Beginner fuzzy blurry images D3500 70-300 DX VR

Messages
18
Name
Michal
Edit My Images
Yes
I could use some help please.
My photos show blur / fuzz to the point that a face can not be recognized.

Nikon D3500 with 70-300 DX VR.
Photos taken at 300mm. Distance to subject is 100+ meters.
Stationary subject.

I tested with several different settings.
I wouldn't be bothering you guys if I hadn't tested everything I am aware of.
VR on VR off - same results.
1/1000 1/4000 - same results.
Compression fine / raw - same results.
File size S / M / L - same results.
ISO 200 / 800 - same results.
f6.3 - f32 - same results.

I can't fault the compression algorithm because I get pin sharp images (fine JPG) from the 18-55 kit lens.
Could the VR of the 70-300 have failed?
I can't blame the optics because, I see everything perfectly sharp in the viewfinder.
Do I expect too much from this setup?
Obviously some cameras are capable of taking pin sharp photos at 100+ meters.

What do you think?


In the samples focus was on the ANIMAL truck back and roof ridge in the middle of the image.
 

Attachments

  • DSC_6542.JPG
    DSC_6542.JPG
    74.5 KB · Views: 81
  • DSC_6543.JPG
    DSC_6543.JPG
    69.2 KB · Views: 81
My first observation is that these photos were taken in bright sun yet are very dark. Underexposure is not going to help even though it will not cause softness.

How big do you have to enlarge the photos before you see the softness?
 
Histogram shows marked underexposure and the EXIF has been stripped so no idea what settings you have used, or indeed if you have cropped it/them.

Is this the same lens that you posted about last year that you thought had dust in it..................and did you, as you spoke of it, part dismantle it to 'clean it up'?

Re: the one with the "animal" name showing ~ the image can be corrected to lift the highlights and does sharpen enough for potentially acceptable use. Obviously, with the original full sized file more could be done.

The underexposure and the apparent vignette makes me wonder if you had a Polariser Filter on the lens and the fact that the vignette is uneven was a CPL rather than a Linear one?

Now, setting aside your 'settings' and the query over a filter on the lens, you mention that your short zoom is pin sharp but the 70-300mm is not based on the posted images. Using a longer lens even with VR requires IMO good cameracraft...............so does the sharpness improve if you take the same type of picture with the combo on a tripod?

Lastly, apart from pilot error and/or settings any picture using a long lens and expecting pin sharp pictures of a small element with the whole of the frame can be that the lens is less than stellar at the 300mm (long end of its zoom range. Plus, atmospheric disturbance can be a factor i.e. heat haze......especially if the above images are from the late afternoon and the road and all the surroundings are "hot".

PS if the the D3500 has MFA it is possible to correct any slight focus issues with a lens but you need to confirm that it is indeed the lens by trying a controlled test i.e. tripod mounted to prove there is an issue or not. Does the D3500 have live view, if so when you do some controlled tests use OVF focusing and also Live View focusing to see if there is any difference or not?
 
My first observation is that these photos were taken in bright sun yet are very dark. Underexposure is not going to help even though it will not cause softness.

How big do you have to enlarge the photos before you see the softness?
Hi
f32 was intended as this was the last thing I tested with the hope of achieving better results.
When viewing on PC 100% (or 1:1) - not pixel snooping, faces are not recognizable, can barely make out where the 'face-blob' has eyes!
Same when viewing on the camera itself.
 
f32, really!!!

At that small an aperture you are well into diffraction effects on image sharpness. :banghead: Try the same 'test' at between f7.1 to f11 and see if that helps the focus sharpness you seek?
 
Histogram shows marked underexposure and the EXIF has been stripped so no idea what settings you have used, or indeed if you have cropped it/them.

Is this the same lens that you posted about last year that you thought had dust in it..................and did you, as you spoke of it, part dismantle it to 'clean it up'?

Re: the one with the "animal" name showing ~ the image can be corrected to lift the highlights and does sharpen enough for potentially acceptable use. Obviously, with the original full sized file more could be done.

The underexposure and the apparent vignette makes me wonder if you had a Polariser Filter on the lens and the fact that the vignette is uneven was a CPL rather than a Linear one?

