WeChat tests "Voice Clarity Enhancement" in limited rollout: one-tap local AI noise reduction for voice messages

WeChat drops a big move! Gray-scale testing of "Voice Clarity Enhancement," the savior for those who can't hear voice messages clearly is here. Jaygee 24 minutes ago. Never worry about not hearing clearly again.

On November 2, a netizen posted that WeChat is quietly running another gray-scale test, an update related to voice and sound.

On the sound side, a "dialect-to-text" feature has been added. When WeChat recognizes that what the other party is saying is a dialect, it can convert it to text for easy reading.

Image source: WeChat screenshot

It has also rolled out a gray-scale launch of the "Voice Clarity Enhancement" feature, with the entry point under "Me—Settings—General—Accessibility."

Xiao Lei Bibi (ID: xiaoleibbb) puts it simply: this feature is one-click AI noise reduction for the voice messages you receive.

Image source: WeChat screenshot

After enabling it, when you play a voice message again, the system will automatically isolate the human voice, optimize and brighten it, while suppressing background noise such as wind and traffic, making the spoken content more prominent.

For example, for a voice message recorded on the street, before enabling it the traffic noise drowns out the human voice; after enabling it, the human voice is noticeably clearer.

However, it must be said here that this feature only works on the playback end. Put plainly, if you enable it, it only lets you hear other people's voice messages more clearly; the voice messages you send out will not become noise-reduced just because you enabled this.

And if it is a voice message whose audio quality is already very poor, enabling it may instead amplify the original background noise.

It is worth mentioning that this noise reduction relies entirely on the phone's local AI algorithm to separate human voice and noise in real time. It does not invoke the user's microphone, nor does it upload voice content to a server. All processing is completed locally on your phone, so you don't need to worry about chat privacy being uploaded and analyzed.

Image source: Weibo

Xiao Lei Bibi (ID: xiaoleibbb) says frankly that nowadays many apps, the moment they mention AI features, default to requiring an internet connection to upload data. WeChat this time chose local processing, which in terms of privacy can be considered a consideration of users' needs.

A few months ago, WeChat also optimized the voice-to-text feature, making major changes to the interaction of the input box.

Previously, when Xiao Lei wanted to convert voice to text, he had to first tap the voice button next to the input box, then switch to text conversion mode. Not many steps, but every time it required one extra tap.

Now WeChat has removed the standalone voice-to-text button and changed it to long-pressing the input box to speak directly and convert to text, releasing to send. This is equivalent to hiding the text conversion entry inside the input box itself, shortening the operation path by one step.

Xiao Lei believes that from a product logic perspective, this update is one part "listening experience," using local AI for noise reduction to improve voice intelligibility. The other part is "sending efficiency," improving the efficiency of text conversion and lowering the cost of converting voice into standard text, so that voice messages can also adapt to more formal communication scenarios.

Image source: WeChat

The direction is right, because when users use WeChat now, many have to balance both: casual voice chatting in daily life, and formal communication at work. The needs of both sides must be met.

In addition to the above updates, WeChat is also gray-scale updating the following features: an input box prompt "hold to convert to text" switch, the ability to make video/voice calls simultaneously while sharing real-time location, support for conditional filtering when searching contacts, the ability to create group chats through multiple remark/nickname keywords, support for saving Live Photos in WeChat video calls, adjustments to the WeChat running interface, and more.

Anyway, all the features mentioned above are still being rolled out in batches for gray-scale testing. If you have already updated to the latest version of WeChat but cannot find the entry point or use it, that is normal.

Here Xiao Lei Bibi (ID: xiaoleibbb) reminds everyone not to readily believe in installing unofficial-channel installation packages in order to "grab the gray-scale." It is very unsafe. If your account is not on the gray-scale list, it still won't work. Rather than messing around, it's better to honestly wait for the official release. Anyway, according to WeChat's pace, generally after one or two months of gray-scale testing it will be fully rolled out.

In conclusion.

Xiao Lei wants to ask everyone: when you usually send WeChat messages, do you prefer sending voice or typing? For this wave of updates, which feature are you most looking forward to? Welcome to leave a comment in the comment section and let's discuss together.

Source: Leikeji

Images in this article are from: 123RF licensed image library. Infringement report. Back. Share. Related articles. All comments.