In the limitations, it said "Current solution has a maximum limit of 200 users connected at the same time to a given stream with the same captions language selected." If we want to support more users, what should we do? Thanks!

We want to use Chinese transcribe. Basically, people speak Chinese and get transcribe.
Tried to following suggestion to change
"This demo was built with English (en-US) as the language used for auto-captions.
You can change the default language/locale by editing LANGUAGE_CODE, DEFAULT_LANGUAGE_CODE and
MEDIA_SAMPLE_RATE_HERTZ in the constant.js file.

You can see a list of supported languages by Amazon Transcribe here: https://docs.aws.amazon.com/transcribe/latest/dg/supported-languages.html"

Diff is like following.

1. in ./serverless/transcribe-server/src/constants.js
   change
   LANGUAGE_CODE: process.env.LANGUAGE_CODE ?? "en-US",
   to
   LANGUAGE_CODE: process.env.LANGUAGE_CODE ?? "zh-CN",

change
MEDIA_SAMPLE_RATE_HERTZ:
process.env.LANGUAGE_CODE == "en-US" || process.env.LANGUAGE_CODE == "es-US"
to
MEDIA_SAMPLE_RATE_HERTZ:
process.env.LANGUAGE_CODE == "zh-CN" || process.env.LANGUAGE_CODE == "zh-CN"

change
DEFAULT_LANGUAGE_CODE: "en",
to
DEFAULT_LANGUAGE_CODE: "zh",

2. in ./serverless/translate-server/src/constants.js
   change
   DEFAULT_SOURCE_LANGUAGE_CODE: "en",
   to
   DEFAULT_SOURCE_LANGUAGE_CODE: "zh",

3. in ./deployment/translate-languages.json
   add
   "English [en]": true,

4. in ./serverless/transcribe-server/build/task-definition-dev.json
   change
   "value": "en-US"
   to
   "value": "zh-CN"

5. in ./serverless/transcribe-server/build/task-definition-prod.json
   change
   "value": "en-US"
   to
   "value": "zh-CN"

But didn't seem to work.
Do I miss any file that need to change but I didn't?
Also tired to just change 1 ./serverless/transcribe-server/src/constants.js, and doesn't work neither.
Do you have any idea how can I make it work with another language?
Thanks a lot!

run bash deploy.sh error:

console:

aws console:

Hi team. Please add this line to the docker file for Stream Server

RUN echo "deb http://archive.debian.org/debian stretch main" > /etc/apt/sources.list

Or change the base image please.

Regards!

Lambda runtime is nodejs14.

Works if changed to nodejs20

I'm trying to adapt the transcription example to allow an incoming microphone stream to be forwarded to the Transcribe service but can't work out how to pipe the data from the incoming socket to the AudioStream async iterator. NodeJS code:

io.on('connection', function (socket) {
  console.log(`Client connected [id=${socket.id}]`);
  socket.emit('server-ready', `AWS STT Server ready [id=${socket.id}]`);

  socket.on('stt-start', async function () {
    await startRecognitionStream();
  });

  socket.on('stt-end', function () {
    stopRecognitionStream();
  });

  socket.on('stt-data', function (data) {
    // Incoming microphone audio data
    // How to get this into the transcribe AudioStream???
  });

  const transcribeIterator = async function* () {
    for await (const chunk of ???) {
      console.log(chunk);
      yield {
        AudioEvent: {
          AudioChunk: chunk,
        },
      };
    }
  };

  async function startRecognitionStream() {
    speechClient = new TranscribeStreamingClient({
      region: process.env.AWS_DEFAULT_REGION,
      credentials: {
        accessKeyId: process.env.AWS_ACCESS_KEY_ID,
        secretAccessKey: process.env.AWS_SECRET_ACCESS_KEY,
      },
    });
    
    const startStreamTranscriptionCommand = new StartStreamTranscriptionCommand({
      LanguageCode: 'en-GB',
      MediaSampleRateHertz: 16000,
      MediaEncoding: 'pcm',
      AudioStream: transcribeIterator(),
    });

    const startStreamTranscriptionCommandOutput = await speechClient.send(startStreamTranscriptionCommand);

    for await (const transcriptionEvent of startStreamTranscriptionCommandOutput.TranscriptResultStream) {
      if (transcriptionEvent.TranscriptEvent.Transcript) {
        const results = transcriptionEvent.TranscriptEvent.Transcript.Results;
        console.log(results);
      }
    }
  }

Can anyone help?

Hi,

Tried the code. Doesn't seem work as this.
We have some questions.

1. Looks like the download address for pcre-8.44.tar.gz is not available, need the following changes.
   In ./serverless/stream-server/Dockerfile, need to change.
   wget https://ftp.pcre.org/pub/pcre/pcre-8.44.tar.gz &&
   to
   wget https://sourceforge.net/projects/pcre/files/pcre/8.44/pcre-8.44.tar.gz && \

2. In aws configure, how should we set "Default output format"?

3. With the change 1 which I listed above. When run "bash deploy.sh", do you have any idea what this error means and how can I fix it?
   Waiter StackCreateComplete failed: Waiter encountered a terminal failure state: For expression "Stacks[].StackStatus" we matched expected path: "ROLLBACK_COMPLETE" at least once

Generating environment variables file for Player App...node:internal/fs/utils:344
throw err;
^

Error: ENOENT: no such file or directory, open 'stack.json'
at Object.openSync (node:fs:585:3)
at Object.readFileSync (node:fs:453:35)
at Object. (/Users/tianyu/Downloads/amazon-ivs-auto-captions-web-demo-2.0.2/deployment/generate-player-app-env-vars.js:13:33)
at Module._compile (node:internal/modules/cjs/loader:1101:14)
at Object.Module._extensions..js (node:internal/modules/cjs/loader:1153:10)
at Module.load (node:internal/modules/cjs/loader:981:32)
at Function.Module._load (node:internal/modules/cjs/loader:822:12)
at Function.executeUserEntryPoint [as runMain] (node:internal/modules/run_main:81:12)
at node:internal/main/run_main_module:17:47 {
errno: -2,
syscall: 'open',
code: 'ENOENT',
path: 'stack.json'
}

Thanks a lot!

I apologize if this is not the correct place to ask - I glanced at repost and tried googling and nothing obvious appeared.

I'm playing with the code in this example and have some questions about design choices, e.g. using web sockets for the communications between the translate and transcribe servers and the send transcription lambda, instead of say SNS.

I don't want to bother the developers (beyond asking this question) as:

• This project an illustrative example and, as pointed out, not a production ready service,
• So I don't expect each design choice to have been fully examined,
• And I'd rather they create more of these types of things than waste their time defending design choices

But, I also wouldn't mind discussing, why certain things were done certain ways as a way to figure out best practices.

So, can you point me to a discussion group?

Thanks.

I am using alternative approach using websocket-server.
After creating webSocket-server file , removed unwanted code and modify with websocket-server in cloud formation and related bash file.
I was able to create ECR image but deployment is failing.
getting the below error:

Could anyone, please provide the cloudformation.yaml for an alternative approach?

NOTE: We are formalizing our plans to enter AWS SDK for JavaScript (v2) into maintenance mode in 2023.

Please migrate your code to use AWS SDK for JavaScript (v3).