Now, setting aside your 'settings' and the query over a filter on the lens, you mention that your short zoom is pin sharp but the 70-300mm is not based on the posted images. Using a longer lens even with VR requires IMO good cameracraft...............so does the sharpness improve if you take the same type of picture with the combo on a tripod?

Lastly, apart from pilot error and/or settings any picture using a long lens and expecting pin sharp pictures of a small element with the whole of the frame can be that the lens is less than stellar at the 300mm (long end of its zoom range. Plus, atmospheric disturbance can be a factor i.e. heat haze......especially if the above images are from the late afternoon and the road and all the surroundings are "hot".

PS if the the D3500 has MFA it is possible to correct any slight focus issues with a lens but you need to confirm that it is indeed the lens by trying a controlled test i.e. tripod mounted to prove there is an issue or not. Does the D3500 have live view, if so when you do some controlled tests use OVF focusing and also Live View focusing to see if there is any difference or not?
HI
I did not take the lens apart nor sent it to Nikon. It has dust in it.
There is a filter attached! I did not think about it! I slapped it on a year ago to protect the lens from scratching and just left it on. - so thank you for pointing that - testing tomorrow without it.
Correcting is not an option. The images are used as evidence. They must be straight of the camera. No editing whatsoever.
LiveView - will test that. But only for testing purposes. My subjects are moving. The LiveView delay is not acceptable as I would miss most shots.

There is a sample photo, i found on google, of the same lens apparently:

https://sorinvacaru.files.wordpress.com/2013/08/dsc_4178_021_nx2.jpg

I can only dream of that sort of results...

Above all, how it is that I can see everything perfectly sharp in the viewfinder?!
Viewfinder is using the same optics, right?
 
The viewfinder image is very, very small. That makes a large difference to apparent sharpness.

Given that the viewfinder is sharp, is the live-view screen also sharp? This is the image as seen by the sensor before capture. If live-view is sharp and the captured image is not then the problem is mechanical rather than optical - aperture (f/32!), shutter, operator.
 
Without controlled testing I have no idea what might be wrong, if it is?

Re: OVF vs captured image ~ for fear of stating the obvious, the image as shot is what you are seeing in the OVF and as @john.margetts says the OVF is very small, so how do you know the face in the OVF was sharp......but if you are enlarging (zooming into the image on the PC) you are no longer seeing the 'in life view'.

So, just how big a percentage of full sized image is the face you mention.....this factor alone might be the key to your "problem" e.g. your expectations are too high..... perhaps a bit like your level of cropping in ?
 
Last edited:
f32, really!!!

At that small an aperture you are well into diffraction effects on image sharpness. :banghead: Try the same 'test' at between f7.1 to f11 and see if that helps the focus sharpness you seek?
This is the correct and definitive answer. If you don't know what diffraction limitation is, then look it up, but as a rough guide it will start to occur if you use a lens aperture smaller than f/11 on your camera, which has an APS size sensor. Just a very basic understanding of physics would have prevented that problem.

Why do some manufacturers make a lens that can be stopped down far more than it should be? Don't know, just as I don't know why some cars can travel at twice the legal speed limit or more, same thing.

And then we get underexposure, which really doesn't help.
And then there's the filter, which often adds an additional layer of image degradation. There are some people who believe in always having a filter fitted to the lens - that's their choice but it is always very likely to reduce image quality, as well as to introduce lens flare. My own view is that the only time that a filter should be fitted is when it is actually doing a useful job.

And, as the exif date isn't showing we don't know about the ISO setting you used, if it's too high then again this will impact on image quality.

And then there's the possibility of camera shake.
Take a correctly exposed shot at f/ll with the camera anchored on a good quality, solid tripod, without a filter. If you're still having problems, that will be the time to suspect the lens (or at least the focussing).

Hope this helps.
 
@mee how

Michal

You mention the images are needed/going to be used as evidence and therefore no editing is to be done.......please bear in mind that even an enlargement might be seen as manipulation ~ you would need to get clarification on that aspect.

