DJI GO In-Fly Life Video Pixelation/Smearing

[wingbotics]

Hi,

This happened right after having been taking off. It'd lasted for about 20 seconds and it was only cleared after I did a camera tilting downwards perpendicularly. The action was likely to trigger an H.265/H.264 I-frame refresh to flush the bad P-frames. Did anyone experience this? Why? What could cause the damaged frames? It is clearly a video codec bug in the M2P system. Yes, I'm on the latest software. I'd never seen the problem in the Mavic Pro.
At the time of the problem, neither did I do recording nor capturing a snapshot. Too bad.

Thanks in advanced.

 

[Thomas B]

If it happens again please try to capture. If not, let us know. Might be one “on of those things”... a random software load/operation occurrence.

 

[Not A Speck Of Cereal]

If this is a video issue, Wingbotics would already have footage.

If not, then we could simply be talking about a transmission glitch of the video feed to the controller (not recorded video).

Chris

 

[wingbotics]

I don't think that it is conclusive to say that it was the feed to the controller or the source, i.e. the drone. It could've been either

1. the encoding side (drone side),
2. the transmission packet drops,
3. the decoding side (the controller as the receiver) or
4. the iPad.

I did not enable the recording at the time of taking off. I noticed the bad video on my iPad after it ascended and was hovering about 200 feet above my home point. I attempted a little panning but the video was very much pixelated then my fingers were faster than my brain and had my camera pointing down instantly. By then the smearing was cleared and the video was very good. This occurred on Sunday (a few days back).

Yes, I will prepare my cell phone and also will start with recording in my next fly sessions to hopefully capture this. I will post if I have one.

Thank you.

If this is a video issue, Wingbotics would already have footage.

If not, then we could simply be talking about a transmission glitch of the video feed to the controller (not recorded video).

Chris

 

[Not A Speck Of Cereal]

Okay, so it's not video recording. The "H.265/H.264 I-frame refresh" comment in your OP is not relevant to live video feed between the controller and craft (so the comment confused the subject).

The OcuSync 2 video transmission should not have such glitches, especially at close range, unless a) you have strong interference, or b) possibly have a bad antenna orientation (if the craft is directly above you, make the antenna parallel to the ground so the flat surface, not the tip, is point straight up).

As you say, it could be either end of the signal (transmission / reception). If you've made modifications to the aircraft (even skin), that could be suspect.

If those ideas are not of any help, you may have a defect and should consider contacting DJI service.

Chris

 

[wingbotics]

Ok, this time it happened at the last phase of home landing time and I did take a snapshot of my iPad screen as seen in the attachment once the landing was complete. The pixelation was very bad before I took the picture but you can see the bottom of the frame. At the moment of the picture taking, the M2P was 8 feet from the controller. I've not yet check the video recording during the fly. I'll follow up later on this.

Thanks.

 

[Not A Speck Of Cereal]

It would be interesting of that image pixelation was on the recorded video.

If it's not on the video, the transmission pixelation is no big deal, but I wonder if it couldn't be related to interference near the porch I see in the screen shot. Are there electrical lines running under those boards? Do you have a home security system. Are there any kind of strong wireless transmissions emanating from the house?

Chris

 

[wingbotics]

I'll check the recording later.

As for your questions, I have to say that I flew the Mavic Pro just about the same time and had absolutely no issue at all. So why did the Mavic 2 Pro experience the signal problem resulting the pixelation.

Specific to your questions:
1. Home security: yes.
2. Power lines underneath the deck: no.
3. Wireless transmission emission: yes and 30ft away.

But again why is the M2P so sensitive?

Thank you Chris.

It would be interesting of that image pixelation was on the recorded video.
...
Are there electrical lines running under those boards? Do you have a home security system. Are there any kind of strong wireless transmissions emanating from the house?

Chris

 

[Not A Speck Of Cereal]

I think it's just the nature of radio transmission to not be absolutely impervious to surrounding / nearby interference.

If it happened all the time, like during all / most of a flight, it would be a bigger concern to me. But I've only experienced it when I was around a lot of stuff, like the downtown city waterfront (as compared to out in the country with no potential electrical / magnetic sources of interference) and even then it was only momentary.

The good news is that it shouldn't effect the recording of video from the camera to the SD card (no transmission involved in that, of course). So if it doesn't interrupt / disrupt the flight, I think you'll find that DJI would tell you that it's not unusually to get the occasional glitch.

Here's hopin' someone else chimes in -- I'm not an electrical expert, I'm just speaking from experience.

Chris

 

[wingbotics]

It took me a bit long to get back to and conclude this topic.

First of all, I examined the recording when the pixelation occurred in my Mavic 2 Pro. Yes, the recording is normal and does not show any smearing effect.

In my earlier reporting, I said that I did not see the problem in Mavic Pro. However, I finally reproduced the issue on Mavic pro similar to Mavic 2 Pro. The way I reproduce it (from the same takeoff point) was:

1. No compass calibration (to start with and perhaps not related) was performed
2. Took off and flew 10+ mins. Live video feed was good.
3. Hand-caught to land.
4. Took off by handheld approach (of the same location) and flew a few minutes not more than few hundred feet away and a couple hundred feet in height. Live video feed was good.
5. Auto landed.
6. Took off again and then the video on my iPad was very tardy (no smearing with a little blockiness of video).
7. I continued to fly for another few minutes with keeping LOS (within 200-300 feet away). During this time, video was very tardy and jerky.
8. I hand-caught to land the M1P.

Note, in between landing and taking off, I did not shut down M1P, the controller nor the iPad. I believe the observed behavior was perhaps random and due to my home radio trransmitter (more or less of 10 feet away and 30 feet above).

So this is a close issue to me.

Thank you all.

I think it's just the nature of radio transmission to not be absolutely impervious to surrounding / nearby interference.
...
The good news is that it shouldn't effect the recording of video from the camera to the SD card (no transmission involved in that, of course). So if it doesn't interrupt / disrupt the flight, I think you'll find that DJI would tell you that it's not unusually to get the occasional glitch.
...
Chris

 

[Thomas B]

It took me a bit long to get back to and conclude this topic.

First of all, I examined the recording when the pixelation occurred in my Mavic 2 Pro. Yes, the recording is normal and does not show any smearing effect.

In my earlier reporting, I said that I did not see the problem in Mavic Pro. However, I finally reproduced the issue on Mavic pro similar to Mavic 2 Pro. The way I reproduce it (from the same takeoff point) was:

1. No compass calibration (to start with and perhaps not related) was performed
2. Took off and flew 10+ mins. Live video feed was good.
3. Hand-caught to land.
4. Took off by handheld approach (of the same location) and flew a few minutes not more than few hundred feet away and a couple hundred feet in height. Live video feed was good.
5. Auto landed.
6. Took off again and then the video on my iPad was very tardy (no smearing with a little blockiness of video).
7. I continued to fly for another few minutes with keeping LOS (within 200-300 feet away). During this time, video was very tardy and jerky.
8. I hand-caught to land the M1P.

Note, in between landing and taking off, I did not shut down M1P, the controller nor the controller. I believe the observed behavior was perhaps random and due to my home radio trransmitter (more or less of 10 feet away and 30 feet above).

So this is a close issue to me.

Thank you all.

Thanks for resolution to the issue.