Poor image quality

[erhard]

As others have stated, no lens is going to be perfect across all of the corners and even expensive ILC lenses have to be stopped down to get the best corner performance. Even with your Mini 2 photo (flower image) I do see distortion and loss of sharpness in the top left and bottom right corners. With that being said, I see a greater amount of distortion and loss of sharpness in the top and bottom left corners of the Mini 4 photo (beach view). The middle and right side of the Mini 4 photo look acceptable so my first thought is that your Mini 4 lens has a slight amount of decentering which favors the right side. However, without seeing additional examples where the entire frame is on the same plane of focus with good lighting, it is difficult to conclude what is happening.

thanks for the comment. i checked many images and as you say the top left and bottom left are the worst. yes, fixed f1.7 does not help with lens distortion. but would you not agree that even the center sharpness seems to be better in the mini 2 than mini 4 pro? peoples comment that i should not use leaves etc to judge sharpness but that is exactly where it shows. if tree leaves or grass looks like a fudge it is a bad sign for me. at 100% view i expect a crisp sharp image. i took some screen images of special test patterns and it shows the same loss of sharpness in the respective corners. sure it would be great to be able to stop down the the aperture but it is what it is

 

[erhard]

2 things that occur to me about that bottom left corner especially:

1. That section of bush is a bit taller than the rest of it, yet the focus is dialled in for the main centre of the image, which is about 10 ft further down, so that bit might be blurry in that case ?

2. Being the highest bit of bush is it possible it was moving in the wind, which also would have produced the blur ?

...notwithstanding everything everyone said about images going a bit ropey in the corners of consumer level gear, which is my XP too.

not sure which picture you are referring to but in general the exposure time was 1/8000s. so motion blur would not be the issue and what i see is not motion blur but actual distortions

 

[Droning on and on...]

not sure which picture you are referring to but in general the exposure time was 1/8000s. so motion blur would not be the issue and what i see is not motion blur but actual distortions

Was that a typo and you meant 1/800?

with ISO 100 and the brightest possible sunny day, I just can't make that exposure work with a quick calculation, but I'm probably screwing something up.

 

[erhard]

The exit further up in the thread says 1/8000s

 

[erhard]

thanks everyone for sticking around with me. i finally had a chance to take a mini2 and my mini4p for a direct comparison. the drones where set side by side and not flying and pointed to the same object. the drones were not flying just sitting there and taking the picture.

the link is here

DJI compare



in this folder are 4 images total.
2 from each drone

the shorter file name is mini2 (...0074, ...0075)

i hope you agree that the mini4 pro pics are no good, like out of focus, compared to the mini2.
i paid clear attention to when i took the pic with the mini4p, pushed the button halfway to get focus lock and then pushed full to get the picture


i am not even mentioning the distortion or softness in the upper left corner



thanks
erhard

 

[Meta4]

i hope you agree that the mini4 pro pics are no good, like out of focus, compared to the mini2.

You are fooling yourself

 

[erhard]

You are fooling yourself

Ok and how so?

 

[Felix le Chat]

thanks everyone for sticking around with me. i finally had a chance to take a mini2 and my mini4p for a direct comparison. the drones where set side by side and not flying and pointed to the same object. the drones were not flying just sitting there and taking the picture.

the link is here

DJI compare



in this folder are 4 images total.
2 from each drone

the shorter file name is mini2 (...0074, ...0075)

i hope you agree that the mini4 pro pics are no good, like out of focus, compared to the mini2.
i paid clear attention to when i took the pic with the mini4p, pushed the button halfway to get focus lock and then pushed full to get the picture


i am not even mentioning the distortion or softness in the upper left corner



thanks
erhard

Morning Erhard, I noticed that the shots in your drop-box are still JPG's. If you want to see a completely objective view of whatever aberrations are physically present in the M4P camera, you need to set both the M2 and the M4P to capture JPG+DNG and then view and assess the DNG (latent image - the digital equivalent to exposed but undeveloped film. What the sensor actually 'sees', not what it spits out).

This will show you all the ugly stuff like lens distortion, possible sensor/lens misalignment, dead pixels etc. This will be definitive and not open to argument.

If your DNG's show localized mushiness/lack of focus, then the fault is mechanical and you need to go back to the vendor for either an exchange or a refund.

Take the DNG comparison shots along as proof - they really can't be argued with, but in-camera processed JPG's can.