Right, crossing over with your other thread re: 18-200mm vs 70-300mm
If you are photographing distant subjects the longer the focal length the better and 300mm is not that long!
You need to fill the frame as much as possible with the subjects. NB the more of the frame the subject fills the less 'zooming in' on the computer will be needed by the authorities to see the face(s) you talk about.
Re: OVF view ~ the camera shows the view at the widest aperture and only closes to the chosen aperture on pressing the shutter button
try Tv (Shutter priority mode) using at least 1/500th of a second and if the D3500 has Auto ISO use that but with them max ISO of say 3200 NB the camera will pick the aperture and most likely wide open i.e. the smallest number e.g f3.5

HTH and practice, practice, practice :)
 
Last edited:
The images are under exposed by a lot and I wonder if you are adamant that they should be un-manipulated you have maybe turned a few too many things off in your camera settings. For example I have had a look and the softness can be corrected by sharpening them and they come out as semi-acceptable working from such a small JPEG. With such a long focal length lens you will have a lot of foreground and background that is out of focus even at f/32, so just try to focus on the exact subject and try not to use f/32 to stop diffraction and to keep your shutter speed higher. If they are to be used for this mysterious 'evidence' I'm sure normal corrections would be allowed, I mean if it was film the choice of film and the lab that processed it and the paper it is printed on could be called 'manipulation'. You can't view a digital image without it being manipulated by software so let the software take the strain. If the 'other party' object you can give them the original file especially if you use RAW.
 
Last edited:
Thank you all.
The volume of information is overwhelming! I just wonder I should have stick with compact 50x camera.
Anyway.
Tested with f11 - no difference to the softness. I previously tested with a range of f-stops, the f32 image was just the last one I took, yet there was no noticeable difference to softness.
I left the camera on f11 from now on.
Tested with removed filter and hood - no difference to the softness.
Rested the camera on a box: some images were sharper than before but by far. Still quite soft. This confirmed, to me, that the VR actually works.
Tested with LiveView and voila! Excellent result. What does it mean? Does not matter much because LiveView does not allow to capture quick moving action.
Finally, the issue is resolved by opening the window. Fairly sharp image.
Yes, I was taking the photos through a dirty window. You can laugh. Next time you will know where to start with amateurs like myself.

To satisfy your curiosity: The photos I take, are used in court as a support to street CCTV's and body cameras. While the video evidence shows an offence in progress, it often isn't detailed enough to identify the offenders.
I am volunteering as a 'neighborhood watch'.

Thank you all again.
 
OP posted solution while I was typing!!
 
Last edited:
I wrote the bit below before you posted your latest reply:
The biggest single problem for us is that we have very little information about the example shots that you took, mainly because you stripped out the EXIF data, no doubt accidentally.
So, in the absence of information, all that we can do is to make assumptions based on backworking from the known situation.

So, we know that the "correct" exposure here, ISO 100, f/ll, would be about 1/200th second, and we know that you took the shot at f/32, so that would need ISO 1600. But we can see that it's about 2 stops underexposed, which brings the ISO back to around 400.

I've looked at the shot with the L200 (?) greatly magnified and it's clear to me that there's no camera shake. Most competent photographers would need a shutter speed of around 1/500th to avoid camera shake at this magnification, there's absolutely none, so I'm guessing that the shutter speed was around 1/2000th. That raises the ISO needed to 4000. That's just an educated guess but will be in the right ball park.

Looking at the sky in that shot, the noise level is horrendous which, even allowing for the increase in noise that occurs with underexposure, confirms that a pretty high ISO was used.
Now, we all know that modern digital cameras are pretty good at fairly high ISO settings, but ISO 4000 will inevitably affect image quality, and high ISO settings should only be used when absolutely necessary, and it wasn't necessary here.

If you'd shot at f/8 then the image quality would have been much higher optically, with no diffraction limitation, and the ISO setting would have been reduced to just 500 at correct exposure.
+++++++++++++++++++++++++++++++++++++++++++++++++++++++
And then you said this:
I just wonder I should have stick with compact 50x camera.
No, you just need to learn a bit more - which you're doing. And, like almost everyone else, you're learning by making mistakes:)
Tested with f11 - no difference to the softness. I previously tested with a range of f-stops, the f32 image was just the last one I took, yet there was no noticeable difference to softness.
I left the camera on f11 from now on.
It will have made a massive difference to the sharpness, my guess is that the problems caused by the high ISO (which you still haven't confirmed) and shooting through a window made such a massive impact that you can't seen the negative effect of shooting at f/32
Tested with removed filter and hood - no difference to the softness.

