Can anyone tell me according to the photos enclosed , why X3 camera showed a much higher sharpness than the X5???????
Same showed in the videos .
I already calibrate the X5 camera with AF Focus Assistant, but I still notice the X5 is not as sharp as the X3.
And another thing, when I tried to use the Manual Focus Assistant to calibrate the lens, it do not allow me to do so?????

Is I did something wrong, since the X5 is much more expensive than the X3,its a profession camera,  the result should not be like that?

Thanks for explaining.

Hello.
Did you set all of their camera settings the same?
Before trying to adjust the focus manually, have you toggled on the MF Assist in the DJI GO App (in Camera setting)?
Please visit our website for more comparison of X3/X5/X5R cameras: http://www.dji.com/products/compare-xcamera

Thanks.

Just some thoughts:

1) Aperture and diffraction: the x5 image shows an aperture of f16 - at this aperture usually quite some diffraction happens to occur leading to an "unsharp" image.
2) Exposure time: f16 compared to f2.8 - the exposure time must have been almost 6 times longer with the x5 (f16) than with the x3 (f2.8), so movement might have blurred the image even though the Osmo is stabilized.
3) jpg-settings: you can choose some jpg settings in the options, one of it should be sharpness (of the jpg algorithms). Are they on the same level on both cameras? I really doubt that the lenses of the x3 are supposed to be so much "better" than the lens of the x5, maybe the x3 images are just oversharpened?
4) Focus calibration: yes, the focus calibration might be another source of unsharpness. But usually the calibration optimizes the fine focus only (i.e talking about mm to cm range). When looking at your images the distance to the houses on the other side of the water is so far away and the wideangle lenses have such a large DOF, that this should not influence your sharpness too much.

Go try to take some pictures with same camera settings, e.g. using f5.6 and low sharpness setting? And maybe choose an object closer to the lens with a detailed structure to be able to better judge the sharpness of the images.

DJI-Adela Posted at 2016-5-25 14:24
Hello.
Did you set all of their camera settings the same?
Before trying to adjust the focus manually ...

Yes, the X3 and the X5 camera's setting are the same, sharpness 0, saturation 0, contrast0.
Here is what I have done before the adjustment of the MF.

1. Switch from AF to MF on the X5 lens.
2. Turn on the Goapp.
3. Steps as shown in photo 1 enclosed.
4.    -"-                    photo 2
5.    -"-                    photo 3
6.    -"-                    photo 4      

MF calibration failed.         

martin94b Posted at 2016-5-25 21:25
Just some thoughts:

1) Aperture and diffraction: the x5 image shows an aperture of f16 - at this ap ...

Thanks martin for your kind explanation .
Re your 1: I use f16 because I think f16 will give the deepest  depth of field , the picture will be sharp
                from the nearest point to infinity.
Re your 2: Very reasonable explanation.
Re your 3: the 2 cameras setting are exactly the same, sharpness 0, contrast 0, saturation 0
Re your 4: DJI GO App instruct to aim at the object 150ft away to calibrate,  thats what I have done.
                But I noticed that I never can get sharp pictures at infinity, even I use f16 aperture.
I cannot get successful calibration on MF, can it be the reason?????
I will try to take pictures using F5.6 and choose an object closer to test it.

Thanks you for all the trouble caused, appreciated.

wto6071@bigpond Posted at 2016-5-25 21:08
Yes, the X3 and the X5 camera's setting are the same, sharpness 0, saturation 0, contrast0.
Here i ...

Hi,
Looking at your pictures provided in your post, the picture taken by X5 is zoomed in, right?
Could you please send us the original files for comparison?

Thanks.

wto6071@bigpond Posted at 2016-5-25 15:23
Thanks martin for your kind explanation .
Re your 1: I use f16 because I think f16 will give the d ...

Hi,

yes, f16 will get you deepest DOF, but not necessarily the sharpest image! These are two pair of shoes ;-)
Beginning with ~f11 the diffraction of light at the aperture gets more and more significant, leading to less and less sharpness.

DJI-Adela Posted at 2016-5-26 13:40
Hi,
Looking at your pictures provided in your post, the picture taken by X5 is zoomed in, right?
C ...

Adela,
thanks for the help, both pictures ( each of them) are over 1000kB , can you advise how to send to you?

DJI-Adela Posted at 2016-5-26 13:40
Hi,
Looking at your pictures provided in your post, the picture taken by X5 is zoomed in, right?
C ...

