<VoiceConnection>.on('speaking', (user, speaking) => {...}) just stops without showing error or warning

[abp6318]

Please describe the problem you are having in as much detail as possible: Hello, I'm trying to make a live transcription bot using Google Cloud Speech-to-text API. The issue occurs without any warnings, errors, or breakage. It simply does nothing I can interpret. Upon reaching connection.on('speaking', (user, speaking) => {, the issue occurs. I did several hours of prior research and came across several posts from 2019 relating to the same issue, but none were resolved officially from what I can tell.

https://github.com/discordjs/discord.js/issues/3591 https://github.com/discordjs/discord.js/issues/3524

If you follow the link to issue 3524, you will see a variety of solutions provided by users more impatient than I, but all of these are a year old and I can't seem to make them work for me. A popular solution that I saw, but does not work for me, is to make the bot send out audio (or silence) in order to be able to hear a user.

Any and all hints/help/luck would be greatly appreciated. Include a reproducible code sample here, if possible:

// Place your code here
const connection = await memberVoiceChannel.join()
            const receiver = connection.receiver

            timer()
            connection.on('speaking', (user, speaking) => {
                timer()

                console.log("--\n\n" + transChannelList + "\n\n--")

                if (!speaking) {
                return
                }

                console.log(`I'm listening to ${user.username}`)
                const audioStream = receiver.createStream(user, { mode: 'pcm' })
                const requestConfig = {
                    encoding: 'LINEAR16',
                    sampleRateHertz: 48000,
                    languageCode: 'en-US'
                }

                const request = {
                config: requestConfig
                }
                const recognizeStream = googleSpeechClient
                .streamingRecognize(request)
                .on('error', console.error)
                .on('data', response => {

                    const transcription = response.results
                    .map(result => result.alternatives[0].transcript)
                    .join('\n')
                    .toLowerCase()
                    console.log(`Transcription: ${transcription}`)

                    if(transChannelList.length > 0){
                        for(index=0; index<transChannelList.length; index++){
                            chan = msg.guild.channels.get(transChannelList[index])
                            chan.send(`${user.username}: ${transcription}`)
                        }
                    }
                    if(transDMList.length > 0){
                        for(i=0; i<transDMList.length; i++){
                            mem = msg.guild.members.fetch(transDMList[i])
                            mem.then(function(member){member.send(`${user.username}: ${transcription}`)})
                        }
                    }

                })

                const convertTo1ChannelStream = new ConvertTo1ChannelStream()
                audioStream.pipe(convertTo1ChannelStream).pipe(recognizeStream)
                audioStream.on('end', async () => {
                console.log('audioStream end')
                })
            })

Further details:

• discord.js version: v12
• Node.js version: ^14.4.0
• Operating system: MacOS 10.15.5
• Priority this issue should have – please be realistic and elaborate if possible: My school is 10-15% deaf or hard of hearing students and many of the on campus clubs now have to rely on Discord to hold club meetings (thanks COVID). I want everyone at my school to feel included and make these meetings more accessible. I was hoping to have at least this fixed within the next few weeks so it could be functional for at least the back end of the Fall semester.

Relevant client options:

• partials: none
• gateway intents: none
• other: none

• [ ] I have also tested the issue on latest master, commit hash:

[mdable2]

I am having the same issue. I got it to work for a while and now it does not work after no edits on my part. No errors or anything and when I debug the connection, it prints debug: [WS] << {"op":5,"d":{"user_id":<my_id>,"ssrc":1074919,"speaking":1}} so it looks like it is picking up that I am speaking but the event speaking is not firing or getting picked up.

[mdable2]

Also, I followed the solution from #3524 and it worked a couple times but now again it is not working. I replaced the youtube link for sound with an mp3 and that is working again.

      // Only join if member is part of a voice channel themselves
      if (message.member.voice.channel) {
        // Join the voice channel the member is currently in
        const connection = await message.member.voice.channel.join();
        const dispatcher = connection.play(
          ytdl("https://www.youtube.com/watch?v=n2EjeMx4bSg", {
            range: { start: 0, end: 1 },
          })
        );
        dispatcher.on("start", () => {
          console.log("Scribe: Play Starting...");
        });
        dispatcher.on("finish", () => {
          console.log("Scribe: Finished playing!");
        });
        dispatcher.on("end", (end) => {
          console.log("Scribe: End Finished playing!");
        });

        connection.on("speaking", (user, speaking) => {
          console.log("here");
        });

[shreyasajj]

Where did the new ConvertTo1ChannelStream come from?

[karelkryda]

Same problem, sometimes works, sometimes not

[amishshah]

Duplicate of #5209