DJI Mavic, Air and Mini Drones
Friendly, Helpful & Knowledgeable Community
Join Us Now

Have you seen similar problems on your MM video? (only respond if you have a mavic mini)


  • Total voters
    79
With the exception of the mystery prevailing artefacts at the extreme bottom of frame there seeming isn't a lot to grizzle about here, particularly having regard to the price point for the overall package.

Even the corners at 100% view aren't too bad for a small cheap lens illuminating a tiny sensor.

The artefact mystery is intriguing. We know the mini employs a 3 axis gimble- I wonder if DJI is employing the digital stabilisation which forms part of the feature set of the Ambarella SOC and not cropping adequately before encoding.

The excitement about this issue seems to be growing so hopefully someone from from DJI might buckle and provide some clues.

I certainly feel looking at my footage that it could be a cropping issue and I suspect could be "cured" by software - it only is an issue in the bottom 10 or so rows of the image and to be honest if that was just cropped off completely I wouldn't miss it..

Apart from that the video is generally decent. A little soft at the top right hand corner and over-sharpened across the whole image, but certainly acceptable quality given the price point.
 
  • Like
Reactions: F22arecool
I certainly feel looking at my footage that it could be a cropping issue and I suspect could be "cured" by software - it only is an issue in the bottom 10 or so rows of the image and to be honest if that was just cropped off completely I wouldn't miss it..

Apart from that the video is generally decent. A little soft at the top right hand corner and over-sharpened across the whole image, but certainly acceptable quality given the price point.
I reckon id be hanging onto if it seems to be a common issue. Good chance whatever the factory fix ultimately is they will be able to roll it out in a firmware update.
 
  • Like
Reactions: F22arecool
I certainly feel looking at my footage that it could be a cropping issue and I suspect could be "cured" by software - it only is an issue in the bottom 10 or so rows of the image and to be honest if that was just cropped off completely I wouldn't miss it..

Apart from that the video is generally decent. A little soft at the top right hand corner and over-sharpened across the whole image, but certainly acceptable quality given the price point.
The most recent firmware update did say that they improved video stability and smoothness. In footage from beta software I cannot find any distortion so maybe that's it.
 
  • Like
Reactions: andrei193
Does anybody here know someone who works for DJI?!?!!?? Maybe we can ask them to look into it ??
 
Does anybody here know someone who works for DJI?!?!!?? Maybe we can ask them to look into it ??
You are probably really looking for someone inside Ambarella. You may have some success in getting clues looking for similar issues with products using the H22 SOC.
 
  • Like
Reactions: F22arecool
The most recent firmware update did say that they improved video stability and smoothness. In footage from beta software I cannot find any distortion so maybe that's it.

this is what is says:
  • Optimized the 2.7K video shooting experience, and optimized video recording quality.
  • Optimized image distortion to give images a more natural appearance.
You may be on to something here. That image distortion optimization might have caused more problems than it fixed. Does your footage after the update has this distortion? If any other people have footage before the update can you please verify?

This could very well be a software fix after all.
 
this is what is says:
  • Optimized the 2.7K video shooting experience, and optimized video recording quality.
  • Optimized image distortion to give images a more natural appearance.
You may be on to something here. That image distortion optimization might have caused more problems than it fixed. Does your footage after the update has this distortion? If any other people have footage before the update can you please verify?

This could very well be a software fix after all.
I know on youtube this dude named billy got the drone before the official release date. He posted a video with sample raw footage taken and there is a link to a google drive in the comments. I can't really judge it because I am on mobile but look there.
Edit: The link is here
I hope this can help.
 
  • Like
Reactions: andrei193
I know on youtube this dude named billy got the drone before the official release date. He posted a video with sample raw footage taken and there is a link to a google drive in the comments. I can't really judge it because I am on mobile but look there.
Edit: The link is here
I hope this can help.
I don't see the problem in the old footage but he has very little footage with the new firmware in a video and is above water and can't really tell if the distorsion happens. I think he lives in the USA and I have a hunch that the FCC versions are not affected.
 
I know on youtube this dude named billy got the drone before the official release date. He posted a video with sample raw footage taken and there is a link to a google drive in the comments. I can't really judge it because I am on mobile but look there.
Edit: The link is here
I hope this can help.

It is present if you download the raw files he has uploaded (very evident on the first one of the ship (0050.MP4)). If you just view his YouTube uploads you can barely see it.

I am beginning to think its a universal problem - just some people aren't seeing it (maybe viewing on a lower res screen or its in screen overscan area or just looking at youtube footage which largely seems to crop it off)
 
I read on the DJI forums on different topics about a firmware refresh (like a manual install). Does anyone know how to do it? Can somebody who knows try and see if it solves this problem?
 
