = "output.mp3";
        saveOutput(audioOutputFilename, buffer);

        // await transcribe_and_reply(audioOutputFilename);
        setMuteSelectively(true, false, false, voiceChannel, user.id);
    });
}

async function onStartedSpeaking(userId, voiceChannel) {
    getTimeElapsedSinceLastCall();
    let user = await getUserById(userId);
    logReliably("🟢 " + user.globalName + " started speaking. " + getTimeElapsedSinceLastCall());
    setMuteSelectively(true, false, true, voiceChannel, user.id);

    // voiceReceiver.speaking.removeAllListeners();
    let opusStream = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });
    let audioOutput = new prism.FFmpeg({
        args: ["-f", "opus", "-ar", "48000", "-ac", "2", "-codec:a", "libmp3lame", "-i", "pipe:0"]
    });
    // opusStream.pipe(audioOutput);

    buffer = []
    opusStream.on("data", (chunk) => {
        buffer.push(chunk);
    });

    opusStream.on("end", async () => {
        logReliably("🟠 " + user.globalName + " stopped speaking, saving. " + getTimeElapsedSinceLastCall());
        setMuteSelectively(false, true, true, voiceChannel, user.id);

        const audioOutputFilename = "output.mp3";
        saveOutput(audioOutputFilename, buffer);

        // await transcribe_and_reply(audioOutputFilename);
        setMuteSelectively(true, false, false, voiceChannel, user.id);
    });
}

However, the moment I uncomment opusStream.pipe(audioOutput);, "end" never gets emitted on opusStream So I tried rewiring both events to audioOutput:

audioOutput.on("data", (chunk) => {
        buffer.push(chunk);
    });

    audioOutput.on("end", async () => {
        logReliably("🟠 " + user.globalName + " stopped speaking, saving. " + getTimeElapsedSinceLastCall());
        setMuteSelectively(false, true, true, voiceChannel, user.id);

        const audioOutputFilename = "output.mp3";
        saveOutput(audioOutputFilename, buffer);

        // await transcribe_and_reply(audioOutputFilename);
        setMuteSelectively(true, false, false, voiceChannel, user.id);
    });

audioOutput.on("data", (chunk) => {
        buffer.push(chunk);
    });

    audioOutput.on("end", async () => {
        logReliably("🟠 " + user.globalName + " stopped speaking, saving. " + getTimeElapsedSinceLastCall());
        setMuteSelectively(false, true, true, voiceChannel, user.id);

        const audioOutputFilename = "output.mp3";
        saveOutput(audioOutputFilename, buffer);

        // await transcribe_and_reply(audioOutputFilename);
        setMuteSelectively(true, false, false, voiceChannel, user.id);
    });

But now the stream's "end" gets emitted almost immediately (40ms, independent of when opusStream (the user talking) actualkly ends)... I suspect the args for FFmpeg are wrong to some extent, but I lack the experience with the module to actually tell where

ThePedroo•5mo ago

Try removing -i pipe:0, it should be added by prism-media by default

KaiFireborn•5mo ago

That actually works! The last problem is that the written file is unreadable, yet again.

ThePedroo•5mo ago

If you want to save to a file Use -i filename, don't use fs That way you have faster execution and allows you to not need to save all data to an array If you simply want to send to Discord, you don't need to save to a file though

KaiFireborn•5mo ago

No, I have to send the file to a different API that only accepts files Well I tried, but the moment I add that "-i" flag, it goes back to the 40ms problem

ThePedroo•5mo ago

Unless it's a library, it will always accept buffers -i file.mp3?

KaiFireborn•5mo ago

Yeah args: ["-f", "opus", "-ar", "48000", "-ac", "2", "-codec:a", "libmp3lame", "-i", "output.mp3"]

ThePedroo•5mo ago

oops 🤦‍♂️ Messed up with the arguments remove the -i

KaiFireborn•5mo ago

Oh I wondered, it means input, right?

