Resources for online interpretation

Hello there.

So, you want to organise a multilingual event – online!
We at the [bla] collective support self-organised groups with the technical means to facilitate the technical side of interpretation at events, but we usually do so with hardware, and have little to no experience with online-support.

That is why, despite the currently increasing need for online solutions, we can‘t offer you personal support before and during your event, but rather want to supply you with tips and information for making it happen yourselves. The following list is compiled of points we and some friends have found to be useful – special thanks to our InterpRISE buddies! (https://interprise.nirgendwo.info)

Please consider that technology changes quickly, and services we recommend might have changed quite a bit in the meantime or might no longer be available. If you encounter anything that has changed, or you find new solutions that might be useful to others as well, it would be great if you let us know – so that we can update our info and can let others know as well!

Also, interpretation is never just about technology. Even though we are a tech-collective, we try to give you as full a picture as possible of what you might want to consider regarding interpretation.

In this document you will find:


0) Big vs. small events

1) Video conference systems

2) Interpretation via video systems

3) Interaction between different languages

4) Interpreters’ collaboration in video conferences

5) Additional tools & pads

6) Communication with participants about interpretation

7) Feedback: please share your experience(s) and ressource(s)

0) Big vs. small events
This document was written mostly with bigger events in mind (e.g.: a panel of 5 speakers, 150 listeners in English, 20 listeners in French, 5 listeners in Russian).
Even though it might seem too extensive at first glance, we think the information provided here can still be useful for people planning smaller events (e.g. 15 people in total, of whom 10 speak English, 4 speak French and 1 speaks Russian). With smaller events, you should be able to „scale down“ any of the tips here to your needs. It would probably be a good idea to plan ahead: e.g. make people sign up for workshops beforehand and have everyone be transparent about their language needs, so that you can plan accordingly. Some workshops might need no interpretation at all, so your energy is better used at those where interpretation is actually required!

1) Video conference systems
We assume most events online are done in a video-format. To have a successful event, you will most likely want to chose a stable, but also a trustworthy conference system. Commercial services like Zoom are currently widely known, but there are a lot of other options, e.g. open source etc. There‘s pros and cons to all of them.

Commercial software (e.g. Zoom, Microsoft Teams, Skype)

These are commonly used commercial video conference softwares. Generally, their connection quality is better and more stable than with the free alternatives.

Security-wise, usually connections are encrypted in a way that the participants and the company running the software (and any institution that might have control over them) can listen to your conversations. As they are commercial, only the company running the service provides the server it is run on (you don‘t get to choose a server), and are usually the only ones offering technical support.

These companies usually collect metadata about you, as for example when you are calling, with whom, from where. It is possible that they pass this data on to the authorities.

The software can be used for free by installing an app on the computer or smartphone, or by accessing it through the web browser. Registration with your e-mail address is required. In case of Zoom, meetings have to be restarted every 40 minutes if you are using a free account. An admin account that also allows you for a larger number of participants has to be paid for. Some universities offer their students free access to Zoom.

You can find an analysis of the security of the different programs here: https://foundation.mozilla.org/en/privacynotincluded/categories/video-call-apps/
Here is an article about Zoom: www.schneier.com/blog/archives/2020/04/…

Open source web conferencing systems/ video, chat

  • chatb.org p2p – WebRTC video chat for Chrome/Firefox
  • jabber.thing.net – chat/skype replacement
  • jitsi – chat/skype replacement: Jitsi Meet is an open source software for video conferencing (for max 10 – 15 persons). With 2 participants, there‘s end-to-end encryption. From 3 participants only transport encryption, which means that server operators can have access to data, therefore it is important to choose trustworthy server operators. (sometimes one server is overloaded, then try another instance):
    Systemli : https://meet.systemli.org/
    Autistici/Inventati: https://vc.autistici.org/
    Mayfirst: https://meet.mayfirst.org/
    Immerda: https://meet.immerda.ch/
  • big blue button – web conferencing system , need to have a server/host. Enables breakout rooms. BigBlueButton only supports transport encryption, so it is important to choose a server operator you trust.
  • Signal
  • Linxxnet: bbb.linxx.net/b (telephone dial-up possible)
  • solitech: webinar.solitech.org/b (community provider for bindingly participating groups, contact: solitech@riseup.net)
  • cyber4edu: bbb.cyber4edu.org/b initiative to promote open source software in schools Password protection possible
  • Accelerator: open source software for webinars, which was developed at Reutlingen University of Technology. TU Reutlingen accelerator.reutlingen-university.de source code: github.com/cracker0dks/Accelerator Password protection possible
  • NextCloud Talk: NextCloud Talk is an application that can be installed in a Next-Cloud. A maximum of 4 people can participate in a conversation and can be used mainly to work together in small groups. But you can also invite people to a conversation who do not have a NextCloud account. The communication via audio/video is end-to-end encrypted, which is currently not possible with 4 people in Jitsi Meet, BigBlueButton, Accelerator, or Zoom. Systemli cloud.systemli.org/login (requires Systemli account to create a session) Freedom Cloud: freiheitswolke.org/ueber-uns
  • mumble – web conferencing system , need to have a server/host
    Mumble is an audio conferencing tool that allows larger groups (more than 20 for example) to meet online. To use mumble, you usually need to install a client and you need a little common training to find all the features. If you work together in larger groups for a longer period of time, it is worth trying out mumble. With the desktop application you can also password protect rooms and set up breakout rooms. The following community providers offer mumble:
    – systemausfall: babbeln.systemausfall.org
    – Mayfirst: mumble.mayfirst.org
    – Calyx institutes: mumble.calyxinstitute.org
    Systemli: talk.systemli.org Instructions for installing the mumble app called mumla.

