Mavic Pro-2 vs Zoom and P4P - Line Skipping + Sensor Heat?

[Cliff_622]

Sweet!!!! Your flying the M2 rather than pixel peeping.

Im really hoping people haven't been turned off the M2 by all the guesses as to what issues we might be dealing with.

As to the theories to date it seems apparent now the sensor overheating and line skipping/addition weren't right.

I have bounced some HQ MLOG test takes I shot to a friend who does post product for a living and asked him for an opinion on the quality. Will be interesting what he says- he is very familiar with the footage from the P4P. Now Im wondering how many cartons of beer I will need to part with to get him to make a LUT for me.

Actually, DroneDJ got an official DJI comms response to my video. In it, they confirm that a "special" pixel binning was used and that it does, in fact have a negative impact on image quality. They also go on to state that this was a necessary tactic and that it was the best compromise based on the "complexity" of the M2P design....but they refuse to say "why" it was necessary to do it. The ""why" they did it was the mysterious part. They don't deny or confirm a heat problem that I suspect.

DJI actually does not refute or deny any hypothesis in my video at all. In fact, they solidify several of my (our) findings and everybody else that is unhappy with FOV sub-sampled sensor scanning. It's "real" and not "imagined".

Again, this is all just a three paragraph; "yes,..but we HAD too" statement from DJI

Guys,...I love my M2P. I made 7 flights with it on Sunday. I took great construction site photos and flew entirely in 40mm HQ video mode. I got decent images that I'm reasonably satisfied with. No,...I don't fly in FOV 28mm and will not ever unless it's an emergency.

When used this way, I'm perfectly OK with my M2P. I do have a Zoom too that is great for inspections and a P4P V2.0 for when I need more serious imagery stuff.

Cliff Totten

 

[Lon Denard]

Wow, simply wow! You spent hours in this thread trying to disprove and mock any negative opinion over 4k quality of this drone and yet you were never that big into content and just started with Premiere!? What else can be said? Have fun and good luck!

What's really wow is that of your 4 posts, 3 are directed towards me! Hahaha!

Yea I've used Power Director up until now. The guy that basically started all this has back tracked quite a bit and check the join date. I'm not new by any means. I'm also on the 3DR forum and the GoPro forum because I own a Solo and Karma. Heck, I still have my Parrot AR Drone 1.0 and my Bebop.

This doesn't have anything to do with editing expertise. It's about what DJI promised and what they delivered. I believe that they delivered what they promised. Again, even the guy that basically started all of this has to agree that most of the overblown expectations were brought on by YouTube reviewers. Even the idea that the name "Hasselblad" meant that this drone was going to be superior to the P4P is ridiculous. Red is coming out with a phone, does everyone expect it to be as good as a Dragon? Or will it suffice that it's superior to other camera phones of its type? Some will say that it has to be better than a standard point and shoot, others a low end DSLR, while others will expect the moon and stars.

I took my first photography class in 1981 and have taken pictures all over the world. That doesn't mean that I'm an expert on digital video down scaling by any means but I do know what DJI said at the release event and I can read what was posted on their website. The bottom line is people ASSumed some things, believed things that others told them and are blindly believing what is currently being said by so called experts. While ignoring other equally so called experts. In any case, you don't have to be a doctor to know a bullet to the head can kill you.

 

[Lon Denard]

Actually, DroneDJ got an official DJI comms response to my video. In it, they confirm that a "special" pixel binning was used and that it does, in fact have a negative impact on image quality. They also go on to state that this was a necessary tactic and that it was the best compromise based on the "complexity" of the M2P design....but they refuse to say "why" it was necessary to do it. The ""why" they did it was the mysterious part. They do deny or confirm a heat problem.

DJI actually does not refute or deny any hypothesis in my video at all. In fact, they solidify several of my (our) findings and everybody else that is unhappy with FOV sub-sampled sensor scanning. It's "real" and not "imagined".

Again, this is all just a three paragraph; "yes,..but we HAD too" statement from DJI

Guys,...I love my M2P. I made 7 flights with it on Sunday. I took great construction site photos and flew entirely in 40mm HQ video mode. I got decent images that I'm reasonably satisfied with. No,...I don't fly in FOV 28mm and will not ever unless it's an emergency.