ThePedroo•5mo ago

Yep, I confused it with output

KaiFireborn•5mo ago

Okay, the file is still not playable, and now it seems to trigger on("end") over and over again Because instead of a single "started speaking/stopped speaking" log it's just spam

ThePedroo•5mo ago

Save the opus to a file, and try using ffmpeg via command-line to see what happens

KaiFireborn•5mo ago

// voiceReceiver.speaking.removeAllListeners();
    let opusStream = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });
    buffer = []
    opusStream.on("data", (chunk) => {
        buffer.push(chunk);
    });
    // let audioOutput = new prism.FFmpeg({
    //     args: ["-f", "opus", "-ar", "48000", "-ac", "2", "-codec:a", "libmp3lame", "output.mp3"]
    // });
    // opusStream.pipe(audioOutput);

    opusStream.on("end", async () => {
        logReliably("🟠 " + user.globalName + " stopped speaking, saving. " + getTimeElapsedSinceLastCall());
        setMuteSelectively(false, true, true, voiceChannel, user.id);

        const audioOutputFilename = "output.mp3";
        saveOutput("test.opus", buffer);

        // await transcribe_and_reply(audioOutputFilename);
        setMuteSelectively(true, false, false, voiceChannel, user.id);
    });

// voiceReceiver.speaking.removeAllListeners();
    let opusStream = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });
    buffer = []
    opusStream.on("data", (chunk) => {
        buffer.push(chunk);
    });
    // let audioOutput = new prism.FFmpeg({
    //     args: ["-f", "opus", "-ar", "48000", "-ac", "2", "-codec:a", "libmp3lame", "output.mp3"]
    // });
    // opusStream.pipe(audioOutput);

    opusStream.on("end", async () => {
        logReliably("🟠 " + user.globalName + " stopped speaking, saving. " + getTimeElapsedSinceLastCall());
        setMuteSelectively(false, true, true, voiceChannel, user.id);

        const audioOutputFilename = "output.mp3";
        saveOutput("test.opus", buffer);

        // await transcribe_and_reply(audioOutputFilename);
        setMuteSelectively(true, false, false, voiceChannel, user.id);
    });

Well "end" triggers correctly - but only if that .on("data) event is there Not sure if the file has anything usable inside, but one gets written and it takes 8KB; The ones written by ffmpeg are all 0KB I don't remember ever being this confused before -_- Actually no, ffmpeg doesn't write a file at all I tried saving it into a buffer and putting that into a fs write stream, and the written file is still 0KB. I think the ffmpeg command ight still be wrong

ThePedroo•5mo ago

OOps

KaiFireborn•5mo ago

Okay nevermind, it totally works

ThePedroo•5mo ago

Discord buggy With cmd or with the code?

KaiFireborn•5mo ago

I downloaded a different file and it converted it with those arguments. With cmd Oh not anymore Hang on I've got an idea

ThePedroo•5mo ago

Also, you forgot to .end the opusStream wait nvm

KaiFireborn•5mo ago

Doesn't it end automatically through discordjs? Yeah

ThePedroo•5mo ago

Yep, I was confusing with fs write stream Ensure you're properly saving

KaiFireborn•5mo ago

Okay well, cmd doesn't work if you specify the format to opus However, the part in js still doesn't save the file at all

let opusStream = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });

    let audioOutput = new prism.FFmpeg({
        args: ["-ar", "48000", "-ac", "2", "-codec:a", "libmp3lame", "output.mp3"]
    });
    opusStream.pipe(audioOutput);


    audioOutput.on("end", async () => {
        logReliably("🟠 " + user.globalName + " stopped speaking, saving. " + getTimeElapsedSinceLastCall());
        setMuteSelectively(false, true, true, voiceChannel, user.id);

        // await transcribe_and_reply(audioOutputFilename);
        setMuteSelectively(true, false, false, voiceChannel, user.id);
    });

let opusStream = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });

    let audioOutput = new prism.FFmpeg({
        args: ["-ar", "48000", "-ac", "2", "-codec:a", "libmp3lame", "output.mp3"]
    });
    opusStream.pipe(audioOutput);


    audioOutput.on("end", async () => {
        logReliably("🟠 " + user.globalName + " stopped speaking, saving. " + getTimeElapsedSinceLastCall());
        setMuteSelectively(false, true, true, voiceChannel, user.id);