2) Interpretation via video systems

[bla], as a collective working on overcoming language barriers, advise you to question which languages are spoken predominantly during events and why. Still, we think that from a technical viewpoint on video conferences, it is advisable to decide for one main language of the event used in one main video conference, and to have an additional channel for each additional language. These channels could be separate video conferences or phone conferences.
Interpreters would then listen to the main conference (for example in English), while their simultaneous interpretation is heard in the other conference by participants as well as speakers/panelists/… that do not speak the main language (e.g. in French). In this case, the French speakers will only listen to the French conference and shut off the sound in the English conference, while they can still watch what is going on there.

Separate video conference:
Usually, you can‘t have two programs accessing one camera at the same time, unless you have a setup designed specifically for that. With large groups, people in the main event should shut off their cameras (most of the time) anyway to increase sound and video quality for everyone. With some programs, you might not be able to be in two separate conferences by the same provider at the same time, and would have to use another one – try out different combinations that make sense for you, and also check whether both conferences fulfill your requirements of sound and video quality!

Interpretation into a sign language needs to be done via video, where at least the interpreters have to be visible to the participants. If signers are supposed to be able to ask questions in sign language, your setup needs to support them becoming visible to at least the interpreters as well.

Separate phone conference:
Phone conferences can be used for any spoken languages. If you use a phone conference not run via the internet, it shouldn‘t affect your internet connection, and therefore the main video conference. But some phone conference providers aren‘t free, or depending on people‘s phone plans/contracts and where they are located in combination with the phone operator, participants might have to pay something per minute.
Also depending on the operator, instructions on how to enter the conference room might be available in only some languages. Some operators use noises to „announce“ when someone enters or leaves the phone conference room. You might want to research phone conference providers specific to your country, since accessibility and rates will depend on the country a lot. Check out what works for you!


3) Interaction between the different languages:
Depending on how interactive your event is, there are different things to consider regarding speakers of different languages interacting. As explained above, the easiest setup would be to have one language per video/phone conference. So, what happens when a French-speaking persons wants to ask the English-speaking panelist a question?

Written questions
Some events with a large number of participants decide to not let people speak to the panelists directly, but to have them send their questions to e.g. a facilitator via private chat during the event (that way, the facilitator can also bundle questions that seem similar to them). In that case, people who do not speak the same language would be asked to send their questions to the interpreter via private chat in the separate conference, and in turn the interpreter can translate the question into the main language and send it to the facilitator in the main event just like everybody else.
Since there will most likely not be a break to take time to prepare and/or ask questions in this scenario, you definitely need two interpreters per language (see also section „Interpreters collaboration in video conferences“), so that one can continue interpreting what is being said, while the other can take care of the written interactions.

Spoken interaction
If people in your event interact verbally, you could go either of two ways:
1) Consecutive interpreting would mean that e.g. a french speaker would ask their question in French, directly in the main English-speaking conference – followed up by the English interpretation. This of course takes twice the time a question would normally take, and some people might be uncomfortable with being put in the spotlight of taking the extra time simply because they speak a different language than the majority of participants. Consecutive interpreting might make sense e.g. if a French speaking panelist wants to speak themselves in the main conference. It‘s important to make sure that the French-speaking conference still can hear the question in French – check for technical difficulties beforehand.
2) Simultaneous interpreting means that a French speaker would ask their question in the French-speaking conference. The interpreter, muted in the French-speaking conference and unmuted in the English-speaking conference, will ask the question in English in the main conference. With follow-up questions, this muting and unmuting can be a bit tricky.

Relay languages/several languages…
If e.g. a Russian speaking person wants to ask a question, but not all interpreters of all languages understand Russian, you shouldn‘t run into any problems, since – no matter whether you opted for consecutive or simultaneous interpretation – the main language, e.g. English, will always be heard in the main conference. Therefore any interpreter not understanding Russian could listen to the main conference and then interpret in another target language (eg. French).

Please try to make sure beforehand that technical challenges will not be what prevents people who speak a language other than the main language from actively participating in your event.

