Brick Wall Effect – Micro-stutter during manual flight (4K60 HEVC 10-bit)

Mini 5 Pro
1

Hi everyone,

I would like to report a repeatable issue that I have been investigating for several days.

For convenience, I call it the “Brick Wall Effect”, because I first noticed it while flying laterally alongside a brick wall.

The name is just a label. The issue may also appear on other highly detailed repeating textures.

What I observe

4K60 HEVC 10-bit recording.

The original camera file already contains several very short motion irregularities.

The effect is visible without YouTube compression.

The same irregularities remain after editing and exporting because they are already present in the original recording.

The most important observation

When I repeat the same route using Waypoints, the problem disappears completely.

The camera, settings and scene remain the same.

The main difference is that the drone follows a perfectly smooth programmed trajectory instead of manual stick input.

this is definitely an encoder bug.

I simply want to know whether other users can reproduce the same behaviour.

If you want to help

Please try:

4K60 HEVC 10-bit;

lateral flight;

manual control;

fly close to a brick wall or another highly detailed repeating texture;

then repeat exactly the same flight using Waypoints.

If you notice the same effect, please mention:

drone model;

firmware version;

SD card;

recording settings;

and, if possible, upload a short original clip.

Maybe together we can determine whether this is normal behaviour, a firmware issue, or something else.

2

I cannont help you with the " motion irregularities " , could they just be the camera’s inability to focus on the details of the wall as you fly pass… The camera might pick a spot on the wall to focus, but the drone is moving and then the camera has to refocus, again and again… But this is all a guess…

However I want to offer you a warning about flying along a wall…

I do not know how close you are flying to a wall (brick or not…), but it is not for the faint of heart…

When a drone flies near a vertical surface like a wall, the air pushed down by the propellers bounces off the wall and creates high pressure. This pushes the drone away. However, the air being pulled into the top of the propellers creates low pressure (like a vacuum). This low pressure pulls the drone toward the wall. If you are not actively steering away, the drone may suddenly drift into the wall. Hence the remark, " faint of heart… " :joy:

Since you are using “waypoints”, you are not actively flying the drone and probably more interested in the cinematic effect and imagery…

Most modern drones (like the DJI Mini or Air series) use sensors to see in front of and behind them, and even along side them… Flying along (parallel to) a wall means the drone is sideways. The side sensors on many drones are less advanced or turned off entirely. The drone might think it is completely safe while drifting sideways into wall.

3

Thank you for your reply.

The issue I’m seeing is not a loss of sharpness or the camera searching for focus. The image remains sharp. What I notice are very short motion irregularities in the original video file.
I also repeated the same flight using Waypoints along the same wall, with the same camera settings and approximately the same distance to the wall. In that case, the motion irregularities disappeared completely.

That is why I currently believe the wall itself is not the root cause. It simply makes the issue much easier to see because of its highly repetitive texture.

1 Like
4

In that case, I would recommend that if you still have the video post it for others to see, make a note of the specific time the issues is visible… You have not described what the “motion irregularities” are…

This is a little like a person calling a doctor and saying, “Doc, my back hurts…”

If I was doing it, I would save the video to YouTube and publish it “unlisted.” This way, only those of us who are granted the link can see it. Any one just viewing your YouTube Channel would not see it…

But I still think the cameras is stuttering by focusing it on a particular image and then quickly refocusing… the two flights were made under different circumstance, the first by you and the second, the Drone using way points

5

wall proximity isn’t the subject of this discussion. I’m investigating a repeatable 4K60 HEVC micro stutter, and all tests are performed in calm conditions to eliminate wind as a variable.

1 Like
6

I am not satisfied with DJI’s response.

I provided original 4K60 HEVC footage, exact timestamps, recordings made on internal storage, recordings made on different cards, and recordings made on the Lexar card DJI specifically recommended. I also checked the footage frame by frame with several applications and used independent AI analysis. The same suspicious points were identified repeatedly.

Despite all of that, DJI replied that their technicians found nothing.

That answer looks like the file was simply played once and the case was closed. This is not a serious technical investigation. An intermittent encoding or frame-timing issue cannot be dismissed after one normal playback.

I already did half of DJI’s diagnostic work myself: I excluded the memory card, tested internal storage, tested other cards, used the recommended Lexar card, supplied exact timestamps, and compared the footage on different devices.

DJI makes expensive cameras and drones, not cheap toys. Their technicians should analyse frame timing, duplicated frames, PTS/DTS, GOP structure and repeated playback conditions instead of giving customers a generic “no issue found” reply.

At this point, the problem is not only the micro-stutter. It is DJI’s careless attitude toward a customer who provided more than enough material for a proper investigation.

7

The original video is about 10 GB, and even the extracted problem section is around 3 GB. Uploading it to YouTube would be pointless because YouTube recompresses the footage and may hide or alter the exact issue being investigated.

DJI already provided me with dedicated storage, and I uploaded the original file there without YouTube butchering the quality. So no, YouTube is not the ■■■■■■■ solution here.

8

All I will say for now is that you have not eliminated the wall proximity from your “4K60 HEVC micro stutter”… Just saying so does not make it so… When you do find the cause, please publish your findings and solution.

9

During fast yaw or manual panning, the spatial and temporal complexity of the image increases sharply because most of the frame changes between consecutive frames.

In 4K60 10-bit HEVC, this creates a heavy load for the hardware encoder. With a limited bitrate and only 16.7 ms available per frame, the encoder may produce a difficult GOP, irregular frame sequence or unstable frame timing.

The file may remain technically valid, but the same section can appear as a micro-stutter depending on the decoder, buffering state and system load.

This would explain why the issue appears mainly during manual panoramic movement across highly detailed scenes, especially when moving between structured textures such as brick walls and chaotic detail such as foliage.

The correct way to verify this is to analyse GOP structure, frame sizes, PTS/DTS timing, duplicated or near-duplicated frames and frame-to-frame motion at the exact stutter location.

1 Like