 

[erhard]

Morning Erhard, I noticed that the shots in your drop-box are still JPG's. If you want to see a completely objective view of whatever aberrations are physically present in the M4P camera, you need to set both the M2 and the M4P to capture JPG+DNG and then view and assess the DNG (latent image - the digital equivalent to exposed but undeveloped film. What the sensor actually 'sees', not what it spits out).

This will show you all the ugly stuff like lens distortion, possible sensor/lens misalignment, dead pixels etc. This will be definitive and not open to argument.

If your DNG's show localized mushiness/lack of focus, then the fault is mechanical and you need to go back to the vendor for either an exchange or a refund.

Take the DNG comparison shots along as proof - they really can't be argued with, but in-camera processed JPG's can.

You r right that the raw file has all information. I do have the dang from the mini 4 but not from the mini 2 . Said this my experience is that most jpg conversations today are excellent and o am sure DJI uses the same jpg machine in all there drones. For example I use raw and jpg on my Olympus micro 4/3 and the jpegs are super close to the raw files.
For example the edge unsharpness has nothing to do with jpg process it is simple the optics.
Also since we only compare jpeg from either drone should that be fair too.
I will try to get raw from mini 2 but do not expect a difference in the comparison
Thanks for your advice

 

[erhard]

Morning Erhard, I noticed that the shots in your drop-box are still JPG's. If you want to see a completely objective view of whatever aberrations are physically present in the M4P camera, you need to set both the M2 and the M4P to capture JPG+DNG and then view and assess the DNG (latent image - the digital equivalent to exposed but undeveloped film. What the sensor actually 'sees', not what it spits out).

This will show you all the ugly stuff like lens distortion, possible sensor/lens misalignment, dead pixels etc. This will be definitive and not open to argument.

If your DNG's show localized mushiness/lack of focus, then the fault is mechanical and you need to go back to the vendor for either an exchange or a refund.

Take the DNG comparison shots along as proof - they really can't be argued with, but in-camera processed JPG's can.

ok i went back to do the jpeg side by side and ther is NOTHING really sharp in the mp4 image - no need for DNG file. i checked all areas of the image even the ground where we go from close to a distance and nothing is really sharp.
take for example a look at the lamp. the right mini2 image is super sharp and the left is just fuzzy and we cannot blame jpg for that one.
or look on the left side under the roof. on the right side (mini2) you see clearly the separation of the beams (horizontal lines) whereas on the left it all looks almost like one surface. this is so obvious that something is not right with that camera.
the mini4 focus was on the bush so at least that one should be SHARP but it is not. yes, the mini4p f1.4 will have less DOF than the mini2 f2.8 but something in the picture should be in focus after all.

 

[Droning on and on...]

@erhard, while there's nothing wrong with the advice to check DNGs, as I said earlier what you're expecting is entirely reasonable.

I get better jpgs from my M4P than my Mini 2 images (it's since been sold), and interestingly better than my Mini3P. All shooting in full AUTO.

So my advice would be to send it back as defective and get a replacement. You should be getting stunning, sharp jpgs from the M4P with default settings.

 

[erhard]

@erhard, while there's nothing wrong with the advice to check DNGs, as I said earlier what you're expecting is entirely reasonable.

I get better jpgs from my M4P than my Mini 2 images (it's since been sold), and interestingly better than my Mini3P. All shooting in full AUTO.

So my advice would be to send it back as defective and get a replacement. You should be getting stunning, sharp jpgs from the M4P with default settings.

thanks, that is exactly my thinking too.
with all the rave about the mp4 image quality i should not be bothering all of you and ask if there is a difference - it should obvious that m4p is much better than mini2. i have not gotten much reply from dji support on this subject. one more thing but that is not a real issue. it seems the mini2 come out with a warmer tone right of the drone which i personally like better.
maybe i should wait until they add aperture control on the mini5
btw.: if you don't mind could you please attach one of your m4p images (jpg)

 

[erhard]

You are fooling yourself

i do NOT

 

[erhard]

thanks, that is exactly my thinking too.
with all the rave about the mp4 image quality i should not be bothering all of you and ask if there is a difference - it should obvious that m4p is much better than mini2. i have not gotten much reply from dji support on this subject. one more thing but that is not a real issue. it seems the mini2 come out with a warmer tone right of the drone which i personally like better.
maybe i should wait until they add aperture control on the mini5
btw.: if you don't mind could you please attach one of your m4p images (jpg)

i found on flickr great m4p pics. some are DNG and photoshop processed but they are really great. i see no smeared corners etc. so i think yes the camera if all works can produce great images

