Stream.new

An open-source example application that allows users to record and upload videos using Mux

View Demo · Report a Bug · Request Features

About The Project
- Components
Getting Started
Videos to Test in Development

About The Project

Components:

Mux:

Direct uploads - this is an API for uploading video files from a client to create Mux Assets
Webhook signature verification - webhook signature verification to make sure Mux webhooks are coming from a trusted source
HLS.js - for doing HLS video playback of videos
Mux Data - for tracking video quality metrics.

Slackbot moderator. This examples allows you to configure a SLACK_WEBHOOK_ASSET_READY. When a new Mux asset is ready, an Incoming Webhook for slack will be sent. This is an example of how you might integrate a Slack channel that can be used to moderate content. The Slack message contains the asset ID, playback ID and a storyboard of thumbnails from the video.

NextJS:

SWR — dynamically changing the refreshInterval depending on if the client should be polling for updates or not
/pages/api routes — a couple endpoints for making authenticated requests to the Mux API.
Dynamic routes using getStaticPaths and fallback: true, as well as dynamic API routes.

This app was created with the NextJS with-mux-video example as a starting point.

Getting Started

Step 1. Create an account in Mux

All you need to set this up is a Mux account. You can sign up for free and pricing is pay-as-you-go. There are no upfront charges, you get billed monthly only for what you use.

Without entering a credit card on your Mux account all videos are in “test mode” which means they are watermarked and clipped to 10 seconds. If you enter a credit card all limitations are lifted and you get $20 of free credit. The free credit should be plenty for you to test out and play around with everything before you are charged.

Step 2. Set up environment variables

Copy the .env.local.example file in this directory to .env.local (which will be ignored by Git):

cp .env.local.example .env.local

Then, go to the settings page in your Mux dashboard, get a new API Access Token that allows for "Full Access" against Mux Video and set each variable in .env.local:

MUX_TOKEN_ID should be the TOKEN ID of your new token
MUX_TOKEN_SECRET should be TOKEN SECRET
MUX_WEBHOOK_SIGNATURE_SECRET (optional) - the webhook signing secret if you set up webhooks (see below)
SLACK_WEBHOOK_ASSET_READY (optional) - the slack webhook URL that will be used for the Slackbot moderator feature (see below)
SLACK_MODERATOR_PASSWORD (optional) - this is the password when you want to take actions from the Slackbot moderator feature (see below)
NEXT_PUBLIC_MUX_ENV_KEY (optional) - this is the mux environment key for Mux Data integration

Step 3. Deploy on Vercel

You can deploy this app to the cloud with Vercel (Documentation).

To deploy on Vercel, you need to set the environment variables using Vercel CLI (Documentation).

Install the Vercel CLI, log in to your account from the CLI, and run the following commands to add the environment variables. Replace the values with the corresponding strings in .env.local:

vercel secrets add stream_new_token_id <MUX_TOKEN_ID>
vercel secrets add stream_new_token_secret <MUX_TOKEN_SECRET>

Then push the project to GitHub/GitLab/Bitbucket and import to Vercel to deploy.

Step 4 (optional) Slackbot Moderator

This application uses a slackbot to send message to a slack channel every time a new asset is ready for playback. This requires a few steps for setup.

First, login to your Mux dashboard and in the left sidebar navigation find Settings > Webhooks. Create a new webhook and makes sure you are creating a webhook for the environment that matches the access token that you are using.

For local development you may want to use a tool like ngrok to receive webhooks on localhost. The route for the webhook handler is /api/webhooks/mux (defined in this NextJS app under ./pages/api/webhooks/mux).

Create a Slack 'Incoming Webhook'. Configure the channel you want to post to, the icon, etc.

When you're done with this, you should have a slack webhook URL that looks something like https://hooks.slack.com/services/....

