Mini 3 Pro beats Air2S in both photo and video

[z28lt1]

Some more out-of-the-camera auto mode comparison footage.

Between this and the Pilot Institute, at least in the daylight, beginning to firm my opinion that the video quality is really close. If I freeze and blow up some areas, I can find spots where each might be slightly better, but I think overall the Air 2S video seems to have more spots where I think it "wins", so would lean the Air 2S for these auto/daylight shots. Close enough that if you don't want to spend time correcting/grading the output, you aren't going to worry about any minor differences that you need to blow up the video to see. There was one shot in the Pilot Institute video (close to the end with the blow up of the white tanker) where you saw just a tiny bit of noise/shimmer from the Mini 3 that the Air 2S didn't have. Think the point and shoot crowd will be very happy with either.

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

 

[pilotinstitute]

Here's the way I looked at it. If you're going to use the content straight out of the drone pretty much as is, it would seem that the Mini 3 Pro is configured for exactly that, just like a lot of point and shoots having that little extra sharpness and contrast out of the camera with the image processing tuned specifically for that.

If I had not gotten my Air 2S a few days ago, and saw the Mini 3 Pro release, I still would have gotten the Air 2S. Why? Because I can't do 10-bit color in a DLG/HLG Color Profile on the Mini 3, and for grading work, 10-bit color matters.

But if you're primarily going to use the drone for photography, the Mini 3 does seem to be a much more attractive option especially if you're comfortable utilizing the DNG raws but more so if you just want crisp Jpegs out of the camera.

That's the other thing about the specs though, at 1/1.3" sensor the same aperture (f/2.8) is going to have image degradation due to diffraction compared to the slightly larger 1" sensor. For example on the smaller 1/2.3" sensor of the mini, 2.8 is already on the edge of breaking down the image quality a little. So that f/1.7 aperture helps against diffraction.

The other thing is the Mini 3 is a 24mm equivalent focal length versus a 35mm equivalent. The wider FOV with smaller sensor generally means a larger perceived depth of field (a larger front-to-back perceived sharpness of the focused area), versus a larger sensor that is also a longer lens, that depth of field is shallower and the perceived sharpness can seem softer if it's not locked onto a clear subject, even if the longer lens is stopped down 1.6 stops from f/1.7 lens.

But for less than a grand for a kit that includes a remote that doesn't require an expensive smart phone or tablet? That's kind of hard to beat especially if you are mainly just wanting it pre-processed right out of the drone without the intention of pushing it further in post.

I think that's a great conclusion and analysis!

 

[boldblue737]

Here's the way I looked at it. If you're going to use the content straight out of the drone pretty much as is, it would seem that the Mini 3 Pro is configured for exactly that, just like a lot of point and shoots having that little extra sharpness and contrast out of the camera with the image processing tuned specifically for that.

If I had not gotten my Air 2S a few days ago, and saw the Mini 3 Pro release, I still would have gotten the Air 2S. Why? Because I can't do 10-bit color in a DLG/HLG Color Profile on the Mini 3, and for grading work, 10-bit color matters.

But if you're primarily going to use the drone for photography, the Mini 3 does seem to be a much more attractive option especially if you're comfortable utilizing the DNG raws but more so if you just want crisp Jpegs out of the camera.

That's the other thing about the specs though, at 1/1.3" sensor the same aperture (f/2.8) is going to have image degradation due to diffraction compared to the slightly larger 1" sensor. For example on the smaller 1/2.3" sensor of the mini, 2.8 is already on the edge of breaking down the image quality a little. So that f/1.7 aperture helps against diffraction.

The other thing is the Mini 3 is a 24mm equivalent focal length versus a 35mm equivalent. The wider FOV with smaller sensor generally means a larger perceived depth of field (a larger front-to-back perceived sharpness of the focused area), versus a larger sensor that is also a longer lens, that depth of field is shallower and the perceived sharpness can seem softer if it's not locked onto a clear subject, even if the longer lens is stopped down 1.6 stops from f/1.7 lens.

But for less than a grand for a kit that includes a remote that doesn't require an expensive smart phone or tablet? That's kind of hard to beat especially if you are mainly just wanting it pre-processed right out of the drone without the intention of pushing it further in post.

This! So much this! Even my Mavic 2 Pro was a disappointment in the sense the normal color profile stunk and when you filmed in LOG profiles the drone didn’t have the CPU power to perspective correct the distortion of the lens. Basically you had to live with bad color (but a proper perspective image) or always do LOG and have to both color grade AND perspective correct the image. Now it wasn’t exactly hard, but it was time consuming enough it made me not want to edit as much. Ironically later when I bought a Mini 2 while the image was clearly “inferior” it had nice color and proper perspective and so it was a much nicer pleasure to quickly use the results. I feel for most the results straight out of camera is the most important thing but obviously bigger sensors and color depth etc. have their advantages for those needing it and understanding it.