[/QUOTE]
Unlikely, there's always a difference when using a filter, but with a perfectly clean, high quality filter the difference may not be significant. Having a lens hood on though is a positive step.
Finally, the issue is resolved by opening the window. Fairly sharp image.
Yes, I was taking the photos through a dirty window. You can laugh. Next time you will know where to start with amateurs like myself.
Well, now that you've told us this, the whole situation makes more sense. The dirt won't help but the real problem is that windows are optically very inferior, the window was acting as a terrible quality filter
To satisfy your curiosity: The photos I take, are used in court as a support to street CCTV's and body cameras. While the video evidence shows an offence in progress, it often isn't detailed enough to identify the offenders.
I am volunteering as a 'neighborhood watch'.

Thank you all again.
Well, I don't have any experience of that - don't fancy being punched in the face - but I had a high quality PTZ security camera fitted outside my studio, covering the whole area 24/7. The quality was perfect and the police often popped in to view what it had seen, it cost me a small fortune in coffee and biscuits:)
 
I use the same camera and the same 70/300 lens. Recently spent many days tracking a Rabbit (Human) around all the roads in our town taking very many images. This was a Covid inspired task to bring a smile to many.
Due to the fact that my skills were bad with my new camera and first dSLR images were taken on the hoof (Paw).
I soon found that everything 'Automatic' was my only course and apart from cropping and a light adjust they were virtually all very acceptable.
As a very long time novice dare I ask why you shoot manual in your situation?
If I have totally missed the point... Please forgive me!:mad:
 
I use the same camera and the same 70/300 lens. Recently spent many days tracking a Rabbit (Human) around all the roads in our town taking very many images. This was a Covid inspired task to bring a smile to many.
Due to the fact that my skills were bad with my new camera and first dSLR images were taken on the hoof (Paw).
I soon found that everything 'Automatic' was my only course and apart from cropping and a light adjust they were virtually all very acceptable.
As a very long time novice dare I ask why you shoot manual in your situation?
If I have totally missed the point... Please forgive me!:mad:
I'm guessing that shooting everything in manual may be related to the OP's belief that images taken for evidential purposes must not be manipulated.
I don't know what "manipulated" actually means in this context, and my advice to him would be to seek advice from the people who are likely to use them - the police.
I'm guessing that they will say that the photos must be truthful, i.e. not manipulated to the extent that they give a false impression of reality. Removing or adding people to the scene would obviously be dishonest pixellating the faces or vehicle index plates of the innocent would not, neither would carrying out a reasonable degree of sharpening.

But, I stick with my earlier suggestion; forget about still photos, use a high quality security camera instead.
 
But, I stick with my earlier suggestion; forget about still photos, use a high quality security camera instead.
I would have to lure all crooks to the security camera to pose. They are all over the estate.
Another issue being, they are well aware of cameras locations and cover their faces etc.
When I get a shot while they are in the same vehicle, same clothes within minutes of an offence, they are done.

Council nor police would invest in a proper CCTV system and also they are not permitted to conceal it...
Same applies to private CCTV. I explored this previously. It is fine to capture public areas as long as it is clearly stated and obvious for all passers by.
 
I use the same camera and the same 70/300 lens. Recently spent many days tracking a Rabbit (Human) around all the roads in our town taking very many images. This was a Covid inspired task to bring a smile to many.
Due to the fact that my skills were bad with my new camera and first dSLR images were taken on the hoof (Paw).
I soon found that everything 'Automatic' was my only course and apart from cropping and a light adjust they were virtually all very acceptable.
As a very long time novice dare I ask why you shoot manual in your situation?
If I have totally missed the point... Please forgive me!:mad:
Hi
As my subjects are often moving fast, the only automatic setting was the 'sports'. Unfortunately, the camera would not correctly set the ISO or shutter speed in anything other than perfect lighting. Rain, clouds, tree shade, etc renders the auto mode not so useful. After taking over 15,000 shots, I can now just look outside and by 'feel' know what to set the camera to :) i usually get it close to right. It's satisfying.
I am no expert by far but there is only 3 factors: shutter speed, ISO, f-thing (in the middle).
Shutter speed can be set very quickly between shots using the dial.
f- can be set very quickly between shots using the dial and the +/- button pressed at the same time, however, I just set it to the lowest number and let the lens decide. Unless, I photo flowers, then I set it to ~f11-f22.
ISO is a pain to change but can be done if you have time between shots.