Set the optional environment variables either directly in the vercel UI or by updating vercel.json and setting them as secrets for your organization. The optional environment variables are:

MUX_WEBHOOK_SIGNATURE_SECRET - This is a security mechanism that checks the webhook signature header when the request hits your server so that your server can verify that the webhook came from Mux. Read more about webhook signature verification. Note that in ./pages/api/webhooks/mux the code will only verify the signature if you have set a signature secret variable, so this step is optional.
SLACK_WEBHOOK_ASSET_READY - This is the https://hooks.slack.com/services/.... URL when you created the Slack Incoming Webhook.
SLACK_MODERATOR_PASSWORD - This is the password that will be used to authorize deleting assets from the slack moderator (The button with the red text "DELETE (cannot be undone)")
NEXT_PUBLIC_MUX_ENV_KEY - This is the env key to use with Mux Data. Note this is different than your API key and this environment key can be found on your environment page in the Mux dashboard
TELEMETRY_ENDPOINT - This is an endpoint where instrumented telemetry data is sent. As of this update, the only place we collect/send telemetry data is around upload performance in order to test different configurations of UpChunk, but there may be more in the future.

After all of this is set up the flow will be:

Asset is uploaded
Mux sends a webhook to your server (NextJS API function)
(optional) Your server verifies the webhook signature
If the webhook matches video.asset.ready then your server will post a message to your slack channel that has the Mux Asset ID, the Mux Playback ID, and a thumbnail of the video.

Step 5 (optional) Add automatic content analysis to Slackbot Moderator (Google Vision API)

stream.new can automatically moderate content with the help of Google's Cloud vision API.

Follow these steps to help moderate uploaded content:

GOOGLE_APPLICATION_CREDENTIALS - This is a base64 encoded JSON representation of your Google service account credentials. Follow instructions below.

First, you will need to set up a google developer account at cloud.google.com.
Create a project
Create a service account for your project and enable the "Cloud Vision API" for your project

Export a Google Service Account authentication file in JSON format. If you have a file that is like this:

service_account.json

{
  "type": "service_account",
  "project_id": "",
  "private_key_id": "",
  "private_key": "-----BEGIN PRIVATE KEY-----\",
  "client_email": "",
  "client_id": "",
  "auth_uri": "",
  "token_uri": "",
  "auth_provider_x509_cert_url": "",
  "client_x509_cert_url": ""
}

Get the base64 encoded string of this JSON file like so:

cat service-account.json | base64

^ This command will output one long string. This string is what you will use for the ENV var GOOGLE_APPLICATION_CREDENTIALS.

When the Slackbot Moderator message gets posted to slack, it will now include a "Moderation score (Google)" with 2 dimensions:

"adult"
"suggestive"
"violent"

Each dimension will have a score from 1-5. You should interpret these scores in terms of likelihood that the video contains this type of content. This is based on Google Vision's Likelihood score

1: very unlikely
2: unlikely
3: possible
4: likely
5: very likely

Step 6 (optional) Add automatic content analysis to Slackbot Moderator (Hive AI)

stream.new can automatically moderate content with the help of Hive AI.

Follow these steps to help moderate uploaded content:

HIVE_AI_KEY - This is a base64 encoded JSON representation of your Google service account credentials. Follow instructions below.

First, you will need to set up an account at thehive.ai.
Create a project
Get the API key for your prject

When the Slackbot Moderator message gets posted to slack, it will now include a section titled "Moderation score (hive)" with 2 dimensions:

"adult"
"suggestive"

Each dimension will have a score from 0-1 with a precision of 6 decimal places. These numbers come from the "Classification API" that Hive AI provides. Details here.

Videos to test in development:

When developing, if you make any changes to the video player, make sure it works and looks good with videos of various dimensions:

Horizontal

http://localhost:3000/v/Hi6we01h00uVvZc00GzvVXZW8C02Y8QC8OX7

Vertical

http://localhost:3000/v/UNDUU7tU7vYt02CRMDTlZd1qKjvk41LN6yI5LbHgtxo8

Super vertical

http://localhost:3000/v/seK501Bf00kyqSnGdMwQFi3lgqgdoS00qm5PAiV7Yjf2ew

Also be sure to check: Safari, Mobile Safari, Chrome, Firefox because they all behave a little differently.

drincruz/stream.new