Video seems to jump and seems to lag from time to time

Could you verify what mobile device and Go app version you're running?

Not sure why Jamie bothers to ask questions as the issues are never resolved and never hear back from him..video lag,loss of signal dji my a**e

Is this only at long range or all the time?

It's seems to be varied even while moving slowly
I was using android on my Samsung s7
But as the app has crashed switched to iPad

Probably your card is not fast enough for the recording, get a UHS 1 card, I have a sandisk extreme pro 32gb, great card for only £20

fansa84420ba Posted at 2017-5-8 14:17
It's seems to be varied even while moving slowly
I was using android on my Samsung s7
But as the app has crashed switched to iPad

How many Ram is free for DJI go? I suspect the Go app crashes,if your device is running with low free ram volume. Ios or(and)  android have this problem to slow down with low ram volume.

Bradders Posted at 2017-5-9 00:29
Probably your card is not fast enough for the recording, get a UHS 1 card, I have a sandisk extreme pro 32gb, great card for only £20

Please, the SD card has absolutely nothing to do with the issue. It is to do with image transmission and the ability of the device to decode the video stream.

Geebax Posted at 2017-5-9 01:28
Please, the SD card has absolutely nothing to do with the issue. It is to do with image transmission and the ability of the device to decode the video stream.

OP is talking about the recorded video, not streamed ?

Bradders Posted at 2017-5-9 10:12
OP is talking about the recorded video, not streamed ?

There is no such thing as lag on recorded video. And the SD card has got nothing to do with the problem,, it either records or not, it cannot skip frames.

Geebax Posted at 2017-5-9 16:37
There is no such thing as lag on recorded video. And the SD card has got nothing to do with the problem,, it either records or not, it cannot skip frames.

  Absolutely!

Geebax Posted at 2017-5-9 16:37
There is no such thing as lag on recorded video. And the SD card has got nothing to do with the problem,, it either records or not, it cannot skip frames.

Im sorry. You are unfortunately incorrect, if a recording lags e.g skips frames it is most likely due to the fact that the card cannot keep up with the write speed required, exactly why I suggested a faster UHS-I card.

"It either records or it does not"

So you get a warning saying "Card too slow, cannot record", nope... it skips frames

Bradders Posted at 2017-5-9 23:49
Im sorry. You are unfortunately incorrect, if a recording lags e.g skips frames it is most likely due to the fact that the card cannot keep up with the write speed required, exactly why I suggested a faster UHS-I card.

"It either records or it does not"

No, you unfortunately demonstrate a lack of understanding about the MPEG stream and recording. Certain types of recording, such as motion JPEG and CinemaDNG can skip frames because each frame in a sequence is completely self-standing frame of material, whereas in an MPEG sequence, each frame depends to a large degree upon the preceeding frame(s). Therefore, if the SD card is not fast enough, the recording stops, period. It cannot skip frames, because it would lose the relationship between between the frames, so it simply stops recording. This is a fundamental fact about MPEG recording.

This is known to happen in the Phantom in the earlier P3 series, when recording would simply stop if the card was not fast enough. However the later P4 and P4P systems are able to determine if the card is not fast enough and refuse to record in the first place. In fact, most modern cameras evaluate the capability of the SD card when it is inserted in the camera and if it is judged to be too slow, they warn you to replace the card with a faster one.

And I say again, 'lag' is a subjective term used by people to describe the delay between what is happening in front of the camera and the point at which it is displayed on the phone or tablet. It is again due to the MPEG encoding, which introduces a considerable delay between real time action and the displayed image. The length of the delay is heavily influenced by the processing ability of the receiving or decoding device. 'Lag' does not occur on the SD card, as a recording has no relevance to real time.

Excellent explanation.  So what's the cure ?

fans2d0e74f6 Posted at 2017-5-10 17:30
Excellent explanation.  So what's the cure ?

If you are referring to the 'lag', then the primary thing is to get a display device that has plenty of power to decode the MPEG stream. One with a dedicated hardware decoder is best. Few phones are good enough, the best I have found is the iPad Air 2.

I'm using the Samsung Tab A. I was told it would work fine but the video is glitchy.  

fans2d0e74f6 Posted at 2017-5-11 13:20
I'm using the Samsung Tab A. I was told it would work fine but the video is glitchy.

I don't know who told you that, but the Tab A is not among the list of devices recommended and tested by DJI.

Geebax Posted at 2017-5-11 14:38
I don't know who told you that, but the Tab A is not among the list of devices recommended and tested by DJI.

Actually it is on the list it's just not among the highly recommended devices.

fans2d0e74f6 Posted at 2017-5-13 13:31
Actually it is on the list it's just not among the highly recommended devices.

The tab A is something that I had already owned prior to purchasing the Phantom 4K. I had looked down the list and it was on the list just not one of the highly recommended highly recommended devices.

fans2d0e74f6 Posted at 2017-5-13 13:31
Actually it is on the list it's just not among the highly recommended devices.

That depends upon what list you are referring to. Several sites have their own lists, this is the official one for the P4 and I don't see the Tab A on it:

Android: Samsung S7 edge, Samsung S7, Samsung S6, Samsung S5, Samsung NOTE4, Samsung NOTE3, Samsung tabs 705c, Ascend Mate 9, Ascend Mate 7, Huawei Honor 8, Huawei P8 Max, Vivo X7, Xiaomi 5, MI 3, Google Nexus 6p, Nexus 9, Google Nexus 7 II, LG V20, Sony Xperia Z3, Smartisan T1.