I encourage you to try.

https://photos.app.goo.gl/kf1Xy24ZBUwKg8Mx9
 
Last edited:
Agree with EJB here, just stick it on AUTO till you learn a bit more about your camera, the camera communicates with the lens so knows what focal length you are at and can change aperture and ISO accordingly, all you have to do is concentrate on your subject and having a steady hand.
 
Hi
As my subjects are often moving fast, the only automatic setting was the 'sports'. Unfortunately, the camera would not correctly set the ISO or shutter speed in anything other than perfect lighting. Rain, clouds, tree shade, etc renders the auto mode not so useful. After taking over 15,000 shots, I can now just look outside and by 'feel' know what to set the camera to :) i usually get it close to right. It's satisfying.
I am no expert by far but there is only 3 factors: shutter speed, ISO, f-thing (in the middle).
Shutter speed can be set very quickly between shots using the dial.
f- can be set very quickly between shots using the dial and the +/- button pressed at the same time, however, I just set it to the lowest number and let the lens decide. Unless, I photo flowers, then I set it to ~f11-f22.
ISO is a pain to change but can be done if you have time between shots.

I encourage you to try.

https://photos.app.goo.gl/kf1Xy24ZBUwKg8Mx9

To get a correctly exposed picture the three settings of shutter speed, aperture & ISO need to be set "correctly". Your posted pictures show marked underexposure.

For fast moving subjects I advised in an earlier post that you try Tv mode (shutter priority) this mode allows you to set the shutter speed and the aperture will adjust as appropriate....if auto ISO is available (do read the manual to see if the D3500 has that function) that can be a useful aid to getting it right.

All of the above is predicated on having a longer enough focal length lens to fill as much of the frame as possible with the subject...... combined with the appropriate choice a AF focusing point.

PS do look up the 'Exposure Triangle' for the insight ;)
 
...After taking over 15,000 shots, I can now just look outside and by 'feel' know what to set the camera to :) i usually get it close to right. It's satisfying...

Not wanting to be rude, but the two sample shots you posted with significant underexposure show that actually, while you may be 'close', you are not actually getting it right.
I would suggest you actually shoot in Shutter priority (called Tv on Canon), with Auto ISO (limited to however high you are happy for your camera).
While the camera will not get it right every time - in a typical 'street' scene it will be very close the vast majority of the time, and will have the benefit that as you only have to set the Shutter Speed (perhaps even just leave it at a value you are confident avoids motion blur in the subjects you are shooting), you will be able to take the pictures more quickly - which I suspect is important in capturing the offenders while they are in view and have not noticed you.
 
Not wanting to be rude, but the two sample shots you posted with significant underexposure show that actually, while you may be 'close', you are not actually getting it right.
I would suggest you actually shoot in Shutter priority (called Tv on Canon), with Auto ISO (limited to however high you are happy for your camera).
While the camera will not get it right every time - in a typical 'street' scene it will be very close the vast majority of the time, and will have the benefit that as you only have to set the Shutter Speed (perhaps even just leave it at a value you are confident avoids motion blur in the subjects you are shooting), you will be able to take the pictures more quickly - which I suspect is important in capturing the offenders while they are in view and have not noticed you.

I said the same re Tv mode earlier in post #10 and again above........so far (may have missed his post) he has yet to acknowledge that such settings have been tried?
 
@mee how

Exposure triangle
.

D3500 online manual

PS I note that Nikon call Tv mode S on the control settings dial.....and it says you can use Auto ISO with S
 
Last edited:
Michal, I think you are asking too much of both the camera and even more so the lens....If it can't get fair results on Auto I don't believe any manual control will improve things much.
Poor light is the systems big problem as in any camera and more particularly the lens of that ilk.
I think the setup is extremely capable in reasonable conditions and wonderful value for money...I love it!

Presumable a much more capable second hand lens is out of the question?
 
Back
Top