Last bets regarding image quality vs 2S and Mavic 3

[Chaosrider]

Hyperlapse aside, the Air2s is going to more stable and faster.

I have to scratch my head about portability. I do get that the smallest size and weight is important to some...

But last summer we were all up and down the East Coast. We went for a wedding and visit the kids and friends. I had to make a choice, take my Mini 2 or my M2P. Knowing the visual possibilities I chose the M2P. I loaded it, the controller, 3 batteries, ND filters and anything I could think of into the M2P's FMC bag. Chargers got packed seperately with my other stuff. in my backpack (which the M2P case fit into when I only wanted 1 bag to carry. Not one time was I bothered or did I with I'd brought the lighter drone. Now, an Air2s is almost half the weight of the M2P, though the brick they call a controller adds weight and bulk. I carried that thing on my shoulder for almost two weeks not even thinking about it. That's just me. YMMV

I'm with you.

For me, the whole portability thing is a gimmick. The folding legs are just an annoyance. I have a hard case for my Minis that lets me transport them with the legs unfolded. Not that much bigger than the soft case that comes with the Mini-2 FMC.

My guess, and that's all it is, is that portability is more important outside the US. In the US, if you put another bag in your car...really, so what?

 

[Chaosrider]

But when the Mini 2's (and DJI's) "standard" controller was presented this case was no longer adequate.

I think the Mini-2 controller is vastly superior to the Mini-1, Mini-SE controller.

As far as the "screen under the sticks" configuration...it conjures up the image of flying a conventional fixed wing aircraft, and having to look down toward the floor to look out the window...lame!

Is there anyone here who prefers the "screen under the sticks" configuration? If so, what is it about that configuration that you find appealing?

Thx

 

[vindibona1]

I think the Mini-2 controller is vastly superior to the Mini-1, Mini-SE controller.

As far as the "screen under the sticks" configuration...it conjures up the image of flying a conventional fixed wing aircraft, and having to look down toward the floor to look out the window...lame!

Is there anyone here who prefers the "screen under the sticks" configuration? If so, what is it about that configuration that you find appealing?

Thx

The position of the screen above or below is neither here nor there to me. There is an advantage of the screen below. You raise your hands a few more inches to bring the screen into view and you can see your hands and the controls BETTER at the same time and if aligned with the drone you have all three in view.

Is there really a reason to hold a brick while you're flying? Seriously. More packing space needed. Why? Aside from limited programmable buttons, the dead spots in the new "standard" controller are there whether you notice them or not. It is most noticeable when flying close to narrow spots, particularly if you want to keep speed down. I don't notice any dead spots in the M2's controller. Have you tried flying VR/FPV with any of the DJI camera drones? (Litchi or DroneVR) Those extra programmable buttons make it so much more convenient.

Be honest. The M2's controller is a premium controller. The standard controller that comes now with all the DJI drones up to $3000 is... basic.

 

[Chaosrider]

The position of the screen above or below is neither here nor there to me. There is an advantage of the screen below. You raise your hands a few more inches to bring the screen into view and you can see your hands and the controls BETTER at the same time and if aligned with the drone you have all three in view.

Sorry, I'm not tracking here. I can't imagine why I would ever want to look at my hands while flying. Perhaps it's related to the fact that I learned touch-typing in the 8th grade...

With the Minis, my hands have vanished into the background, and I can align the screen and the drone as much or as little as I like, with no hands in the field of view for clutter.

What's the payoff to being able to see your hands?

One time after an FPV flight, I looked down at the controller to turn it off, and was briefly surprised that I couldn't see it...

;-0

Is there really a reason to hold a brick while you're flying? Seriously.

I never have. Seriously.

I did have a literal lead brick for a while, which I acquired from Caltech. It was called a "Sync Brick". At parties, a test of manhood was how long you could hold it out directly in front of you...

But...seriously...if I'm going to use a tool to control something with my hands, I want that tool to have some heft. Give me an 800 lb motorcycle to control, not a micro-light little screwdriver!

When I wanted to be in the Air Force when I was a little kid, I always wanted to fly bombers, not fighters. The B-52 is my actual favorite.

I have no interest in flying at low speed in tight spaces.

Be honest.

I always am. Annoying, perhaps, but honest!

An odd choice of words...

The M2's controller is a premium controller. The standard controller that comes now with all the DJI drones up to $3000 is... basic.

"Premium" is only meaningful if you're getting something useful or desirable above the "standard". For my interests, coming to droning from the pilot side, I just don't see it.

YMMV

 

[test2000Anafi]

I think the Mini-2 controller is vastly superior to the Mini-1, Mini-SE controller.

As far as the "screen under the sticks" configuration...it conjures up the image of flying a conventional fixed wing aircraft, and having to look down toward the floor to look out the window...lame!

Is there anyone here who prefers the "screen under the sticks" configuration? If so, what is it about that configuration that you find appealing?