When used this way, I'm perfectly OK with my M2P. I do have a Zoom too that is great for inspections and a P4P V2.0 for when I need more serious imagery stuff.

Cliff Totten

Me too. I only fly in HQ since Mad RC's video. I never noticed anything wrong at all! Hahaha! Until your video and all of this hoopla, I had no idea anyone was anything but happy with the M2P. That's the honest truth. Now hundreds of people have returned their drones to DJI and YouTube reviewers have enjoyed a huge upswing in clicks. For what it's worth, I mean this sincerely, I believe that you only posted what you believe and I appreciate that you acknowledge that it wasn't DJI that "hyped" up the M2P camera to the stars, it was YouTube reviewers. In fact, some of the same ones that were praising the drone last week are condemning it now! They pulled all of the "fan boy" clicks and now they're reaping in the DJI hate squad clicks. Hahaha!

 

[Cliff_622]

Guys,....what hurts my feelings the most is that, throughout this entire thing, not even ONE single person mentioned how COOL my LIVING piece of Super Bowl Forty Four 40 yard line grass planter pot was. :-(

Cliff Totten

 

[SteveCA]

Actually, DroneDJ got an official DJI comms response to my video. In it, they confirm that a "special" pixel binning was used and that it does, in fact have a negative impact on image quality. They also go on to state that this was a necessary tactic and that it was the best compromise based on the "complexity" of the M2P design....but they refuse to say "why" it was necessary to do it. The ""why" they did it was the mysterious part. They don't deny or confirm a heat problem that I suspect.

DJI actually does not refute or deny any hypothesis in my video at all. In fact, they solidify several of my (our) findings and everybody else that is unhappy with FOV sub-sampled sensor scanning. It's "real" and not "imagined".

Again, this is all just a three paragraph; "yes,..but we HAD too" statement from DJI

Guys,...I love my M2P. I made 7 flights with it on Sunday. I took great construction site photos and flew entirely in 40mm HQ video mode. I got decent images that I'm reasonably satisfied with. No,...I don't fly in FOV 28mm and will not ever unless it's an emergency.

When used this way, I'm perfectly OK with my M2P. I do have a Zoom too that is great for inspections and a P4P V2.0 for when I need more serious imagery stuff.

Cliff Totten

I too only shoot in HQ for video which is pretty good. I watched your entire video (long version) and found it highly informative. I was very surprised to find that the resolving ability of the MP 2 pro was not even as good as the Mavic air. Could the lens be at fault? What aperture were the test done at? In my unscientific tests, I found the lens is pretty bad at apertures other than 4.0-5.0. Could the lens also be part of the problem?

 

[Roamer105]

The Phantom 4 Pro uses an Ambarella H1. The Mavic 2 Pro is still unclear what image processor they are using. I searched Ambarella and found their quarterly earnings report say that DJI has moved on from Ambarella and is now developing their own video processors. Did DJI create a new image processor on their own? If so,...that could be dangerous! lol Who knows what video processor SOC problems they are having? Did they develop a chip that is already maxed out? If so, I bet Ambarella is laughing their tails off at DJI right now! ( I can her them saying to DJI,..see?,...you never should have left us!" lol

The plot thickens......

Rather than speculating... why not opening the case and take a look? If it is such an issue to you, maybe it would be good therapy.

Btw: A 1” sensor is just a description for the item which in reality has probably only 16mm diagonal size or just a little less since it is supposed to be a Sony sensor. The description goes back to vidicon tubes where a 1” tube had a 16mm diagonal sensor surface.

 

[Roamer105]

I too only shoot in HQ for video which is pretty good. I watched your entire video (long version) and found it highly informative. I was very surprised to find that the resolving ability of the MP 2 pro was not even as good as the Mavic air. Could the lens be at fault? What aperture were the test done at? In my unscientific tests, I found the lens is pretty bad at apertures other than 4.0-5.0. Could the lens also be part of the problem?