        // await transcribe_and_reply(audioOutputFilename);
        setMuteSelectively(true, false, false, voiceChannel, user.id);
    });

Right now it looks like this... Thanks for your time, by the way

ThePedroo•5mo ago

What outputs if you ffprobe inputfile? Np also

KaiFireborn•5mo ago

I don't have an actual input file since it all comes from the opus stream, but trying it with the placeholder one I get

ThePedroo•5mo ago

Save it to a file, and execute ffprobe

KaiFireborn•5mo ago

let buffer = [];
    opusStream.on("data", (data) => {
        buffer.push(data);
    });

    opusStream.on("end", async () => {
        fs.writeFileSync("output.opus", Buffer.concat(buffer));
    })

let buffer = [];
    opusStream.on("data", (data) => {
        buffer.push(data);
    });

    opusStream.on("end", async () => {
        fs.writeFileSync("output.opus", Buffer.concat(buffer));
    })

Like this, right?

ffprobe version 6.1.1 Copyright (c) 2007-2023 the FFmpeg developers
  built with Apple clang version 15.0.0 (clang-1500.1.0.2.5)
  configuration: --prefix=/usr/local/Cellar/ffmpeg/6.1.1_2 --enable-shared --enable-pthreads --enable-version3 --cc=clang --host-cflags= --host-ldflags='-Wl,-ld_classic' --enable-ffplay --enable-gnutls --enable-gpl --enable-libaom --enable-libaribb24 --enable-libbluray --enable-libdav1d --enable-libharfbuzz --enable-libjxl --enable-libmp3lame --enable-libopus --enable-librav1e --enable-librist --enable-librubberband --enable-libsnappy --enable-libsrt --enable-libssh --enable-libsvtav1 --enable-libtesseract --enable-libtheora --enable-libvidstab --enable-libvmaf --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libxvid --enable-lzma --enable-libfontconfig --enable-libfreetype --enable-frei0r --enable-libass --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libopenvino --enable-libspeex --enable-libsoxr --enable-libzmq --enable-libzimg --disable-libjack --disable-indev=jack --enable-videotoolbox --enable-audiotoolbox
  libavutil      58. 29.100 / 58. 29.100
  libavcodec     60. 31.102 / 60. 31.102
  libavformat    60. 16.100 / 60. 16.100
  libavdevice    60.  3.100 / 60.  3.100
  libavfilter     9. 12.100 /  9. 12.100
  libswscale      7.  5.100 /  7.  5.100
  libswresample   4. 12.100 /  4. 12.100
  libpostproc    57.  3.100 / 57.  3.100
output.opus: Invalid data found when processing input

ffprobe version 6.1.1 Copyright (c) 2007-2023 the FFmpeg developers
  built with Apple clang version 15.0.0 (clang-1500.1.0.2.5)
  configuration: --prefix=/usr/local/Cellar/ffmpeg/6.1.1_2 --enable-shared --enable-pthreads --enable-version3 --cc=clang --host-cflags= --host-ldflags='-Wl,-ld_classic' --enable-ffplay --enable-gnutls --enable-gpl --enable-libaom --enable-libaribb24 --enable-libbluray --enable-libdav1d --enable-libharfbuzz --enable-libjxl --enable-libmp3lame --enable-libopus --enable-librav1e --enable-librist --enable-librubberband --enable-libsnappy --enable-libsrt --enable-libssh --enable-libsvtav1 --enable-libtesseract --enable-libtheora --enable-libvidstab --enable-libvmaf --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libxvid --enable-lzma --enable-libfontconfig --enable-libfreetype --enable-frei0r --enable-libass --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libopenvino --enable-libspeex --enable-libsoxr --enable-libzmq --enable-libzimg --disable-libjack --disable-indev=jack --enable-videotoolbox --enable-audiotoolbox
  libavutil      58. 29.100 / 58. 29.100
  libavcodec     60. 31.102 / 60. 31.102
  libavformat    60. 16.100 / 60. 16.100
  libavdevice    60.  3.100 / 60.  3.100
  libavfilter     9. 12.100 /  9. 12.100
  libswscale      7.  5.100 /  7.  5.100
  libswresample   4. 12.100 /  4. 12.100
  libpostproc    57.  3.100 / 57.  3.100
output.opus: Invalid data found when processing input

