Recording a WeChat Video with Free OBS Studio

阅读中文版

By Quwei Ye (叶曲炜, also known as Tree) · A practical guide for Windows

While preparing a lesson recently, I ran into a very specific problem. A WeChat video would make a useful learning reference, but there was no convenient download option. Opening the link in a browser also required scanning a QR code with WeChat before I could continue watching. So I turned to screen recording.

The software was installed, but why was the preview still black? Where had the recording gone? Had the computer audio been captured as well? This article brings together what I learned about setting up, testing, saving, and playing back a recording on a Windows computer.

I used OBS Studio, a free, open-source application for screen recording and live streaming. I asked AI to help check the settings and the results; the recording itself was made by OBS running on my computer. Its built-in features were enough for this task, without an additional recording plug-in. About OBS Studio

1. After installation, first find out where recordings are saved

OBS was already installed on my computer when I began this experiment. If you have not installed it yet, choose the Windows installer from the official OBS download page, download it, and follow the installation wizard.

When you first open OBS, you can run the Auto-Configuration Wizard and choose recording as your main purpose. If you skipped it, you can open it again from Tools > Auto-Configuration Wizard. OBS Quick Start Guide

Open Settings > Output, find Recording Path, and make a note of where your files will be saved.

You can create a dedicated folder, such as “OBS Recordings.” In my case, the files were saved in the computer’s Videos folder. Knowing the location beforehand helps avoid the frustrating feeling of having recorded something but being unable to find it.

2. If the preview is black, check where the picture is coming from

Two terms in OBS are worth understanding first. A Scene is a particular arrangement of content. A Source supplies something within that arrangement, such as a window, a display, an image, or a video file.

My initial setup contained empty Media Sources, so the preview did not show the WeChat video I wanted to record. AI then helped me create a scene, add a Window Capture source, and select the WeChat window that was actually playing the video. The picture appeared correctly.

1. Open the target video in WeChat on your computer and make sure it plays normally.
2. Click the plus sign in the Scenes panel in OBS and create a scene.
3. Click the plus sign in the Sources panel and select Window Capture.
4. Select the WeChat window displaying the video, then check that the preview shows the right content.

If several windows are called “WeChat,” try selecting them one at a time and watch the preview. If the picture does not fit, select the source, right-click, and choose Transform > Fit to Screen, or press Ctrl + F. Keep the original proportions and check that the subtitles have not been cut off. Window Capture documentation · Sources Guide

Window Capture: Captures a particular window. This is the source used for this recording.

Display Capture: Captures the entire display.

Media Source: Adds existing media, such as an audio or video file. For this task, I switched to Window Capture.

The video must already be playable. This recording used a video I could already watch in desktop WeChat. OBS did not replace WeChat’s login or QR-code scanning process.

3. Check the audio separately

Once the picture appeared, I still had to check whether the sound playing on the computer would be included in the recording. For this task, I enabled Desktop Audio to capture computer sound and muted Mic/Aux to keep my voice and room noise out of the recording.

The check is straightforward: play the video and watch whether the Desktop Audio level meter in the Audio Mixer moves with the sound. If it does not, open Settings > Audio, check the Desktop Audio device, and select the speakers or headphones currently in use. OBS Quick Start Guide

Desktop Audio may also capture notification sounds from other applications. Before recording, close unrelated music, voice calls, and message notifications.

Hearing the sound yourself is the first step. Hearing it again in the saved recording completes the check.

4. Record 10 seconds, then open the file and watch it

I began with a small request: “Record a 10-second test and see whether it captures both the video and the computer audio.”

1. Click Start Recording in OBS, then switch back to WeChat and play the video.
2. After about 10 seconds, return to OBS and click Stop Recording.
3. Wait until the button changes back to Start Recording, find the new file, and double-click it to play.

During playback, I checked three things: Was the picture changing? Was the sound clear? Were the subtitles fully visible? The short test captured a normal picture, changing subtitles, and computer audio, so I went on to record the complete video. For a much longer recording, a slightly longer test can also be useful.

5. Start recording before starting the video

When it was time for the full recording, my main question was: “When should I start playing the video?” This experience gave me a simple order to remember: start recording in OBS first, then play the video from the beginning.