Geebax Posted at 2017-5-10 03:00
No, you unfortunately demonstrate a lack of understanding about the MPEG stream and recording. Certain types of recording, such as motion JPEG and CinemaDNG can skip frames because each frame in a sequence is completely self-standing frame of material, whereas in an MPEG sequence, each frame depends to a large degree upon the preceeding frame(s). Therefore, if the SD card is not fast enough, the recording stops, period. It cannot skip frames, because it would lose the relationship between between the frames, so it simply stops recording. This is a fundamental fact about MPEG recording.

This is known to happen in the Phantom in the earlier P3 series, when recording would simply stop if the card was not fast enough. However the later P4 and P4P systems are able to determine if the card is not fast enough and refuse to record in the first place. In fact, most modern cameras evaluate the capability of the SD card when it is inserted in the camera and if it is judged to be too slow, they warn you to replace the card with a faster one.

Your a genius, I bow to your talent

Geebax Posted at 2017-5-9 16:37
There is no such thing as lag on recorded video. And the SD card has got nothing to do with the problem,, it either records or not, it cannot skip frames.

You're right

Geebax Posted at 2017-5-10 03:00
No, you unfortunately demonstrate a lack of understanding about the MPEG stream and recording. Certain types of recording, such as motion JPEG and CinemaDNG can skip frames because each frame in a sequence is completely self-standing frame of material, whereas in an MPEG sequence, each frame depends to a large degree upon the preceeding frame(s). Therefore, if the SD card is not fast enough, the recording stops, period. It cannot skip frames, because it would lose the relationship between between the frames, so it simply stops recording. This is a fundamental fact about MPEG recording.

This is known to happen in the Phantom in the earlier P3 series, when recording would simply stop if the card was not fast enough. However the later P4 and P4P systems are able to determine if the card is not fast enough and refuse to record in the first place. In fact, most modern cameras evaluate the capability of the SD card when it is inserted in the camera and if it is judged to be too slow, they warn you to replace the card with a faster one.

Your explanation is really interesting.
Unfortunately I have a similar problem, not on the video transmitted to the mobile phone, but on the video recorded on the SD card.
The video lacks some very short intervals, it is chopped.
I googled and found that many people complain about the same problem, in particular this link clearly shows the problem (it is not the only one):
https://www.youtube.com/watch?v=YuD3qG4hjms
As you can see the video is not smooth and fluid.
I've tried changing all the recording parameters, practically all the combination of resolutions and frame-rates, but the problem always arises in all cases.
If someone thinks that the SD card I am using is not suitable then we must also establish that a SunDisk Extreme 32GB V30 U3 is not suitable for use in the mavic air 2.
If someone thinks that my PC is the cause of the problem in the playback phase we should understand which system should be better performing than this
     * -memory size: 63GiB
     * -cpu product: Intel (R) Core (TM) i9-9880H CPU @ 2.30GHz
         size: 4446MHz, capacity: 4800MHz
     NVIDIA Quadro RTX 3000 Max-Q

Thanks for your suggestions

Robertodst Posted at 12-24 09:01
Your explanation is really interesting.
Unfortunately I have a similar problem, not on the video transmitted to the mobile phone, but on the video recorded on the SD card.
The video lacks some very short intervals, it is chopped.

I looked at your footage, and what you are seeing is an issue caused by the rolling shutter employed in the camera's sensor. It can be reduced by either slowing down the speed of your pan, and/or by using an ND filter on the front of the lens to reduce the light and then changing the camera shutter speed to around 1/50th of a second.

Geebax Posted at 12-24 13:12
I looked at your footage, and what you are seeing is an issue caused by the rolling shutter employed in the camera's sensor. It can be reduced by either slowing down the speed of your pan, and/or by using an ND filter on the front of the lens to reduce the light and then changing the camera shutter speed to around 1/50th of a second.

Thank you Geebax.
I can also try some filters and other shutter speeds, as you suggest, in order to increase the shutter opening time, but I ask you to consider the two issues separately:
1) the segmentation of fast movements in the presence of a very short shutter time. This is what happens if you take a video shoot like the blades of a rotating fan.
2) the segmentation of a video during a footage.
They are two different things, the first I would say is normal, and has a clear physical explanation. The second can only be explained by a problem at the SW level (in the best case) or at HW level (in the worst case).
In order to support what I write, I am showing you another video taken overnight and showing the same problem, so I don't think it can be solved by adding a filter, but, as I write before, I'll try this too.
Please watch this.
https://www.youtube.com/watch?v=yEaxyZCQa9w&t=283s
If you don't want to see the  whole video, skip to minute 3 for example, or watch only the tracking shots.
Your suggest to do slower tracking shots. Ok, this certainly reduces the annoyance of seeing the video segmentation, because we have the ability to mediate between the images we observe and we perceive less the intermittence of the video, in fact, at the limit, when the image is static you don't notice anything,
but that doesn't mean we're eliminating the defect. It's a bit like saying, when my car goes fast it is too much noisy, so I always go slow.

Robertodst Posted at 12-25 03:43
Thank you Geebax.
I can also try some filters and other shutter speeds, as you suggest, in order to increase the shutter opening time, but I ask you to consider the two issues separately:
1) the segmentation of fast movements in the presence of a very short shutter time. This is what happens if you take a video shoot like the blades of a rotating fan.

I looked at your footage, around 3:00, the part with the Ferris Wheel, but I could not see any issue there.