4) Interpreters‘ collaboration in video conferences
Interpreters do not have to be professional interpreters – often, a person who is fluent in two languages and really knows the topic you are talking about can be just as good a choice as someone who is a trained interpreter (interpreting experience helps, though). So, if you are looking for interpreters, it is often a good idea to look in your pool of participants or in your communities.
Interpreting becomes much easier if interpreters had a chance to prepare (content as much as vocabulary): let them know what the event is about, give them the agenda and/or slides beforehand, give them access to background information (e.g. websites, books, videos) and be open to their questions.
Working with headsets is a good idea because of sound quality and acoustic feedback, as well as because of the interpreters‘ ability to concentrate on what is being said. A team of interpreters might consider using more than one computer and/or screen. Figure out technical issues of your set-up beforehand, e.g. whether one computer and two headsets would fit together.

It is crucial to have two interpreters per language interpretation is necessary: interpreting is exhausting, so interpreters really need to be able to take turns (e.g. every 15 minutes). Because interpreters can recharge in the meantime, working as a team enhances the overall quality of the interpretation, and thus of the access people who do not speak the main language can have. When taking turns, one interpreter is actively interpreting, the other can support them by writing down more difficult things like numbers, names, or helping them out with vocabulary etc. (Forwarding written translations of questions to the facilitators would also be the passive interpreter‘s job, see section on „Interaction between different languages“). Technical difficulties are also easier to deal with if one person can keep concentrating on interpreting while the other informs the event organisers of what is going on.

If everything goes smoothly, the interpreters will not need to interrupt the conference. But if they do, it is for good reason – whether it be a technical problem or a speaker talking at the speed of lightning, if it prevents the interpreters from doing their job properly, it prevents the people who do not speak the main language from knowing what is actually going on. The event‘s facilitation should make sure whether things are going okay for the interpretation team (be it via chat or by listening in on the interpreted conference).

To be able to support one another, many interpreters prefer to be in the same room/sit next to one another. Obviously this might not be possible in every scenario. If interpreters are not in the same room, you need to make sure that they are still able to communicate effectively – depending on their preference, this could be even another phone line, a chat and/or a pad or shared document to write in, etc.


5) Additional tools & pads
Open source online tools – edit documents collaboratively in real-time:

These are not conference software, but are useful to use while you are in a conference – for interpreters collaborating just as much as for the facilitators of a big conference or for a working group taking minutes. They work in the same way as Jitsi Meet: Each document gets a URL that you can share with other participants, and whoever has the link can edit the document, no setup or registration required. Everyone can edit the document at the same time, with changes you make being shown live to everyone else.

  • pad.thing.net – text documents
  • calc.thing.net – edit simple spreadsheets
  • Etherpad Lite for text documents, for example on https://pad.riseup.net/
  • Ethercalc for spreadsheets, for example https://calc.rankenste.in/ (hosted by me)
  • WBO for drawings/charts, for example https://wbo.ophir.dev/
  • FacilMap for maps, for example https://facilmap.org/, https://facilmap.eauchat.org/
  • CryptPad seems to support text documents, code, presentations, spreadsheets, drawings and even more. https://cryptpad.fr/

shared storage space

  • hostb.org 1 click – file transfer
  • Next Cloud: is a cloud application that can be installed on a dedicated server as an alternative to Google Cloud or iCloud. There you can store your documents, use a shared calendar and find many other collaborative tools.
    – systemli: systemli.org/service/cloud.html
    – Freedom cloud: freiheitswolke.org/ueber-uns
    – next.thing.net – next cloud for thing.net users

6) Communication with participants about interpretation
The technical details we discuss here aren‘t very intuitive, if we‘re being honest. Just as you are taking time to get familiar with how all of this could work, you should give your participants the same chance. You could decide to attach a written description of how interpretation will work to the email (message/…) with which you are sending around the invite link (which conferences will they need to enter, what about switching of cameras and mics…). Still, you should plan for a couple of minutes at the beginning of the event to explain the technical setup to all participants – in all languages that you offer!
Apart from that, you need to be transparent and up front about what you can offer before the event even starts. When advertising your event, make sure people know for speakers of which languages the event will be accessible – and try to make sure to also advertise it in all languages in which it will be accessible. You can also ask attendants about their language needs beforehand and try to find interpreters for the languages requested.
Panelists are participants who speak a lot and need to be aware of how to speak interpretation friendly. That means: they need to make complete sentences, not speak to fast and take a short break once in a while – but still, speak naturally and as they normally would. It makes a lot of sense for every speaker to announce which language they are going to be speaking to leave to the interpreters time to e.g. make technical adjustments when there is a language switch.

7) Feedback: please share your experience(s) and ressource(s)

We hope these points of advice did help you to some extent. You might have found other sources of information. Your experience can help other people: How did you organize? Which software did you use? What went well and what could have been done better?

We would love to receive a short feedback from you: please email us at bla [at] inventati.org