Adela, these are the datas for the (2) pictures, hope it will help.
I would like to send you the origionals, if you could tell me how to send it to you, as you can see, each of them is about 3.9MB.

Regards

wto6071@bigpond Posted at 2016-5-26 12:25
Adela, these are the datas for the (2) pictures, hope it will help.
I would like to send you the o ...

Thanks for your reply.
Please upload the original files in Dropbox, tell us the sharing link and then we'll download it for analysis.
I'll let our engineers have a look.

DJI-Adela Posted at 2016-5-26 14:33
Thanks for your reply.
Please upload the original files in Dropbox, tell us the sharing link and t ...

Thanks Adela, I,ll do it after I returned home from work to-night.

DJI-Adela Posted at 2016-5-26 14:33
Thanks for your reply.
Please upload the original files in Dropbox, tell us the sharing link and t ...

https://www.dropbox.com/s/ssfpxflblbpln2x/X3%20.JPG?dl=0
Good morning, Adela, enclosed is the link address for the X3 and X5 pictures.

Have a nice day.

Regards

DJI-Adela Posted at 2016-5-26 14:33
Thanks for your reply.
Please upload the original files in Dropbox, tell us the sharing link and t ...

https://www.dropbox.com/s/u6z4svzhscv6qda/X5.JPG?dl=0

Here is the X5 picture from dropbox. Thanks for helping.

martin94b Posted at 2016-5-26 14:12
Hi,

yes, f16 will get you deepest DOF, but not necessarily the sharpest image! These are two pair ...

Good morning, Martin, thanks for your good advise, very useful indeed.
according to my (2) pictures taken with X3 and X5,
do you think there is some focus problems in the X5 camera?????

wto6071@bigpond Posted at 2016-5-27 00:49
Good morning, Martin, thanks for your good advise, very useful indeed.
according to my (2) pictur ...

Hi,

indeed the image from the x5 does look a little less sharp than the one from the x3. Also the FOV is quite different?!
But is it a focus calibration issue?
If I look at the foreground sharpness and the sharpness of the houses in the back, the DOF seems to be pretty large even with f5.6 (as I assume these images were taken?). So almost everything is within focus anyway. Thats why I doubt that focus calibration is contributing to your problem, while the focus calibration only shifts the focus point a little to the front / back (same as Front-Focus / Back-Focus issue on DSLRs), correct? So it moves the DOF a little to the front / back, but while the DOF in your image already covers the entire picture anyway a better focus calibration would not improve your sharpness (to my understanding).

Can you give me a favour to improve the error analysis?
Why not take 2 images (x3 and x5, again both with same settings and f5.6 or even better f2.8) of a large metering rule going from the foreground to the background (something you use in sports for measuring the distance of a discus or spear)? Then try to focus each camera on the same distance, e.g. put a large object next to the metering rule, preferable in a closer distance rather than large distance.
While the metering rule has a continuous scale it will tell use more precise about the DOF than your current images with lots of water between fore- and background (nice place to live by the way!!!), not really offering marks to judge sharpness.
And the metering scale will also lead us to the "sweet focus spot" with maximum sharpness, enabling us to compare both sweetspots of the x3 and x5.

If I am correct the sweetspot of the x5 will still be less sharp than the one of the x3, and your x5 camera has at least one more issue than focus calibration!
If the sweetspot is not next to the object you focused on you do have a focus calibration issue.

Only if the sweetspot of the x5 is as sharp as the one of the x3 but the sweetspot is not where you focused, then you only have a focus calibration issue alone and the rest of the camera / lens works just fine. But I doubt this is going to happen...

Looking forward to see those images from you ;-)

martin94b Posted at 2016-5-27 14:40
Hi,

indeed the image from the x5 does look a little less sharp than the one from the x3. Also the ...

Thanks Martin for the detailed professional advise. I have to go to work now, I will come back to you
after I do as you advised, and let you know.
Really appreciated.
PS: I had sent (2) Pictures to Adela, as she requested for DjI engineers to examine, I,ll let you know as well.

wto6071@bigpond Posted at 2016-5-27 05:15
https://www.dropbox.com/s/u6z4svzhscv6qda/X5.JPG?dl=0

Here is the X5 picture from dropbox. Thanks ...

Hi wto6071,
Based on the provided information, we can't make a very fair comparison since the two pictures were not really taken in the same settings. Since you set the camera as Auto, the Aperture, their focus lengths and the Exposure may be all different.