It is not the lens or sensor. Just test the (raw) still performance it makes that case real clear. The still resolution due to the outstanding lens quality is way better than any other prosumer drone. I tested stills against high quality SLRs with professional lenses and frankly the little Hasselblad gave them a good across the whole spectrum of aspects.

 

[prothorin]

Which part of the camera made by Hasselblad?

 

[Cliff_622]

Which part of the camera made by Hasselblad?

The lens filter is made by "Hasselblad".

 

[Sockeye]

Actually, DroneDJ got an official DJI comms response to my video. In it, they confirm that a "special" pixel binning was used ......

Are you sure? As the quoted text in that blog just quotes 1:1 "Blade-Strikes" (DJI support) posting in the RC Groups from about a week ago... I wouldn't take anything he says as an official statement from DJI.

RC Groups - View Single Post - Official DJI Mavic 2 ***Owner Thread***

 

[BlueBay]

DJI claim 'iconic image quality' and 'see the bigger picture'.

We all know by now that this doesn't seem to be the case, as I said, work in progress.

Here is the lovely video:

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[FiShBuRn]

What about this guy? And the info he demonstrate? DJI didnt offer him a drone....

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[BlueBay]

What about this guy? And the info he demonstrate? DJI didnt offer him a drone....

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

Yes, that's the best synopsis to date. Everyone should watch this to recalibrate this entire journey of discovery with the M2P.

 

[kilomikebravo]

EXCELLENT TIP! Thanks for the link.

 

[Lon Denard]

I mentioned

I too only shoot in HQ for video which is pretty good. I watched your entire video (long version) and found it highly informative. I was very surprised to find that the resolving ability of the MP 2 pro was not even as good as the Mavic air. Could the lens be at fault? What aperture were the test done at? In my unscientific tests, I found the lens is pretty bad at apertures other than 4.0-5.0. Could the lens also be part of the problem?

I asked about lensing as well but I'm not sure there was an answer.

 

[Sockeye]

What about this guy? And the info he demonstrate? DJI didnt offer him a drone....

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

He must be rather blind, not seeing a detail difference between f2.8 and f11, nor FOV vs HQ...

 

[dwallersv]

Sigh...can there be no sanctuary from people insulting our president? Maybe they are acting like sanctimonious know it alls, very Obama-like?

At great risk, I will publicly reveal that I voted for him, and am to this day a strong supporter.

He also is a Horse's ***. Don't have to be a nice guy to implement good policy.

 

[DidJI]

Which part of the camera made by Hasselblad?

the text "Hasselblad" written on front, in big font

 

[sar104]

At great risk, I will publicly reveal that I voted for him, and am to this day a strong supporter.

He also is a Horse's ***. Don't have to be a nice guy to implement good policy.

Very true. Although it's not generally good policy that comes out of a horse's... well never mind - stray thought.

 

[WithTheBirds]

Actually, DroneDJ got an official DJI comms response to my video. In it, they confirm that a "special" pixel binning was used and that it does, in fact have a negative impact on image quality. They also go on to state that this was a necessary tactic and that it was the best compromise based on the "complexity" of the M2P design....but they refuse to say "why" it was necessary to do it. The ""why" they did it was the mysterious part. They don't deny or confirm a heat problem that I suspect.

DJI actually does not refute or deny any hypothesis in my video at all. In fact, they solidify several of my (our) findings and everybody else that is unhappy with FOV sub-sampled sensor scanning. It's "real" and not "imagined".

Again, this is all just a three paragraph; "yes,..but we HAD too" statement from DJI

Guys,...I love my M2P. I made 7 flights with it on Sunday. I took great construction site photos and flew entirely in 40mm HQ video mode. I got decent images that I'm reasonably satisfied with. No,...I don't fly in FOV 28mm and will not ever unless it's an emergency.

When used this way, I'm perfectly OK with my M2P. I do have a Zoom too that is great for inspections and a P4P V2.0 for when I need more serious imagery stuff.

Cliff Totten

Yes- there is a hardware limitation- the sensor is 5472px wide and 4K= 3840.

Where does the DJI response say anything about special binning? They talk about subsampling, seemingly after sensor readout....

The FOV mode looks fantastic with a bit of post. Nobody will see an issue when its transcoded out into a final clip.