Lens sharpness/calibration

[Peter Connan]

I need help.

I struggle to get sharp pictures. Just some background for starters: My favorite subject is the animal kingdom, and since I can't escape from the city often enough, birds seem to be about 90% of what I shoot. Equipment consists of a Nikon D7000, Nikkor 80-400VR lens and a very old Mamiya tripod fitted with a new Giotto video head.

Recently I have found that I sometimes manage to take some reasonably sharp pictures. After peering at pictures for hours, I have come to the conclusion that the common denominator is focal length: I have no sharp pictures at 400mm, but when I zoom back matters improve greatly.

This would firstly point to camera technique, but since a large majority of my bird photos are taken in Manual mode at 1/1600 SS, at f8 and using auto ISO with a ceiling of 3200, I am not convinced that camera shake is the major problem.

I then took a series of pictures of an object at various zooms (the object was held by my daughter and moved closer or further to fully frame in the viwfinder) and the 400mm photo was much softer than any of the others. However the light was pretty bad and SS was fairly low, such that "object shake" might have had an influence.

So then I did a second string, this time with the object on the ground, the camera on the tripod (with all sections fully retracted for best stability), mirror-up mode and with a cable release. I waited 5 seconds between lifting the mirror and taking the picture.

Again, 400mm was the softest (although the difference was less this time). But, I took two other pictures manually focussed in Live View with maximum magnification, one at 340mm and one at 400mm, and the 400mm one is the sharpest photo in the whole string.

My conclusion is that the problem is not so much with the lens's optical quality, but with the autofocus. I am hoping it is a clibration problem.

My question is, can this be fixed by calibration, how do I do it etc? Also, are there any flaws in my reasoning?

[DaveK]

it sounds to me as if this could be a calibration issue. Suggest repeat your second test: stationery object, good SS, stable camera, MLU and cable, and photograph either a calibration chart (you can download lots of these), or just a calibrated ruler using the centre focus point. Lens wide open for minimum depth of field. Let the camera autofocus take care of focus, and examine the results to decide whether there is front or back focusing.

Northlight images has an alternative method where you can use a computer monitor as the subject. works quite well too.
Once established where the system is focused, I am not sure whether D7000 allows compensation or not: Nikonians will know.

D.

D

[Peter Connan]

Thanks Dave

I know my camera has an adjustment in the custom settings menu, but have no idea how to use them. I have read the article at Northlight, and will try it but am a bit concerned about the distance I need between the screen and camera (50xFocal Length, or 20m) but will have a stab at it.

One question though, how will this affect the shorter focal lenghts (as the adjustment is don for the whole focal length range)?

[Terry J]

I too have an Nikkor 80-400VR and battled with inconsistent results specially at 400mm I found that it was whenever I used the VR as I was not allowing the VR sufficient time to kick in. When I mount the camera on a tripod I use a Manfrotto 055 and turn off the VR I get excellent results. One of the easiest way to do tests is to take a page from a glossy magazine and tape it to a wall, have your camera on a sturdy tripod focus up on the page with the front of your lens centred and parallel with the page and a shutter speed higher than the focal length of the lens. Use the self timer to trigger the shutter. That way the subject does not move and the camera will not have any shake except for possible mirror slap which should not have any effect if the shutter speed is higher than the focal length.

[jab2]

Peter, any zoom lens has a sweet spot, both in aperture, and in focal length. That is just the way of physics and engineering. Design tweaks are done to make it reasonable across the range, but with most telephoto zooms when at the stops at the longest end it displays a little bit of softness. I see this on my Canon EF 100-400mm too. Pulled just a few millimeters back from the 400 end and one notice a visible improvement in sharpness. There is a general rule (OK, there are exceptions too) that says that any zoom which is bigger than 3x in focal length range will have some issues somewhere in the range, as the deviations between wide and tele is just too much to optically correct in an affordable way. (The figure is found by dividing the wide end FL into the tele end FL)

I have no experience with Nikons, but think this might be what you notice. Terry's notes on the VR (Canon = IS) is also valid. There is some operational constraints in the technology, and one must be aware of this so that you can adjust your technique to work around it if possible.

[Peter Connan]

Thanks guys. I normally switch VR off when working with shutter speeds over 1/100th, so it's not that.

I do realize that this lens is a consumer-grade lens and may well struggle, but still I had expected more. Further, the LiveView manual focus result seems to prove that it can do better? I will try the calibration over the weekend, and see if matters can be improved.

Is there anybody who would like to swop me for a 200-400?

[DaveK]

Peter lots of useful comments above. Do that calibration test. In controlled conditions, you will immediately know, particularly if you focus on a calibrated 'gradient' of some sort, whether your lens , at your chosen focal length, can achieve sharp focus at all, and if it can, whether this is at the point selected by your autofocus function in camera. This will tell you: lens problem or calibration problem and give you a good departure point.
Regarding calibration focal length for a zoom: I have seen various recommendations. Most sense to me is to calibrate at the most used focal length, which for you will most likely be 400mm for avian work. Some recommend choosing a midrange setting, to get some balance between possible calibration differences between the two ends of the zoom range. For sure, zooms may require different calibration at different focal lengths, and some high end bodies now allow this type of calibration!!
GOod luck

D

[Peter Connan]

Thanks Dave.

Will do it at 400mm, as you are correct, I generally use it at between 200-400, but probably at least 80% at 400mm.

[ThalliumZA]

Just to throw my 2cents in on an issue i had with my 120-400mm.. the first thing I did when i got that lens was slap a UV filter on for protection, I was about to pull my hair out and send it back, as it would just never focus properly on the higher zoom range. and backgrounds always looked a bit distorted. During the weekend i put on a circular polarizer to test with and my lens actually got "drunk". pictures literally looked like those old school 3D pictures where you need those red and green glasses to see them. The Higher the zoom, the worse it got.

I took the UV filter and the Polarizer off, and Voila.. Everything went sharp and clear. out of focus backgrounds no longer looked weird either.

I had the same issue, Zooming back was fine, but at the higher end it was always out of focus. It was caused by a cheap UV filter. if you have a UV filter on, and its cheap. I would suggest to try it without the filter.

[Peter Connan]

Thanks, I do have a UV filter on. Will take it off and see.

[ThalliumZA]

Have a look at the difference i get from that UV filter. I quickly made this.

http://imageshack.us/photo/my-images/18/75883845.jpg/

[Peter Connan]

Web filer does not allow me to see this? Can I PM you my e-mail?

Thanks for the effort.

[ThalliumZA]

Web filer does not allow me to see this? Can I PM you my e-mail?

Thanks for the effort.

Yeah sure.. No Problem

EDIT: I mean Yeah sure about the email.. and No Problem for the effort

[Peter Connan]

I have made an attempt at calibrating this lens (although it seemed a lot more difficult doing it than the instructions made it sound). In any event, I am much happier with the results, there is a significant improvement.

Thanks for all the assistance!