Invalid data... I'm not sure if I'm saving it correctly though; When we saved it as .ogg, it used to work

ThePedroo•5mo ago

I'm curious If you add ogg head to it, and try to ffprobe it, what happens? FFmpeg probably doesn't support opus like that

KaiFireborn•5mo ago

Well I still have an ogg file saved from yesterday Nothing in that respect should've changed, so


Input #0, ogg, from 'output.ogg':
  Duration: 00:00:02.56, start: 0.000000, bitrate: 59 kb/s
  Stream #0:0: Audio: opus, 48000 Hz, stereo, fltp

Yeah it works properly Maybe we have to use some other library after all. I didn't come far with it previously, but maybe discordjs/opus? It has some kind of opus encoder/decoder, but I couldn't find any documentation

ThePedroo•5mo ago

Just add ogg heads and pass to ffmpeg Done

KaiFireborn•5mo ago

But for that I would need the prism-media alpha version again Wouldn't it break the FFmpeg class? Yeah I have no idea anymore.. Guess I'll try to continue tomorrow, but I'm pretty sure I'm doing that part wrong

ThePedroo•5mo ago

Not really, you'll just need to pipe 2x

KaiFireborn•5mo ago

let opusStream = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });
    const oggStream = new opus.OggLogicalBitstream({
        opusHead: new opus.OpusHead({
            channelCount: 2,
            sampleRate: 48000
        }),
        pageSizeControl: {
            maxPackets: 10
        }
    });

    let buffer = [];
    oggStream.on("data", (chunk) => {
        buffer.push(chunk);
    });

    

    opusStream.on("end", async () => {
        logReliably("🟠 " + user.globalName + " stopped speaking, saving. " + getTimeElapsedSinceLastCall());
        setMuteSelectively(false, true, true, voiceChannel, user.id);

        oggStream.destroy()
        let ws = fs.createWriteStream('output.ogg')
        ws.write(Buffer.concat(buffer))
        logReliably("Finished writing output.ogg")
        ws.end()
        buffer = []

        let audioOutput = new prism.FFmpeg({
            args: ["-ar", "48000", "-ac", "2", "-codec:a", "libmp3lame", "-i", "ogg_output.ogg", "mp3_output.mp3"]
        });

        // await transcribe_and_reply(audioOutputFilename);
        setMuteSelectively(true, false, false, voiceChannel, user.id);
    });

    opusStream.pipe(oggStream)

let opusStream = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });
    const oggStream = new opus.OggLogicalBitstream({
        opusHead: new opus.OpusHead({
            channelCount: 2,
            sampleRate: 48000
        }),
        pageSizeControl: {
            maxPackets: 10
        }
    });

    let buffer = [];
    oggStream.on("data", (chunk) => {
        buffer.push(chunk);
    });

    

    opusStream.on("end", async () => {
        logReliably("🟠 " + user.globalName + " stopped speaking, saving. " + getTimeElapsedSinceLastCall());
        setMuteSelectively(false, true, true, voiceChannel, user.id);

        oggStream.destroy()
        let ws = fs.createWriteStream('output.ogg')
        ws.write(Buffer.concat(buffer))
        logReliably("Finished writing output.ogg")
        ws.end()
        buffer = []

        let audioOutput = new prism.FFmpeg({
            args: ["-ar", "48000", "-ac", "2", "-codec:a", "libmp3lame", "-i", "ogg_output.ogg", "mp3_output.mp3"]
        });