Before recording, pause the video at the start, adjust the window size, and make sure the preview and audio meter look right. Click Start Recording, then play from the beginning. Keep the window and picture as steady as possible during playback. Wait until the video has actually finished before returning to OBS to stop the recording.

The video in this example was Understanding Machine Learning in 10 Minutes (《10分钟看懂机器学习》), by “51学通信.” Its actual running time was approximately 11 minutes and 1 second. The “10 minutes” in the title was therefore not a reliable point at which to stop.

Along the way, I encountered waiting time, playback-position changes, and automatic playback of the next video. I eventually asked AI to restart OBS and record the entire video again from the beginning.

The finished file was approximately 11 minutes and 1 second long and 69 MB in size. The picture and audio were checked, and the complete closing sequence was retained.

6. After stopping, how do you save and play the recording?

This was one of the steps I found most confusing: once recording was over, where was the Save button? OBS writes the recording to the path set earlier. After clicking Stop Recording and waiting for it to finish, you can find the file in that folder. There is no separate Save As step like the one used when editing a document.

Stop recording → wait for completion → open the recordings folder → double-click the file to play.

Use File > Show Recordings in OBS to open the folder, or check the Recording Path under Settings > Output. Once you find the file, double-click it to open it in a video player.

If you need an MP4 file, use Remux Recordings

If your file has an MKV extension and you would prefer an MP4 for sharing or editing, use File > Remux Recordings. The official OBS recording guide describes this option and explains that MKV offers better protection for a recording if recording is interrupted unexpectedly. Standard Recording Output Guide

Remuxing addresses the file format. Removing waiting time or cropping window borders is a separate editing task. The finished recording in this example was trimmed and cropped after recording. Clicking Stop Recording does not make OBS decide automatically which waiting scenes should be removed.

When something goes wrong, check these points first

A black preview, or the wrong picture: Check whether the source points to the WeChat window that is actually playing the video.

A picture with no sound: Check that Desktop Audio is enabled and the correct playback device is selected. Make a short recording and play it back.

You cannot find the recorded file: Make sure recording has stopped, then check Recording Path under Settings > Output.

The next video appears at the end: Watch for automatic playback. Stop recording promptly, or remove the extra section afterward.

Official references

• OBS Studio
• Official downloads
• Quick Start Guide
• Window Capture Sources
• Sources Guide
• Standard Recording Output Guide

7. A small recording task became a reusable learning process

Looking back, this experiment helped me work through three questions: Where does the picture come from? Where does the sound come from? Where does the file go?

I began by looking for a free tool, tried it myself, asked AI for help, checked a short test, and made a complete new recording. At the end, I had a file I could play. None of the individual steps was large, but each produced something I could check.

It also gave concrete meaning to my six-part learning framework, “Record, Think, Ask, Act, Review, and Transfer” (记、思、问、行、复、化): record what happens, think through the problem, ask for something actionable, try a small step, review the result, and carry the effective method into future work.

Next time I face a similar task, I can follow this sequence: make sure the video plays, choose the right picture source, check the audio, make and review a short test, record the whole video, then stop and save. Even a small screen-recording task can become an exercise in learning to work with AI.

About the author

Quwei Ye is a professor whose work focuses on questioning skills and innovation in learning in the age of AI. He has more than 40 years of experience in higher education teaching, administration, and educational evaluation. He formerly served as Vice President of Peking University Founder Education Group and President of Xi’an Technology and Business College (西安工商学院).

He holds a master’s degree in Artificial Intelligence and Expert Systems from the National University of Defense Technology and was a visiting scholar at Connecticut State University in the United States. His research has included knowledge discovery in databases (KDD) and data mining.

In recent years, he has focused on generative AI and AI agents, exploring their connections with problem-based learning, project-based learning, and social and emotional learning. His six-part framework—Record, Think, Ask, Act, Review, and Transfer—connects authentic records, independent thought, effective questions, action and verification, reflection, and the transfer of learning to new contexts.

Website: www.treeye.cn
WeChat: tree_ye1961
Email: shuye_1961@qq.com