Thx

Probably so that they can cover their neighbours body on display if her husband sneaks up on the pilot? I also prefer above btw

 

[EsaT]

But when I did this with the Nano+ there is some weird stuff going on at the pixel level of that drone and it's not great and I think the Mini 3 will be the same.

Quad Bayer design itself has far worser than advertised pixel count colour resolution:
Instead of individual pixels missing two of three colour channels, 2x2 pixel size chunks are missing that data.
Hence demosaicing has far more guessing to do.
Whole design is really intended for producing images of quarter of advertised MP count.

And on top of that instead of standard green colour using Bayer filter Autel went for "RYYB" filter replacing those two green pixels with yellow pixels.
On the basis that yellow filter lets in both green and blue light onto those pixels increasing efficiency in capturing light.
What marketing fails to tell is that green channel matches closely luminance/details our vision senses and there's contaminated mess in place of green colour channel data.
So there's need for some complex processing and some kind "green" colour data needed to produce usable colour image is gotten only after mutilation of signal of those yellow pixels.
Combine that to Quad Bayer design and it's no wonder if pixel level data is riddled with both colour resolution and colour accuracy issues.


Situation which isn't that different from Foveon sensor:
Foveon uses "stacked" RGB subpixels based on the idea of different wavelengths penetrating different distance into silicon.
(short wavelength least penetration, long wavelength best penetration)
Theoretically that gives 100% light capture by completely avoiding the need for RGB filter.
(besides 100% colour resolution with all colours captured per pixel)

But in reality that red wavelength photon is more likely to be captured by either "blue" or "green" photosite than "red" photosite at bottom of the stack.
Also that red photosite receives some green wavelength photons.
Of course also middle of stack "green" photosite captures some red photons, but it also gets blue photons.
And "blue" photosite at top obviously captures little bit of red and more than little green photons.
So once that mess has been cleared up to get usable colour data, final efficiency is actually worser than with Bayer sensor.


That's the reason why RGB filter Bayer sensor keeps dominating camera markets despite of filter throwing away light and colour resolution being less than pixel count.

 

[test2000Anafi]

Quad Bayer design itself has far worser than advertised pixel count colour resolution:
Instead of individual pixels missing two of three colour channels, 2x2 pixel size chunks are missing that data.
Hence demosaicing has far more guessing to do.
Whole design is really intended for producing images of quarter of advertised MP count.

And on top of that instead of standard green colour using Bayer filter Autel went for "RYYB" filter replacing those two green pixels with yellow pixels.
On the basis that yellow filter lets in both green and blue light onto those pixels increasing efficiency in capturing light.
What marketing fails to tell is that green channel matches closely luminance/details our vision senses and there's contaminated mess in place of green colour channel data.
So there's need for some complex processing and some kind "green" colour data needed to produce usable colour image is gotten only after mutilation of signal of those yellow pixels.
Combine that to Quad Bayer design and it's no wonder if pixel level data is riddled with both colour resolution and colour accuracy issues.


Situation which isn't that different from Foveon sensor:
Foveon uses "stacked" RGB subpixels based on the idea of different wavelengths penetrating different distance into silicon.
(short wavelength least penetration, long wavelength best penetration)
Theoretically that gives 100% light capture by completely avoiding the need for RGB filter.
(besides 100% colour resolution with all colours captured per pixel)

But in reality that red wavelength photon is more likely to be captured by either "blue" or "green" photosite than "red" photosite at bottom of the stack.
Also that red photosite receives some green wavelength photons.
Of course also middle of stack "green" photosite captures some red photons, but it also gets blue photons.
And "blue" photosite at top obviously captures little bit of red and more than little green photons.
So once that mess has been cleared up to get usable colour data, final efficiency is actually worser than with Bayer sensor.


That's the reason why RGB filter Bayer sensor keeps dominating camera markets despite of filter throwing away light and colour resolution being less than pixel count.

If a quad bayer configuration is used for the dual iso or dual exposure, is it worth it then ? If you are only shooting around 4000 horizontal pixels.

so 12 MP standard rgb vs 12 MP quad bayer (48MP), which has the best dynamic range, lowest noise or both?

 

[EsaT]

Is there anyone here who prefers the "screen under the sticks" configuration? If so, what is it about that configuration that you find appealing?

That keeps remote perfectly balanced in hand and very comfortable to hold for long times.
Having screen moved above sticks makes remote forward weighted.
Unless other end is ballasted with extra weight, or made longer.

 

[EsaT]

so 12 MP standard rgb vs 12 MP quad bayer (48MP), which has the best dynamic range, lowest noise or both?

If sensor sizes are identical and lens is same, both receive same amount of photons.
Variations in that would favour one.

After that we would need to find sensors using all the same level technical tricks.
Not exactly fair comparison if one sensor uses some years old design and other all the latest design tricks.
And with standard compact digicams having pretty much died years ago have the feeling of traditionally their sensors having been neglected in R&D with anything phone camera related having far higher priority.