Just did a firmware refresh and the problem is still there.
From what I can gather my theory is this: As there seem to be different hardware variations also the firmware must be different in terms of coding for each of the versions. Initially I tought that it might be a hardware problem but as the distortion does not manifest consistently, I believe that somehow the new image distortion optimization present in the new firmware affected the european version the most. That might explain why I can't seem to find any conclusive FCC version with this problem. Also it does not manifest in stills so I think a warped lens or sensor is not an explanation.
 
Just did a firmware refresh and the problem is still there.
From what I can gather my theory is this: As there seem to be different hardware variations also the firmware must be different in terms of coding for each of the versions. Initially I tought that it might be a hardware problem but as the distortion does not manifest consistently, I believe that somehow the new image distortion optimization present in the new firmware affected the european version the most. That might explain why I can't seem to find any conclusive FCC version with this problem. Also it does not manifest in stills so I think a warped lens or sensor is not an explanation.
What evidence do you have of variation in hardware? DJI has said the model differences (compliance with regional RF standards) are implemented through firmware flashed at the factory. No difference in hardware between FCC/CE.
 
What evidence do you have of variation in hardware? DJI has said the model differences (compliance with regional RF standards) are implemented through firmware flashed at the factory. No difference in hardware between FCC/CE.
What I know is that DJI said It's not a matter of firmware lock. It's hardwired. Now if I am wrong I am wrong, I said it's only a theory.
 
What I know is that DJI said It's not a matter of firmware lock. It's hardwired. Now if I am wrong I am wrong, I said it's only a theory.
It’s not about being wrong- it’s about us understanding what the reality is. You may be correct. In case you haven’t seen it this is DJI’s statement made to the FCC in seeking compliance certification.
 

Attachments

  • 097B2979-2E28-44FE-B2B4-9A886443681B.jpeg
    097B2979-2E28-44FE-B2B4-9A886443681B.jpeg
    378.4 KB · Views: 18
It’s not about being wrong- it’s about us understanding what the reality is. You may be correct. In case you haven’t seen it this is DJI’s statement made to the FCC in seeking compliance certification.
This actually does not prove that the FCC version (model MR1SS5) is the same hardware wise with the CE version (model MR1SD25 ). By the way, it's not even about the hardware, is the fact that these different models most likely have different firmware versions.
 
This actually does not prove that the FCC version (model MR1SS5) is the same hardware wise with the CE version (model MR1SD25 ). By the way, it's not even about the hardware, is the fact that these different models most likely have different firmaware versions.
Fair enough- it seemed you were sighting hardware differences. It seems almost certain the factory firmware is different fire FCC and CE.
 
Fair enough- it seemed you were sighting hardware differences. It seems almost certain the factory firmware is different fire FCC and CE.
Sorry for the confusion.
Check this out:
Mavic mini videos keep popping on my youtube feed and most of them are from USA and no sight of the problem on any of them. Take a look at this video:
To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.
from 4:42 to 4:44. keep your eyes on the river. The algorithm that compensates the distortion when rotating is correctly applied and is exactly in the region where most of us notice the ripple effect but we mostly see it at the bottom and top edges because thats where the strecthes are noticed when apllied for barrel distortions and maybe that's why is not consistent (sometimes the distortion is at the botom, sometimes is at the top and sometimes is on both the top and bottom), because the algorithm fails to understant what needs to be corrected and how in moving from left to right and vice versa.
 
Last edited:
  • Like
Reactions: F22arecool
I too have a problem with 2.7k video on the new Mavic Mini but it’s a different problem to yours. When you film in 2.7k and either pan left/right or fly to the left or right a ripple effect or artefact appears across both the top and bottom of the video clip. It’s hardly noticeable in 1080p but stands out like a sore thumb in 2.7k. Lots of people have reported the same problem but as yet we haven’t had a definitive answer from DJI what’s causing the problem or if it can be fixed. Some drone owners have already exchanged their drone only to find the problem is still there. Very frustrating. It’s not what I expected given that I own other DJI products.
 
  • Like
Reactions: F22arecool
I'm having exactly the same issue here, CE version (UK), with no cage, not a windy day, simply Quickshots mode zooming out, the problem is very noticeable. DJI please explain the problem to all of us, thanks.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.
 
Lycus Tech Mavic Air 3 Case

DJI Drone Deals

New Threads

Forum statistics

Threads
134,445
Messages
1,594,852
Members
162,983
Latest member
Roel Hopstaken