Thanks.

martin94b Posted at 2016-5-27 14:40
Hi,

indeed the image from the x5 does look a little less sharp than the one from the x3. Also the ...

Martin,
X3 camera use on Osmo does not have manual focus, it does auto focus by itself.
X5 camera has one touch focus, then how can I focus each camera on the same distance?
And
Excuse me for asking, what is the metering scale look like? Can you email me a photo of it????
Is it the retractable ruler we used to use at working site with marking on it?
As you mentioned, a large metering rule going from foreground to background, what is the length I need to take the photo??
Can you contact me by: wto6071@bigpond.net.au
Thanks

wto6071@bigpond Posted at 2016-5-27 15:22
Martin,
X3 camera use on Osmo does not have manual focus, it does auto focus by itself.
X5 camera  ...

Hi,

correct, the x3 does have autofocus only, but that ain´t a problem?!
And since we are talking about focus calibration of the x5 you should use the autofocus there as well.
Thats the reason why to have a "large" object next to scale: to use it for the autofocus (tap on that object on the display of your smartphone to focus) and then compare the position of the sweet spot of the metering scale. Preferrable the large object is flat like a sign facing perpendicular to your lens so you really only focus on a plane object.

Here´s some more information about focus calibration, and the fancy tool that they use is basically what we want to rebuild on a much larger scale for calibrating / checking the wideangle lens focus:

https://photographylife.com/how-to-calibrate-lenses

Just use the largest metering scale you can get ;-) E.g. 50m should be perfect, but even 10m can do just fine.
Try to find a scale where the lines and numbers are large as well, otherwise you won´t be able to see them on the picture.

Looking forward to see your results.

I had a good look at the large images from the dropbox links in Photoshop, and for me it just confirms what a vastly superior camera the X5 is over the X3.

Look at the blue sky.. nice and smooth in the X5, but the X3 shows all that noise. Look at the trees, and also the distant mountains.  The X3 image shows fuzzy pixelation, while the X5 is much better.

Finally, (and probably obvious to many) like with DSLR photography, the way to sharpen properly is in post, not in-camera. That means a soft image is what you want to start with.  The X5 is softer but that doesn't mean the detail cannot be brought out with sharpening in post. Remember the old rule - you can't soften a sharp image without losing detail and quality, but you absolutely can sharpen a soft  image and bring out detail you thought was never there. I learned this fact way back in 2003 when I lurked the dpreview forums with fellow DSLR owners.

wto6071@bigpond Posted at 2016-5-27 23:22
Martin,
X3 camera use on Osmo does not have manual focus, it does auto focus by itself.
X5 camera  ...

After receiving very good advises from Martin, Frypan and Adela. I must admitted that DJI Osmo forum
really helped much  , at least know where the problems is. (Though its not solved yet).

After I raised the question comparing the sharpness for my X3 and X5 camera, its not after I read the replies from the above (3) good guys that made me realised that "Is it I am over demanding, over suspicious"
on the quality of X5 camera.
May be I was wrong, because it give me a thinking that, IF  , there is something wrong with the X5, how come from the forum, there is no other user complaining about the focus problem, and I am the only one to do it.
May be the problem is from me, not the camera itself!

For these reasons, I am practicing and trying to calibrate the camera at different distance.(DJI GOapp advise for the Auto focus, at least 150ft distance). Now I tried to calibrate it at 50ft, 100ft, 150ft, facing very big signs
like big road signs, supermarket signs . Afterwards, examined these photos to find out the truth.
It takes time, of course, but its worth it.

Finally, thanks for the really great help and effort for those who spent time to explain to me. Really appreciated.

frypan Posted at 2016-5-30 19:47
I had a good look at the large images from the dropbox links in Photoshop, and for me it just confir ...

Frypan,

you´re absolutely right, the x5 image is less noisy (esp. in the sky) than the x3. And of course the larger sensor size contributes to this effect. BUT: the image of the x5 is also much brighter than the x3 image, so the sky is brighter as well, also leading to a less noisy image (the SNR - not the absolute noise level - is the important value and the SNR is always much better at the MSB than at the LSB range of the image, right?) And while the total noise level in the x3 image isn´t too disturbing at all I would not call the x5 a vastly superior camera (not because of these images, there might well be other comparing images that do proof this statement though).