        // await transcribe_and_reply(audioOutputFilename);
        setMuteSelectively(true, false, false, voiceChannel, user.id);
    });

    opusStream.pipe(oggStream)

Okay I tried real quick, and the problem with end triggering over and over is there again. I'll see what I can do in way of debugging tomorrow, but Okay so being me I couldn't just let the problem hang and focus on my other work So in the end, I just decided to run the ffmeg command through the terminal instead of using that prism-media class to convert the written .ogg file to .mp3.

async function onStartedSpeaking(userId, voiceChannel) {
    let user = await getUserById(userId);

    let opusStream = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });
    const oggStream = new opus.OggLogicalBitstream({
        opusHead: new opus.OpusHead({
            channelCount: 2,
            sampleRate: 48000
        }),
        pageSizeControl: {
            maxPackets: 10
        }
    });

    let buffer = [];
    oggStream.on("data", (chunk) => {
        buffer.push(chunk);
    });

    opusStream.on("end", async () => {
        oggStream.destroy()
        let ws = fs.createWriteStream('output.ogg')
        ws.write(Buffer.concat(buffer))
        logReliably("Finished writing output.ogg")
        ws.end()
        buffer = []

        exec("ffmpeg -y -i output.ogg -ar 48000 -ac 2 -codec:a libmp3lame output.mp3")
    });

    opusStream.pipe(oggStream)
}

async function onStartedSpeaking(userId, voiceChannel) {
    let user = await getUserById(userId);

    let opusStream = voiceReceiver.subscribe(userId, {
        end: {
            behavior: EndBehaviorType.AfterSilence,
            duration: allowedPauseDurationForInputVoice
        }
    });
    const oggStream = new opus.OggLogicalBitstream({
        opusHead: new opus.OpusHead({
            channelCount: 2,
            sampleRate: 48000
        }),
        pageSizeControl: {
            maxPackets: 10
        }
    });

    let buffer = [];
    oggStream.on("data", (chunk) => {
        buffer.push(chunk);
    });

    opusStream.on("end", async () => {
        oggStream.destroy()
        let ws = fs.createWriteStream('output.ogg')
        ws.write(Buffer.concat(buffer))
        logReliably("Finished writing output.ogg")
        ws.end()
        buffer = []

        exec("ffmpeg -y -i output.ogg -ar 48000 -ac 2 -codec:a libmp3lame output.mp3")
    });

    opusStream.pipe(oggStream)
}

Once again, thanks a ton for your help - it's probably what you meant earlier, too. Definitely couldn't have done it alone..

ThePedroo•5mo ago

(Also, prism-media also executes through shell, soo..)

KaiFireborn•5mo ago

(Funny, couldn't get it to work that way)

discord.js - Imagine an app Join

Support server for discord.js, a Node.js module to interact with the Discord bot API.

58KMembers

View on Discord

Want results from more Discord servers?

Add your server

More Posts

Pre-fill message input with slash command on click buttonthe purpose of what im doing is to have a message send from my bot with button under, when a user cl Client as GuildMemberHello, this has probably been asked a million times, and I'm sorry for probably re-opening a topic. How can i remove my bot from 20k+ serversIt is been a lot of servers since i created the bot so i just want to rip them off. There is 35 sha Need a help in Music commandI have done with my music commands but the main problem is that the song is not playing. Full code Returning bot's missing permissionsHey, i made a function, which is supposed to check for the missing permissionsn in the array of perm Why isn't messageCreate getting triggered?```js import { Client, GatewayIntentBits, Partials } from 'discord.js'; const client = new Client({ Permission OverwritesDoes the bot need administrator permission to set permission overwrites on a channel? It's only work How Can I Make File Attachment Optionhttps://media.discordapp.net/attachments/1197096601987072000/1198442780394000534/image.png?ex=65beeb Cannot read properties of null when editing a automod rule## Description When attempting to update an automod rule i get an error from discord.js that it can Problem with Reddit APIWhen I run the /meme command, the console replies Unexpected token '<', "<!doctype "... is not a va