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

Photo quality?

A reference picture by the M2P would have been nice.
It would have shown if the advantage is the calculated one stop ...😆

There's definitely noise reduction in the JPG automatically, so DNG is the way to go for direct comparison.
But the quality looks fine.


JPG is always compressed lossy.
Do not have the M2P anymore. I do know what it was capable of though. I added a link to the DNG for that one and another. After the first one was edited, I got an update for PS/ACR and more control was definitely there. My mention of no compression only meant I set it to the lowest before saving ;)
Here is a link to video from last night:
To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Cheers, Jon
 
The Mavic 2 Pro has issue with the glass & aperture in some case making it soft. I got the M3 last night & did a few shots in the dark & I am impressed! This image is resized 50% with no compression.
f/5, 1/15s, ISO 100. Underexposed to get better shutter speed. I was not going for long shutter. My M2P could not have been this good. The shadows came out much better without introducing more noise! I would also say it's better than my GH5! I also attached a crop.
Update Link to Sample folder DNG: M3 Samples - Google Drive

Cheers, Jon
These look really good I sometimes wonder if the mastery in how DJI/Hasselblad convert to JPEG actually produces a better job than what a RAW converter can easily do as some of the RAW drone images I have seen have been average but the JPEGS quite outstanding.
 
My old lightroom applies the correct corrections for the lens, so it doesn't need an update as it is written in the open spec.
It will pick up the basic corrections, yes, but Adobe added profiles for the M2P and A2S specifically in the current CC version. I will be interested to see how one for the M3 will improve on the standard DNG changes.
 
@heyengel
That's looking good (bit soft though).
Now think of, if the software would give you the full potential ... how sweet that would be.
Yet it again proves, light, composition and framing is key to pictures. 😄
 
Last edited:
I tested 20 different M2P because the quality control of the lens is so poor to find the 2 perfect models I have now!
These 2 are sharp edge to edge at all apertures. From my testing there is a 1 in 10 chance of getting a perfect lens.

In the 20 I tested I saw a lot of issues with mid field sharpness which was the worst issue as it was fairly central ruined images , sometimes it got worse at f4.

I do not want to go through the nightmare I had with the M2P so for now I will stick with my 2, I only use them for photography so there isn't enough for me to upgrade, plus the wider lens is less use for me.

Well, what we have seen in the past, that this is certainly is applicable in "traditional" SLR-/SLM photography with well-built lenses, but as with the M2P, changing the aperture higher than f 4 made the image overall softer and soft corners not sharper. It even was no real difference beyond f 2.8 most of the times. It may have changed with the bigger sensor, but I guess, we have to live with soft or shifted softness or test a bunch of them and decide which is best as a better lens formulare certainly would need more space. I returned my first M2P for ridiculous soft areas (thread might be here somewhere). My 2nd is ok, while the left corners a still a bit softer than the right.

Interesting to see more real world DNG and not from promotional or selected drones.
PS: Certainly the converters will gain in performance, but visible declining sharpness from centre to corners without any post-sharpening is something that should not be primarly corrected in software (if even possible to the same extent).;)
 
You are luck to get an acceptable lens in 2! Hasselblad only helped design I think, they certainly do not have anything to do with the manufacture.
I fully agree! My first MP2 had a very bad lens with a big blurry spot on the left sight. I could reorder my purchase. My second one has an acceptable lens and is good enough for its purpose.
But the name Hasselblad for this lens is a laugh!
Ton, 100% photographer.
 
Yes - that's a fine example of Moiré. Geometrically it's still there in the JPEG, but the color artifacts have been removed.
Based on these samples, I may never buy a Mavic 3, the lens looks terrible. It seems to be soft quite close to the centre. I know f4 & f5.6 may help, but I use f2.8 a lot to get low light shots so it would be useless.
 
The shots from the main camera in your samples look bad even at these small sizes!

Care to elaborate? The JPEGs were optimized for small file sizes.

It would also help if you can share what you consider are "good" photos from your collection of masterpieces.
 
