Offer your articles as audio in another language, without redoing the work
Part of your audience does not read French, or you are aiming beyond France. WeDispatch's multilingual editions give a single article one audio version per language. What it produces, what it inherits from the original, and what it really costs.
Many sites have an audience that spills past French without their quite noticing: a tourism office that welcomes foreign visitors, an institution whose reports travel outside France, a brand present in several markets. The translated text often already exists. What is missing is the audio version in each language, so that the non-French-speaking listener hears it in their own rather than giving up. That is what WeDispatch's multilingual editions do: a single article gets one audio version per language, each with its own page and its own declared language. Here is what that produces in concrete terms, what settles itself, and the one point you should not kid yourself about, the cost.
An edition is one more reading
Let us start with the plain truth, because that is where the misunderstandings hide. An edition in another language is not a free variant of the French audio. It is an article read one more time, in another language, and so a separate audio file. Text to speech is paid for by the character: a text read twice, in French then in English, costs two readings, not one and a half. There is no hidden discount for translation, and that is as it should be: translating an article to have it read is producing another audio, as long as the first.
That is why the screen shows you the projection before you add a language. The figure is not invented for the demo: it is drawn from what your account has actually consumed over its current period, the same source that keeps your volume. If you already have two editions and you add a third, the product tells you what the month will look like with that extra reading, and whether it tips you over your volume. You decide knowing the number, not after discovering it on the invoice. For how the per-character price works, we laid it out in how much article audio costs.
What settles itself, and what you keep control of
The good surprise is somewhere other than the price. A translated edition does not arrive bare: it inherits the signature voice and the section of its source article. In other words, if your French article is signed and filed under "Culture", its English version is too, without your having to say it again for each language. That detail avoids a silent problem: without this inheritance, every translation would have fallen under unsigned articles and skewed your classification the moment a multilingual account appeared. You of course take back control whenever you want, but the default is the right one.
The reading language matters for the rest of the chain, not only for the ear. Each edition declares its language on its page, which lets a search engine know it is dealing with the Japanese or Spanish version of a piece of content, not a duplicate of the French. It is the same rigour as for the choice of voice: one voice per language, because a French voice droning out English hurts more than it helps, a point our page on French text to speech develops.
How far it goes, and what not to expect from it
The available reading languages run from French to Japanese, by way of Spanish, German, Italian, Portuguese, Arabic, Hindi and Korean. It is broad, and that is exactly where you have to stay honest about the nature of the thing. An edition is a reading of a translated text, not an acted dub that would carry a performer's intonation from one language to the next. The voice reads the target language well, but it reads it, it does not stage it. And the quality of the Japanese or Arabic version depends on the voices available in that language, not on the care we put into French: our speciality remains reading French, and we do not claim to be as sharp on twenty other languages as on our own.
Still, for an article that is already translated, the matching audio edition is the shortest path to an audience that text alone was letting slip. You do not redo the translation work, you do not reset the voice or the section each time, and you know what each extra language will cost you before you launch it. For a multilingual site that wants to speak to each reader in their language, with no nasty surprise on volume, it is a sober and predictable setup. If your need is rather a bilingual site with two voices on the same article, look instead at one voice per language, which covers that case closely.
Give your articles a voice with WeDispatch
This blog is itself voiced by WeDispatch. Curious how it sounds on your content?
Book a demo