For work reasons I am restricting myself (just this once!) to facts, not opinions.
I first set up the Mavic, controller and an iPad
Mini 4 and made sure video transmission was working as expected, i.e. lag was in the milliseconds range.
I then powered everything down, disconnected the iPad and connected the
Moverio running the same (latest) version of DJI Go.
* Despite spending an hour trying every option available in the DJI software, the lowest lag I obtained was about 8 seconds at the beginning of the test.
* This lag slowly increased to about 20 seconds.
* There were occasional and periodic dropouts to black and digital breakup noise.
I might well be wrong, but my guess is that the
Moverio hardware is decompressing the OcuSync video stream in software and hasn't finished decompressing frame
n when
n+1 comes along. At this point one strategy for the software would be to drop frame
n+1 and continue with
n+2, however I do not think this is occurring. I think the buffer just gets bigger until a size limit is hit, which is when visual artifacts appear. At that stage the lag probably goes back to a lower value and starts increasing again.
All of this is a shame because the display portion of the
Moverio system is pretty good, but (in this case) it is let down by the processing hardware, specifically it seems as if a GPU is required to efficiently decompress an OcuSync stream.
I decided not to fly the Mavic using the
Moverio system since real-world results seemed unlikely to be superior to the (literal) lab test, and it was snowing and I hadn't had lunch.