 

[TRogan42]

Ok Pilot Institute's daylight comparison video is out between the Mini 2 / Mini 3 Pro / Air 2S.

The results show the Mini 2 is much worse than the others but I find the image between the Mini 3 Pro and Air 2S essentially close to identical with only slight variations that give one or the other the "win" based on preference and specific lighting. The moment they zoom into 4x on the image though you see the Air 2S suddenly is a clear winner with far less noise and warbling of the image thanks to its larger sensor and 10 bit color.

It just goes to show what a powerhouse the cheaper and smaller Mini 3 Pro is that the image is the equal of the Air 2S straight out of camera, but if you are a color grading master it might still benefit you to keep your Air 2S so you have the added headroom of changing the image it provides.

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

.




at the end with 4x zoomed footage air2S looks like soap, much worst than Mini2, there issomething wrong with their A2S.



.

 

[karlblessing]

.




at the end with 4x zoomed footage air2S looks like soap, much worst than Mini2, there issomething wrong with their A2S.



.

The mini 2 to me looks better than both the Air 2S and Mini 3 Pro zoomed in at the end (in their examples), the Mini 2 is a little less saturated/contrast (But can see a tiny but more detail as a result), the Air 2S bumps some saturation but seems softer in their copy. The Mini 3 Pro while it seems crisper, I'm seeing very noticeable noise artifacts.

Course with that much clutter of details and pattern, repeated 3 times, youtube could have been a bit aggressive on their compression algorithm (which is usually why I prefer Vimeo or to be able to download clip samples directly).

 

[Mid-AtlanticDroneServices]

I have an Air2S. While it looks like the Mini 3 pro is better at still photos and slightly better at video in daylight, I wonder about lower light conditions. I often fly at dusk or right after dusk and I have found the 1” CMOS makes a massive difference. I haven’t compared them directly, but I wonder. All of those photos and videos are in bright daylight in a desert-like environment.

Now, if we’re comparing the drones overall, there are several reasons one might choose a 2S. In my case, I’m not a photographer but more of a recreational pilot. So, bleeding edge photo quality isn’t that big a concern for me. I do wish the still photos with my drone we’re better. They’re still really good. Of course, if you’re interested in various accessories like a drop kit, a megaphone, LED board, etc., you will want the more powerful lift of the Air2s. Moreover, you may want to use the new RC pro, which isn’t compatible with the Mini. I have a smart controller but I’m going to be upgrading to an RC Pro pretty soon. Basically, the air is more versatile. It depends on your use. If you want portability, lower cost and the best video/photo in normal light (aside from the Mavic 3), the Mini 3 seems to be best choice.

 

[ForeverFree]

Would love to see iPhone-level out-of-the-box JPEG quality …

 

[karlblessing]

Would love to see iPhone-level out-of-the-box JPEG quality …

Depends on which iPhone, is it actually that good or just processed real punchy out of the box?

 

[KS-6]

That's the other thing about the specs though, at 1/1.3" sensor the same aperture (f/2.8) is going to have image degradation due to diffraction compared to the slightly larger 1" sensor. For example on the smaller 1/2.3" sensor of the mini, 2.8 is already on the edge of breaking down the image quality a little. So that f/1.7 aperture helps against diffraction.

But then at f/1.7 you might have more corner softness or de-centering issues to deal with as is seen in regards to the latter with the samples on dpreview.

I'll know pretty soon after my first flight with the Mini 3 if it is going to work well enough for vertical stills compared to cropping or stitching my MP2.

 

[karlblessing]

But then at f/1.7 you might have more corner softness or de-centering issues to deal with as is seen in regards to the latter with the samples on dpreview.

I'll know pretty soon after my first flight with the Mini 3 if it is going to work well enough for vertical stills compared to cropping or stitching my MP2.

You are correct regarding corner distortion, which is more an issue the wider you try to get the field of view. Tiny sensors can pair decently with telephoto lens, but not so much with wide angle.

I am curious why they didn't choose something like a 24mm or 28mm equivalent on the Air 2S (35mm equiv), but on the mini, mini 2, mini 3, and Mavic 3 they went with 24mm equivalent lens.

 

[KS-6]

You are correct regarding corner distortion, which is more an issue the wider you try to get the field of view. Tiny sensors can pair decently with telephoto lens, but not so much with wide angle.

I am curious why they didn't choose something like a 24mm or 28mm equivalent on the Air 2S (35mm equiv), but on the mini, mini 2, mini 3, and Mavic 3 they went with 24mm equivalent lens.

I wish there were like a $6 to $8K price point, 3-4 pound foldable drone, horizontal / vertical gimbal, 4/3rd sensor at 24MP with clean ISO 800 and 24mm / 50mm lens options.

