Two people in two places can send to the same studio: each gets an input of their own, and one of you cuts between the two on the desk. The inputs accept SRT from anywhere; the distance between you plays no part.

Two things you should know beforehand. The viewers only ever hear the person who is in the program right now. And the studio has no return channel: you do not hear or see each other through the studio.

Setting it up

  1. Set the number of inputs. Whoever owns the studio sets at least two inputs on the “Eingänge” (inputs) tab. That only works before the studio is started. In a running studio the number cannot be changed.
  2. Pass on the address of input 2. Every input has its own address in the form srt://<host>:<port>?streamid=<id>. The second person gets the address of input 2. Some apps want host, port and stream ID in separate fields.
  3. Both set up their source. Each person sends with what they have: OBS on a computer, an SRT app on a phone or a hardware encoder. The settings are in The right encoder settings.
  4. Start the studio and check. As soon as both are sending, two scenes with picture and level appear on the desk.
  5. Go on air. At least one destination has to be switched on for that.

An input's address normally stays the same for every stream. So the second person sets up their app once.

Who operates the desk

One of you cuts. The desk runs in the browser and shows program and preview with picture, sound and levels. With “Take” you swap the two, which is enough for a conversation between two people: one is in the program, the other on preview, and one click changes.

The change is a hard cut. Showing both in the picture at the same time, side by side or as a small picture in the large one, is not possible in the studio.

The sound: only who is in the program is heard

In the studio the sound follows the picture. What is heard is the sound of the input that is in the program, plus the overlay's sound. The two inputs are not mixed together. If person A is in the program and person B answers, the viewers do not hear B until you cut.

For a conversation in which both voices are to be heard the whole time, one side therefore has to send a signal that already contains both voices. This is how:

  1. You are both in a call, in Discord for example.
  2. One of you, best whoever sits at a computer with OBS, mixes the call's sound with their own microphone in OBS and sends both to their input.
  3. This input now contains both voices. As long as it is in the program, the viewers hear the whole conversation.

If both voices are also to be heard when the other camera is in the program, the second side has to do the same and mix the call into its signal too. If it cannot, because it only sends from a phone for instance, cut to its camera only briefly, or let that person speak alone during that time.

Set the volume of both inputs on the desk so that the change does not jump.

Talking to each other

Because the studio has no return channel, the person who is not operating the desk only learns something of the program by watching the stream on the platform. It arrives there with several seconds of delay and is therefore no good for a conversation.

So you need a way of your own to hear each other. A call in Discord or a similar service is the simplest for that. Mind two things:

  • Headphones for both. Otherwise the microphone picks up the other person's voice from the speaker.
  • The call in the stream only where you want it. Whoever does not mix the call into their signal on purpose makes sure it is not sent along by accident. Mute it in your sending app as needed.

Over the call, the person on the desk also announces who is in the program right now.

Before the first stream

  • Both inputs show picture and level on the desk.
  • Both are wearing headphones.
  • You have settled who cuts and whose signal contains both voices.
  • You have done a trial and listened to the result on the platform.

Read on