Better translation quality and less delay for fast speakers, a Welcome slide with your QR code, listener analytics, and Post-production for your sermons and teachings.
This is everything that has changed in MulingStream since our last update, with an explanation of how each part works.
The translation has been improved in three ways at once: the quality of the text, the smoothness of what listeners read and hear, and the delay when the speaker is fast. The text appears first and the voice follows, and the voice is steadier than before.
One problem is gone. In Japanese and Chinese, one spoken sentence could be shown and spoken more than once. That has been corrected, and the same correction covers short pieces in other languages, so a few words no longer come back as half a sentence.
Four settings in the room let you choose how the translation behaves. They are in the room settings, under Smart Translation.
Translation Length is a slider. At one end, Full length stays closest to the speaker's words, and can fall behind a fast speaker. At the other end, Sync up keeps pace with the speaker. Natural pace and Interpreter sit in between. Move it towards the end that suits your speaker.
Auto keep-up keeps pace with a fast speaker by itself. When the speaker slows down, it returns to the setting you chose.
Faster delivery sends lines to your listeners sooner when the speaker uses long sentences.
Live Interpreter on Stage is for a service where a person on stage repeats the speaker in another language. MulingStream recognises the repeat and skips it, so listeners hear each thing once.
When the speaker mutes the microphone, the room pauses. Listeners see a message that the speaker has paused and that the translation continues when the speaker talks again. No credits are used while it is paused, your passes stay as they are, and the automatic stop waits until the microphone is on again.
If a pass ends while the service is still going on, the room can load a new one after 30 seconds' notice. It uses a pass you already have running. If there is none, it uses your smallest pass that covers the languages of the room. You switch this on in the room settings. The list of passes you can pick from shows only the passes you have taken out, so you never choose one that is loaded in another room.
A pass counts only after ten minutes of listening in total, summed over everyone in the room. A short test costs nothing.
A listener now chooses a language before joining, and the screens that ask for the microphone explain what they are for. There is a try it room with no account needed, and four new How it works pages: an overview, one for listeners, one for leaders, and one for sound and media.
The Invite Listeners tab in the room holds everything a listener needs to join. Show the QR code, or show the login details with the room passcode, and copy the invitation link, which shows the room address. A preview shows what listeners see on their phones.
The QR code and the passcode can be reset, for example if a code reached someone it should not. A reset is possible once an hour for each room. After a QR code reset, the old code and the room link stop working at once, so printed handouts, posters and slides with the old code need to be replaced. After a passcode reset, the old passcode stops working at once, and listeners who join with the passcode need the new one.
The handout has been redesigned and is ready to print. You choose its language, and the file is named after the room and the language.
The Welcome slide is a picture for your projector screen that carries your room's QR code, so everyone can join in their own language. You choose the room, the languages shown on the slide, and one of three styles:
Need translation?, which asks the question in English and in the languages you choose, around the QR code. Joining guide, which shows how to join in three steps, in any language, for the screen or to print and hand out. Phone preview, which puts the QR code beside a phone showing what people see when they scan it.
English is always on the slide. You can place the MulingStream logo in a corner or on top or bottom, or leave it off, and use your own colours. Download the picture for the screen, or print it on A4. People scan the code with their phone camera, choose their language and listen. No app is needed.
Each screen you use for captions is a card with its own switch for on and off. You see at a glance which screens are showing captions and which are off.
You can now see how many people listened, and how. Open any session or pass to get detailed insight into its listeners. The Listener analytics page gives you the overall picture across all your gatherings: listeners per month, listening hours per month, the number of gatherings, and the browsers used. It also tells you how much of the listening was on a computer, since people on a computer usually follow from home rather than in the room. Choose the last 24 hours, 14 days, 30 days or 12 months, and download the page as a PDF.
After a gathering, you can now download, share and sell your sermons and teachings, in text and in audio.
MulingStream finds the speaking parts of the gathering and labels them: sermon, teaching, prayer, worship, announcements and so on. You tick the parts you want and the languages you want.
Text is free when you let us share it in the MulingStream library, with every private person left unnamed. Later you can share it on your own MulingStream page too. Publishing sermons is free. Text comes as subtitles (SRT and VTT), plain text (TXT) and PDF, in the languages you choose.
Translated audio is paid. It comes as MP3, and you can listen to it on the page or download it. Choose a voice for each language and hear a sample first. You can add a short spoken notice at the start of each audio file, with the title, the speaker and the church or conference, saying that the translation is made with AI. Later you can share the audio on your own MulingStream page too.
You have the right to share and sell your texts and your audio.
Everything you exported is on one page, Your exports, with a progress bar while the files are made, and each file can be downloaded as soon as it is ready. Replay shows a gathering as your listeners saw it. Parts such as announcements, prayer and testimonies can name people and share personal matters, so you decide how they are used, and we do not use, keep or publish them. Rooms bought with credits get the same tools as church rooms.
Smart Translation has two new subcategories under Christianity, Catholic Mass and Orthodox Liturgy, with the church words of each tradition kept in every language. The pages, the app and the emails are in more languages, and MulingStream translates between 200+ languages with their regional variants.
Quick Actions on your dashboard now has eight cards, four to a row: create a room, see your rooms, become a member, top up credits, Post-production, Your exports, Welcome slide and Listener analytics.
If you have a question about any of this, write to us or reply to the last email we sent you.