I'm starting to doubt the Mini will give me the IQ I need and I will just return it and stick to my MP2. If I see more DNG's with super soft outer image area I will cancel it outright.

 

[karlblessing]

I wish there were like a $6 to $8K price point, 3-4 pound foldable drone, horizontal / vertical gimbal, 4/3rd sensor at 24MP with clean ISO 800 and 24mm / 50mm lens options.

I'm starting to doubt the Mini will give me the IQ I need and I will just return it and stick to my MP2. If I see more DNG's with super soft outer image area I will cancel it outright.

If you're used to shooting professionally with mirror less cameras such as Olympus/Panasonic/Sony/etc most things less than a 1" is going to be hard to fit the bill, you usually just have to take a compromise to some extent in exchange for having a camera in a place your bigger camera can't reach (outside of mounting it on a drone rig).

Like if I wanted to get close to what I use for my non flying set up, I would need an Inspire 2 with at a minimum an X5S camera.

 

[KS-6]

If you're used to shooting professionally with mirror less cameras such as Olympus/Panasonic/Sony/etc most things less than a 1" is going to be hard to fit the bill, you usually just have to take a compromise to some extent in exchange for having a camera in a place your bigger camera can't reach (outside of mounting it on a drone rig).

Like if I wanted to get close to what I use for my non flying set up, I would need an Inspire 2 with at a minimum an X5S camera.

Oh, I am more than used to it. With a 35 year career under my belt I shoot 35mm to 8x10 in Nikon, Leica, Hasselblad, Sinar, Chamonix, etc. I still own and use the gyro that is my user name from all kinds of aircraft. I have no problem with the price jump of an Inspire 2 / Z7 but the weight and size are the main drawbacks.

I do manage well in stitching verticals with non-dynamic subject matter with the MP2, especially with first tier processing in DXO then on to ACR's image quadrupling "Enhance" feature. When things get moving like fast moving cloud cover or other moving things is where I wish for a vertical sensor orientation.

 

[karlblessing]

Oh, I am more than used to it. With a 35 year career under my belt I shoot 35mm to 8x10 in Nikon, Leica, Hasselblad, Sinar, Chamonix, etc. I still own and use the gyro that is my user name from all kinds of aircraft. I have no problem with the price jump of an Inspire 2 / Z7 but the weight and size are the main drawbacks.

I do manage well in stitching verticals with non-dynamic subject matter with the MP2, especially with first tier processing in DXO then on to ACR's image quadrupling "Enhance" feature. When things get moving like fast moving cloud cover or other moving things is where I wish for a vertical sensor orientation.

I understand where you're coming from, I don't have as many years but I've shot almost everything from a minox B to a linhof 5x7, and have done several 6 feet tall by 12 to 18 foot long printed panoramic stitching from a 12 and 16 megapixel camera that came out flawless and printed no less than 150dpi.

So I can be picky, but I also can't afford the price jump of what I want hence knowing how to get around some of it.

 

[mark_wilkins]

Other than availability, I really see no reason to buy an Air 2S over a M3P right now.

The M3P's camera (judging from those chart images) is pretty impressive, and it does seem to outperform the Air 2s camera in some meaningful ways. Not really a surprise, considering it's a couple years newer.

However, in addition to some of the comments about video above, if regulation is not an issue for an application, the greater mass of the Air 2s will give it some additional stability in moderate wind. The effect of this may have at least as much impact on sharpness in real-world shooting conditions when that's an issue (and it's a concern for me, living in Iceland.)

Regardless, the differences between these cameras at least in still-image performance in daylight will be very subtle in real-world use, and packing similar performance into a smaller, lighter, cheaper package makes a very compelling case for the M3P over the Air 2s for most users.

(I'm speaking as a happy Air 2s owner who's unlikely to buy another drone until the European classification markings start appearing on them.)

As far as the test images from Pilot Institute, I guess it's possible that sample variation is an issue. Some mass-produced lenses from other manufacturers have suffered from some sample variation in sharpness, and it's likely that a pre-production M3P provided as a test article may have been more rigorously checked before delivery for performance than Pilot Institute's off-the-shelf Air 2s. However, DJI has the resources to make big steps forward in their optical designs between product generations, and that's a more likely explanation for differences.

These products were not designed to be clearly differentiated from each other, and there will likely be an Air 3 soon that makes the picture much clearer.

 

[karlblessing]

I wonder if anyone with a Mini 3 could check something.

I know both the Air 2S and M3P are stated to record 150Mbps for their video, but that apparently is only the H.264 rate. When shooting at H265 (the only option when shooting in DLog/HLG 10-bit) the maximum/overall set bit rate is 125Mbps in the file. A 75Mbps H.265 file is roughly the same visual quality as 150Mbps H.264 (Long-GOP).

