Microsoft Edge, is this what you do?

No, it’s Microsoft’s artificial intelligence.

Let’s start with the Edge browser, which is certainly no stranger to you, as it has long since “abandoned the dark side” and replaced its kernel with Chromium, which was in this article when it was first rumored to be in beta.

However, despite Edge’s strong growth, some extensions and scripts on Edge are not as compatible as Chrome, and for habitual reasons, Chrome is still my primary browser.

However, I have recently discovered a feature on Edge that works extremely well and is not available on Chrome, which makes me envious.

This is not a recent feature, but we don’t care too much about it, and we don’t study it too much, but believe me, reading this article today will definitely make you want to “toss” it.

Because, although it is not human, but almost can be faked.

Read aloud

This is Edge’s exclusive “Read Aloud” feature, which can generate TTS voice reading from the text in the browser. This feature is not new, as many software and online websites have it, but these tools call on a common engine, and the voice sound is synthetic, so it is impossible to listen to it.

But Edge’s “read aloud” effect is different, you know Microsoft has many years of experience in the field of artificial intelligence speech synthesis, backed by the giant hard tree, Edge’s “read aloud” effect is comparable to the pronunciation of real people, you really can not tell unless you listen carefully.

Of course, it is better to listen to it directly than to say a thousand words.

Let’s listen to the female voice in Mandarin, Lady first.

Then listen to the Mandarin male voice.

How about this, the “two” Mandarin is not “ordinary” it, spit clear, the words are round, not rigidly read down, and even with a little intonation.

In short, the voice is not at all like those other software text-to-speech so mechanical and stiff, especially some video bloggers with voiceover, are such software text-to-speech, not a bit of emotion.

And what is another point? I do not know if you have found, that is, they read aloud when the sentence break is more accurate, that they can accurately determine the sentence break position.

As you know, when you are asked to read an article aloud for the first time, most people are probably not able to read it aloud fluently, and there will definitely be some mistakes in the middle, but as you can hear, the new Edge’s “Read Aloud” feature basically does not make such mistakes.

What’s more is that this feature can be used directly in the new Edge without installing any extensions, which makes it very easy and convenient to use.

This feature comes in handy when you want to listen to a novel on the web on your computer, or when you are tired of looking at the computer and want to relax and listen to the web content.

You can see the read aloud function by directly clicking on the menu bar.

Microsoft Edge, are you a human being?
Or you can click the right mouse button directly on the page.

Microsoft Edge, is this a human thing to do?
If you don’t need to read aloud the whole text on the web page, then select the text and right click to read aloud the selected content.

When you start reading aloud, some control buttons will appear at the top of the page, such as pause or switch paragraphs, and you can also adjust the reading speed and switch the voice in the voice option on the right.

Microsoft Edge, are you doing this for people?
When choosing a voice, I suggest choosing the first two in the red box for Mandarin, which are the two just shown, you can also choose Cantonese and the dialect of Taiwan Province.

For example, the Mandarin ones are Xiaoxiao and Yunyang, which are the names of girls and boys respectively. They are both part of the public voice in Microsoft Azure Cognitive Services Speech Synthesis.

Other supported languages, looking at the current mainstream foreign languages are not in the picture.

Microsoft Edge, is this what you do with people?
Everyone can try it, but the best result is still recognized as “xiaoxiao”, which is the example at the beginning of the article.

I think those video bloggers can totally use the new Edge’s read aloud function to dub their videos, so that would require opening the text content with the new Edge browser and creating a new text document, which can be opened directly in the new Edge browser.

Microsoft Edge, is this what you do with people?
Then the voice generated by reading aloud is saved inside the record, and the effect absolutely crushes the effect of the marketing number video on Jitterbug.

Mobile use

Unfortunately, the read aloud feature can only be used on the computer side of the new Edge, but thanks to a cool user named “丨丨丨丨” (yes, that’s his ID) who integrated Microsoft’s voice service into the App, you can then replace the phone’s built-in TTS engine with Microsoft’s, so you can call the “read aloud” feature on your phone, but only on However, it can only be used on Android phones.

After installing the app, first click on the system TTS settings to change the preferred engine to Read Out Loud, and below that you can also adjust the speed and pitch of your voice, and tap Play to try it out.

Pictures

Microsoft Edge, are you a human being?
Then click SSML speech synthesis marker language, you can see that it defaults to the voice of Xiaoxiao, the girl just now.

Microsoft Edge, are you a human being?
Then what is SSML speech synthesis markup language?

According to Microsoft’s official explanation.

Speech Synthesis Markup Language (SSML) is an XML-based markup language that allows developers to specify how to convert input text to synthesized speech using a text-to-speech service. Compared to plain text, SSML allows developers to fine-tune syllables, pronunciation, speech rate, volume, and other attributes of text-to-speech output. SSML can automatically handle normal pauses (for example, pausing for a moment after a period) or use the correct pitch in sentences that end with a question mark.

Images

Microsoft Edge, is this what you do with people?
Simply put, with this technology, Xiaoxiao can read aloud with more style, or emotion.

Microsoft Edge, are you a human being?
Microsoft Edge, are you a human being?
How does it work? For example, the code below sets Xiaoxiao’s angry style AI voice, copy and paste it into the input box after clicking SSML speech synthesis markup language in front and click OK to change the style.

Microsoft Edge, is this what you do?
When replacing other styles, you can change angry to other words, we recommend that you must try (pampered affectionate) this effect, very soulful.

(However, I found in the actual test process of the above code directly copied, some cell phones do not take effect, repeated the test back and forth more than N times, still can not solve the problem, suspected that the problem in the WeChat dialog box line marker and the editor line marker inconsistent.

In order to ensure that everyone can use it, the last solution we found was to save the code as a TXT notepad and reply to the password in the background to get it, which ensured that it would take effect, in order to fix the problem last night to one o’clock. (The first thing you can do is to use the App.

But this can only be used when you use the read-aloud function in the App, it does not change the phone’s own voice engine, such as Xiaoxiaoyou.

For example, when listening to a book with the previous Amway Reading App, you can first set different reading styles in the Read Aloud App, and then check the box to follow the system when reading aloud in the Reading App.

Microsoft Edge, is this what you do?
As for which one to use, it depends on what style you like. Speaking of which, I guess you will say “I have a bold idea” in the comments again, you know.

I believe there are many people who like to listen to books, but not all novels are read aloud by real people, so you can listen to anything with this.

Conclusion

This is the end of the story, from the new Edge read aloud function, Microsoft’s artificial voice synthesis effect has been good enough, but this is not the end.

In the words of one of our partners, the intonation, endings, accents, and even the intonation of each sentence are too much like those of ordinary people.

I don’t want to analyze more, I just want to say: I am the same.

Microsoft Edge, is this what you do?
Microsoft Edge, is this something that people do?
Microsoft Edge, is this something that people do?
Microsoft Edge, is this something that people do?
If you are interested, you can check out the video

However, the current codename F201 of the human voice is currently not open to use, I believe that there is no technical difficulty, precisely because the effect is too realistic, if open to use may bring unexpected hidden problems.

The problem now is that the ball has been kicked back to the humans, leaving really little time for the human team to represent.