@heyengel
That's looking good (bit soft though).
Now think of, if the software would give you the full potential ... how sweet that would be.
Yet it again proves, light, composition and framing is key to pictures. 😄

Thanks! I exported these JPGs at 70% quality. The originals from the 1X camera look really sharp at 100%. The 7X images do come out soft, especially the ones I messed up the focus on.

Agreed that people obsess too much on pixel peeping and forget about what matters.
 
  • Like
Reactions: globetrotterdrone
Care to elaborate? The JPEGs were optimized for small file sizes.

It would also help if you can share what you consider are "good" photos from your collection of masterpieces.
I mean I can see the flaws in the lens at These small sizes, its nothing to do with your jpegs.
 
Like what? Also post samples of good images for comparison.
I don't need to post samples of good images to compare, you can compare from the centre of the images to the edges themselves, it is not consistent. I can see flaws in the shots posted even in the small sizes, so at 100 percent it must look terrible on the 1x lens.

In your samples in this link, the top right image, you can see the buildings in the to right go blurry. It is one issue to see that at 100%, but another to see it in a small jpeg like this. Disappointing but I expected it to be like this based on my M2P experiences.

In the bottom image here, if you look at the railings you can see that the image goes blurry even near the centre, but for that you have to zoom in 100%.
 
Last edited:
Bbased on the corners of your M3 being better than the M2P it seems you got one of the many bad M2P out there with optical flaws. from what I have seen compared to my M2P the mavic 3 has great quality but worse corner detail. But I tested many to find a good one.

you mentioned chromatic aberration Is an issue, but did you mean moire? That is just because the lens is too sharp for the sensor.
I care about one thing above all else, photo image quality ... so immediately compared Mavic 3 to my old Mavic 2 Pro. Here's the whole nine yards: Mavic 3 Photo Quality vs. Mavic 2 Pro – Kevin Davis Photography

While it's not a slam-dunk, I have concluded that I will retire my Mavic 2.
 
Last edited:
I don't think you would notice a ground-breaking difference between the overall image quality of 1 inch v.s. 4/3 sensor. M2P's one inch sensor has a dynamic range of 12 stops, the 4/3 sensor in M3 has 12.8 stops. They both have the same resolution.

I think the biggest difference lies in the high ISO performance, which I am yet to see any real-world comparisons between the two.

I made a real comparison in difficult conditions and with different ISO. Look in my topic. There's a link to a video of this comparison. And in the description of the video there is a link to the source code.
 
I care about one thing above all else, photo image quality ... so immediately compared Mavic 3 to my old Mavic 2 Pro. Here's the whole nine yards: Mavic 3 Photo Quality vs. Mavic 2 Pro – Kevin Davis Photography

While it's not a slam-dunk, I have concluded that I will retire my Mavic 2.
Thanks, very useful. The chromatic aberration section - the Mavic 3 shot shows moiré on the railings rather than chromatic aberration. Other tests have shown the same. Maybe when Adobe create a profile, they will be able to improve this.

I was all set to get a Mavic 3, but as it's my first drone, I decided an Air 2S would be more sensible with it's more mature firmware. Also, the large camera housing seems more vulnerable in minor crashes on the M3.
 
Thanks, very useful. The chromatic aberration section - the Mavic 3 shot shows moiré on the railings rather than chromatic aberration. Other tests have shown the same. Maybe when Adobe create a profile, they will be able to improve this.

I was all set to get a Mavic 3, but as it's my first drone, I decided an Air 2S would be more sensible with it's more mature firmware. Also, the large camera housing seems more vulnerable in minor crashes on the M3.
Moiré is physically inevitable in any Bayer-based imaging system since it's the product of the spatial frequency of the object's image plane with the spatial frequency of the sensor. I've never fully understood how it is detected and removed in the RAW conversion process, but presumably something fairly clever is happening there.
 

DJI Drone Deals

New Threads

Forum statistics

Threads
134,485
Messages
1,595,523
Members
163,013
Latest member
GLobus55
Want to Remove this Ad? Simply login or create a free account