But back to the topic of sharpness (even though you can never divide sharpness from noise level...):
Again you are right that post-processing can visually increase the impression of sharpness dramatically. But beside the fact that increasing sharpness wil also increase the noise level at the same time you will never be able to "create" details (=information) that was not within the original image.
So it makes total sence to try and optimize your equipment and way of working to collect as many details as possible upfront (while taking the image). And I am not talking about the "sharpness" parameter in the JPG settings ;-)
And from what we see in the images from wto6071 the images of the x3 are sharper than the x5 ones. But again, unfortunatly they were made with different settings so IQ is not really directly comparable...

martin94b Posted at 2016-5-31 12:36
Frypan,

you´re absolutely right, the x5 image is less noisy (esp. in the sky) than the x3. And o ...

True we need a more scientific test with accurate images.

Using what we have in this thread though, it's not hard though to find the X5 quality advantage. I admit the improvement would appeal more to those who really need it for pro work, or just fussy about image quality! My "vastly superior" is another person's "slightly better". And yes I know the X5 image is not as wide, so it's not a fair comparison. Look at the coner of building anyway,  and  the trees. I think we know where all that blocky mess on Youtube videos comes from when shooting trees with the X3. It sends the youtube compressor nuts, and the detail breaks down. Not so with X5, and this counts for a lot.

wto6071@bigpond Posted at 2016-5-31 07:12
After receiving very good advises from Martin, Frypan and Adela. I must admitted that DJI Osmo for ...

Thanks for your comment about DJI Osmo Forum.
Thanks to all of our kind people for sharing their experiences and ideas. There're lots of professional and knowledgeable people here.  

I'm happy that all of you would find helpful information here and enjoy our discussion.

For any issue, please contact us freely.

frypan Posted at 2016-5-31 07:22
True we need a more scientific test with accurate images.

Using what we have in this thread thoug ...

Not really a good comparison, though, as you:

(a) Have had to magnify the x3 image to match that of the x5

(b) Then magnified them both waaaay beyond their natural size.

The X5 in the dropbox examples does show less sharpness; but sharpness by itself, of course, is no reflection on the amount of detail held by the images.

Another way to compare the two images is to *reduce* the x5 until the objects are the same size. Then a completely unscientific comparison would show that the the x3 image is sharper, but that the x5 holds more detail.



The images provided can't really be used to prove anything.

LenMiddx Posted at 2016-5-31 17:16
Not really a good comparison, though, as you:

(a) Have had to magnify the x3 image to match that  ...

^ I actually reduced the X5 image down slightly to match the X3, not the other way round. I then applied a bit of sharpening to the reduced X5 image. Even with manipulating the X5 image this way, it still wins easily.

We don't need both images to match exactly anyway. Simply zoom both images to 200% in Photoshop and look around the image with your own eyes at the quality. The X5 image retains details and edge smoothness at 200% no matter what part of the image you look at - near objects or far. The X3 image on the other hand, looks rough at 200%.. edges are not clean, noise and artifacts. Zooming in Photoshop does not change the pixels, it simply shows you the image up close.

frypan Posted at 2016-5-31 13:37
^ I actually reduced the X5 image down slightly to match the X3, not the other way round. I then a ...

The problem is that we're comparing apples with asteroids.

The x3 has a tiny 1.23" chip compared with the X5's MFT-sized -- that's a difference of 8x, which gives the X5 a lot more pixels with which to derive its final image(s). i.e. Higher resolution.

So, there's really no winner/loser here -- the cameras are just too different. The images of the X3 will always be inferior to those coming off a camera with a larger sensor -- especially if you look close enough at a still image.

To get back to the OP's original comments -- yes, the x3 is applying a high degree of 'sharpening' -- edge contrast -- seen as the slight halo where the sky meets the buildings. Some people would call that 'over-sharpening, and that's down to the X3's software.

My guess is that the X5 software is simply applying less sharpening in the image shown.

LenMiddx Posted at 2016-5-31 23:29
The problem is that we're comparing apples with asteroids.

The x3 has a tiny 1.23" chip compared  ...

Careful. You're assuming the whole sensor is used when making video. That's an assumption you should confirm is actually true before drawing conclusions about where the better video quality comes from!

I'm not Googling before posting either, but I'll put my money on the camera not using all 16 million pixels and somehow downsampling on the fly to 4K all in real time. I'm pretty sure that's not how they work. And if that's true, then we're back to both cameras using same amount of pixels to capture video - the video specs then equal in that regard and the better quality of X5 coming from somewhere else.

Not only that, but the "camera software" you refer to, surely you mean the Osmo software? That's where camera sharpness is set.