I'm curious if the Mini 3 is set to do 125Mbps as well when selecting H.265 or if they drop the rate down for H.265 with the Mini 3. (can check using Media Player Classic, file -> properties -> MediaInfo tab, and look for Overall Bitrate).

 

[Paradigm]

The M3P's camera (judging from those chart images) is pretty impressive, and it does seem to outperform the Air 2s camera in some meaningful ways. Not really a surprise, considering it's a couple years newer.

However, in addition to some of the comments about video above, if regulation is not an issue for an application, the greater mass of the Air 2s will give it some additional stability in moderate wind. The effect of this may have at least as much impact on sharpness in real-world shooting conditions when that's an issue (and it's a concern for me, living in Iceland.)

Regardless, the differences between these cameras at least in still-image performance in daylight will be very subtle in real-world use, and packing similar performance into a smaller, lighter, cheaper package makes a very compelling case for the M3P over the Air 2s for most users.

(I'm speaking as a happy Air 2s owner who's unlikely to buy another drone until the European classification markings start appearing on them.)

As far as the test images from Pilot Institute, I guess it's possible that sample variation is an issue. Some mass-produced lenses from other manufacturers have suffered from some sample variation in sharpness, and it's likely that a pre-production M3P provided as a test article may have been more rigorously checked before delivery for performance than Pilot Institute's off-the-shelf Air 2s. However, DJI has the resources to make big steps forward in their optical designs between product generations, and that's a more likely explanation for differences.

These products were not designed to be clearly differentiated from each other, and there will likely be an Air 3 soon that makes the picture much clearer.

I've found the wind stability, based on the one video of the British guy fighting the winds by the sea, to be quite impressive. In my experience, I think most people just simply don't fly when such conditions are present. So, if there is any wind resistance advantage between the 2S and the M3P, I think the likely negligible difference is moot.

The untested, arguably, negligible advantage the A2S has isn't worth picking over the M3P. Heck, even for the "first upgrade" that the Air 2S is positioned to be in DJI's lineup, the difference in price, performance, and future operating cost greatly skews towards the M3P. Plus, you'll be getting a better remote. The N1 is barebones at best, in my experience.

Also, I think the M3P is a safer drone to fly than the A2S. And that's something that people don't often consider when choosing between what drone to purchase. When one of these falls from 399', I'd much rather have it be the M3P.

I 100% agree with your last sentence. This is how DJI operates -they release a drone at lower tier that bests the one above it, and then they release one at the tier above to recapture the performance difference.

The low level drones see quicker iterations released, as they serve almost as a testing ground for newer tech.

Side bar (Mod Removed Language) on a personal note, I am and have been dismayed that the US can't or won't compete in the high end camera drone industry.

 

[Paradigm]

Well, here you go. Some good reasons the Air 2S is better choice. D-Log 10 Bit video, 5K (which they failed to show) and higher zoom options.

Bottom line, the Mini 3 Pro is a great little drone, but the Air 2S has better Pro quality video.

NOTE, They got the specs wrong, the Air 2S has Ocusync 3 not Ocusync 2.

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

I did not see this reply. I will watch it and circle back as soon as I have a chance, likely later on today. Thank you for linking to it!

 

[ScottTX]

I have an Air2S. While it looks like the Mini 3 pro is better at still photos and slightly better at video in daylight, I wonder about lower light conditions. I often fly at dusk or right after dusk and I have found the 1” CMOS makes a massive difference. I haven’t compared them directly, but I wonder. All of those photos and videos are in bright daylight in a desert-like environment.

Now, if we’re comparing the drones overall, there are several reasons one might choose a 2S. In my case, I’m not a photographer but more of a recreational pilot. So, bleeding edge photo quality isn’t that big a concern for me. I do wish the still photos with my drone we’re better. They’re still really good. Of course, if you’re interested in various accessories like a drop kit, a megaphone, LED board, etc., you will want the more powerful lift of the Air2s. Moreover, you may want to use the new RC pro, which isn’t compatible with the Mini. I have a smart controller but I’m going to be upgrading to an RC Pro pretty soon. Basically, the air is more versatile. It depends on your use. If you want portability, lower cost and the best video/photo in normal light (aside from the Mavic 3), the Mini 3 seems to be best choice.

I unexpectedly gravitated towards night flying and have an Air 2 - As I've already invested in an iPod Mini 6 and I believe that can reuse the standard Air 2 controller, the Mini 3 Pro is very enticing.

Here is a late dusk shot (12 MP) that I took with my Air 2:



I like this shot but the noise is pretty bad.

 

[karlblessing]

Here is a late dusk shot (12 MP) that I took with my Air 2:

View attachment 148383

I like this shot but the noise is pretty bad.

The noise is supposed to be much better on the Air 2S over the Air as it does temporal denoiseing in the sensor. Compared to my previous mini and mini 2 it seems to work, lil loss of detail but not as much chromatic noise even at 800+