DJI Mini 4 Pro

Still images and videos taken exclusively using the cameras of the DJI Mini 4 Pro.

www.flickr.com

 

[Occams Razor]

i found on flickr great m4p pics. some are DNG and photoshop processed but they are really great. i see no smeared corners etc. so i think yes the camera if all works can produce great images

DJI Mini 4 Pro

Still images and videos taken exclusively using the cameras of the DJI Mini 4 Pro.

www.flickr.com

First, I do see distortion and lack of sharpness in some of the corners for some of the Flickr photos. Also without knowing the post processing workflow on the Flickr photos, it's difficult to compare images.

In your recent tests, the aspect ratios are different so the resolutions of the photos between the Mini 2 and Mini 4 are not the same so that is not a good comparison. You are comparing a 9MP image to a 12MP image. Also, the perspectives are different in that the MP4 was further back than the Mini 2 which puts the MP4 at a disadvantage. The different aspect ratios probably won’t have an impact but the perspective difference can make a major difference.

The Mini 2 has a fixed focus that basically tries to optimize the DOF based on the aperture/sensor size of that camera. The Mini 4 Pro allows you to choose the plane of focus so it's possible to have parts of the frame out of focus. When you use autofocus of the M4P, unless you know specifically what it is focusing on, it's possible that it could be focusing on an unintended subject. You should redo your test but place the focus point directly on the subject that you want in focus.

I am not saying that there isn't an issue with your MP4 but your test comparisons are severely flawed.

 

[erhard]

First, I do see distortion and lack of sharpness in some of the corners for some of the Flickr photos. Also without knowing the post processing workflow on the Flickr photos, it's difficult to compare images.

In your recent tests, the aspect ratios are different so the resolutions of the photos between the Mini 2 and Mini 4 are not the same so that is not a good comparison. You are comparing a 9MP image to a 12MP image. Also, the perspectives are different in that the MP4 was further back than the Mini 2 which puts the MP4 at a disadvantage. The different aspect ratios probably won’t have an impact but the perspective difference can make a major difference.

The Mini 2 has a fixed focus that basically tries to optimize the DOF based on the aperture/sensor size of that camera. The Mini 4 Pro allows you to choose the plane of focus so it's possible to have parts of the frame out of focus. When you use autofocus of the M4P, unless you know specifically what it is focusing on, it's possible that it could be focusing on an unintended subject. You should redo your test but place the focus point directly on the subject that you want in focus.

I am not saying that there isn't an issue with your MP4 but your test comparisons are severely flawed.

all good points.
the aspect ration was the direct comparison with my friend's mini2 because i lost mine. right i noticed only later that he was on 16:9 and mine at 4:3. so yes that makes the corner sharpness comparison invalid but i look also at the overall sharpness. also i compared before to my mini2 4:3 pics and do not see the corner problem.
the resolution is the same because the wide screen pics are simply a crop of the 4:3.
the reduced DOF should be taken care of by the autofocus system and i concluded NOTHING in the entire mp4 pics is really crisp.

from my many hours of looking at various pics i am convinced the m4p are no good.
i called DJI and thanks to my care package they allowed me to send it in for repair or whatever they do. i hope i get a mini4p back in 2 weeks that takes better images. i am sure will let you know.
thanks for your critical input.

 

[pingaddict]

The quality issues are linked to jpeg processing by the drone. I own a Mini 4 Pro and it delivers lame jpegs, but crazy good RAWs. DNG is the only way to go if you want sharp and rich images.

 

[Meta4]

DNG is the only way to go if you want sharp and rich images.

I see comments like this all the time, but I disagree.
Contrary to popular opinion you can also get great images shooting jpg.

 

[erhard]

I see comments like this all the time, but I disagree.
Contrary to popular opinion you can also get great images shooting jpg.

If jpegs are low quality that would be a no go for a good product. Why would DJI have a poorer jpeg machine in mik4p than mini2. Does not make sense to me. I am surprised about all the different controversial comments. Again , I will update u when I get the drone back.

 

[Meta4]

If jpegs are low quality that would be a no go for a good product. Why would DJI have a poorer jpeg machine in mik4p than mini2. Does not make sense to me. I am surprised about all the different controversial comments. Again , I will update u when I get the drone back.

It's hard to tell what you are trying to say.