 

[test2000Anafi]

If sensor sizes are identical and lens is same, both receive same amount of photons.
Variations in that would favour one.

After that we would need to find sensors using all the same level technical tricks.
Not exactly fair comparison if one sensor uses some years old design and other all the latest design tricks.
And with standard compact digicams having pretty much died years ago have the feeling of traditionally their sensors having been neglected in R&D with anything phone camera related having far higher priority.

The one with dual ISO is favoured is what you are saying? Everything else being identical.

I believe the sensor/aperture/lens combo is too good on the mini 3 pro. You see this in the footage, he is under exposing everything. it needs to ND filters or variable aperture for sure. I don’t want to add motion blur in post lol

Not exactly fair comparison if one sensor uses some years old design and other all the latest design tricks.

I never said it was fair, I only said it would beat the air2s magic 1inch sensor in image quality and low light. How it does it, changes nothing about the result

 

[Bussty]

The one with dual ISO is favoured is what you are saying? Everything else being identical.

I believe the sensor/aperture/lens combo is too good on the mini 3 pro. You see this in the footage, he is under exposing everything. it needs to ND filters or variable aperture for sure. I don’t want to add motion blur in post lol


I never said it was fair, I only said it would beat the air2s magic 1inch sensor in image quality and low light. How it does it, changes nothing about the result

Really want to see the comparison using stills...

 

[Rilot]

I'm really hoping it's going to be a 3:2 sensor aspect ratio for stills like the Air2S, P4P, M2P. Unlikely though. Expecting 4:3 like the Minis, Mavic 3, Air 2, etc. which will be a shame.

 

[RogerDH]

I think the Mini-2 controller is vastly superior to the Mini-1, Mini-SE controller.

As far as the "screen under the sticks" configuration...it conjures up the image of flying a conventional fixed wing aircraft, and having to look down toward the floor to look out the window...lame!

Is there anyone here who prefers the "screen under the sticks" configuration? If so, what is it about that configuration that you find appealing?

Thx

I really prefer the Fimi X8 SE controller, which expands sideways to put the device in the middle: It's always perfectly balanced.

 

[test2000Anafi]

I really prefer the Fimi X8 SE controller, which expands sideways to put the device in the middle: It's always perfectly balanced.

And if you need to use the touchscreen?

 

[RogerDH]

And if you need to use the touchscreen?

Well, if you can explain why you think that's a problem, maybe I can explain why it's not.

 

[test2000Anafi]

Well, if you can explain why you think that's a problem, maybe I can explain why it's not.

Guess not everybody uses iPad mini

 

[Chaosrider]

Probably so that they can cover their neighbours body on display if her husband sneaks up on the pilot? I also prefer above btw

Ah, of course, that must be the reason.

Thx!

;-)

 

[Chaosrider]

That keeps remote perfectly balanced in hand and very comfortable to hold for long times.
Having screen moved above sticks makes remote forward weighted.
Unless other end is ballasted with extra weight, or made longer.

I don't find the Mini-2 controller to be unbalanced in the slightest. The FPV controller is similar in overall size, and I literally forget it's there when I'm flying.

Balance is easy enough to engineer in. If the Mini-2 controller had a built in screen above the control sticks, rather than a phone in a holder above the control sticks, there's no obvious reason that would need to be any heavier or less well balanced.

Assuming equal weight and quality of balance, would you still prefer the screen under the sticks?

Thx.

 

[RogerDH]

Guess not everybody uses iPad mini

It holds tablets up to 10 inches.

 

[footfly]

As of today, May 1st 2022, it appears me, dealsdrone, and Jasper Ellens are like the only people that believe Mini 3 Pro will beat Air2S image quality (not limited to low light), not to mention it will trade blows with Mavic 3. With YouTube compression, dynamic range and resolution will be near indistinguishable from Mavic 3 even. Gimbal angle might be the main differentiating factor. (as well as portrait shots/video).

We all know the sensor is only 1/1.3, whereas the Air2S is 1 inch and the Mavic 3 even bigger.
We also know the mini 3 is the cheapest, with the Mavic 3 being about 3 times more expensive.

So, what are your last bets:

Will Mini 3 beat Air2S image quality?
Will it compete with Mavic 3?

Or will it be a slightly improved Mini 2 version that has worse image quality than even Air 2 (non S) ?

Place your bets as in about a week, the entire world will know.

Here is my 2-cent worth of thought, since Mini 3 officially is not available to public and therefore, much of this discussion is hypothetical:

Mini 3 Pro is likely not to have better image than Mavic 3 but might be about similar to Air2S (maybe less). Why? DJI has to still have good reasons to have people buy Air2S and the Mavic 3 in terms of money. If the Mini 3 Pro was going to be better image quality, than people would buy Mini 3 Pro over the Air2s and Mavic 3. In that case, with the current hypothetical price of Mini 3 Pro, DJI would lose money. It needs to be equivalent or likely less image quality, in order for DJI to continue to make money off the Air2S and Mavic 3.