Is there a problem with my 100-400mm mkii?

Messages
32
Name
Jason
Edit My Images
Yes
IMG_5286.JPG IMG_5278.JPG Hi everyone,

I recently bought the 100-400mm mkii and have been using it on my 70d. A lot of photos come out soft. Some photos are okay, but a lot of the time a shot that you would think the camera and lens would find easy to get sharp is either out of focus or has a slight amount of what looks like motion blur, even though shutter speeds are about 1/500 or even 1/1000 with the lenses IS mode 1 engaged. I have autofocus micro adjusted the lens to be best of my ability (W-2 T-1 seems like all it needed) but nothing seems to work. Live view focusing is no sharper than viewfinder.

I have tried every focus point and operation, shutter speeds all the way down and all the way up, high and low ISOs, lens IS modes, photos do get sharper with a narrower aperture, but only a bit and this lens should be sharp at f5.6.

Am I missing something? Or is there a problem? I have included a couple of photos, one cropped image of where I was focusing and one of the back of the camera so you can see he settings. This image was taken at ISO 800 which I know is quite noisy on a 70d but that doesn't explain the blurry look.

Would appreciate any advice!
 
Firstly, if you think that you have a focus problem you need to get the camera on a tripod for any testing. Low ISO and centre point AF, single shot AF. Choose a day when you have good light so shutter speeds are fairly high. Try with IS off and on to see if there is any difference.

It doesn't need to be scientific. When I buy a new lens I get a seed packet and pin it to my fence then set the camera/tripod up at a reasonable distance away (if I was using the 100-400 on a crop sensor I would set up about 10-12 metres away). Take a shot, then microadjust +5 and repeat, then microadjust -5 and repeat. They I check the shots on the LCD. The reason I use a seed packet is that the writing on the back is different sizes and at the bottom is usually about 2mm high but anything similar will do. If I can read the smallest writing there isn't a lot wrong with the set up. From this quick test you can get an idea if the lens is there or there about.

If I see that the print is sharper at +5 or -5 then I start to do further testing and I have a simple homemade set up for that.
 
Did you take this test shot on a tripod? Looks a little like motion blur to me, even if very slight.

It does sound to me like it could be a poor performing lens tho, might be worth getting in touch with who you bought it from (if it was a retailer).
 
I've tried all that. Tripod set up, shutter release cable, tried every shooting mode (quite mode, mirror lock up, etc), the target was a black metal meter ruler with white markings, on a tilt so I could see the front and back focus. It was about 12 meters away. I did micro focus adjust it to where I think the sharpest point was but when taking real world photos it's just all over he place.

The photos I posted were hand held.
 
Try on a tripod, IS turned off, use a remote release or 10sec timer delay.

Have you got a filter on the lens... if so also try without the filter.
 
target was a black metal meter ruler with white markings, on a tilt so I could see the front and back focus
Not the best way to do things, the focus sensor is larger than the viewfinder indicator and may not focus where you think it has, test on a good contrast target that is parallel to your sensor.
 
Not the best way to do things, the focus sensor is larger than the viewfinder indicator and may not focus where you think it has, test on a good contrast target that is parallel to your sensor.

That's a good point. The ruler is about 2 inches wide and has good contrast (black and white) but I will perform another test on a bigger flat surface.
 
You haven't got a filter on the lens by any chance?

Edit....someone has asked
 
Last edited:
If you know exactly which focus point the camera is using and you still have soft pictures once you check a high res picture, have you consider the glass having haze/fungus? Ideally you can test the camera with a different lens of same specs or your lens on a different camera.
 
If you know exactly which focus point the camera is using and you still have soft pictures once you check a high res picture, have you consider the glass having haze/fungus? Ideally you can test the camera with a different lens of same specs or your lens on a different camera.

The lens is brand new so fungus wouldn't be an issue. I used to have a Tamron 70-300mm vc usd and that worked fine. Also my canon 100mm f2.8 is super sharp.
 
Dont understand Canon but if it was Nikon i would ask did you shoot jpeg, on Nikon the in camera jpeg settings would need tweaking to get sharp image.
 
Dont understand Canon but if it was Nikon i would ask did you shoot jpeg, on Nikon the in camera jpeg settings would need tweaking to get sharp image.

Looking at the LCD shot, it is large JPEG but as I shoot raw I cannot recall the 'order' that the picture style info is indicating i.e. whether no he has left default or the 3 is an increase to sharpening setting.

Whichever......when I shot jpeg I left all such settings at 0 because I always pp'ed on computer and all needed to a lesser or greater degree some sharpening!

Edit ~ I wonder if the 70D at ISO 800 is introducing noise making the edge definition soft.

I note the OP is using a natural subject to express the problem but says he has done appropriate empirical tests.......I would rather see those controlled condition images with full exif and descriptions of setup etc..........the only way to identify potential pilot error whilst handholding.
 
Last edited:
Looking at the LCD shot, it is large JPEG but as I shoot raw I cannot recall the 'order' that the picture style info is indicating i.e. whether no he has left default or the 3 is an increase to sharpening setting.

Whichever......when I shot jpeg I left all such settings at 0 because I always pp'ed on computer and all needed to a lesser or greater degree some sharpening!

The only reason i mentioned it is because often people who shoot jpeg dont want to do any PP
 
The only reason i mentioned it is because often people who shoot jpeg dont want to do any PP
And good point too, hence my canon view of your post ;)
 
if your getting sharp images with the ruler when you do test shots i would suggest it's not then lens

i use more or less the same technique by setting a ruler on a flat surface stood on it's edge on a 45 degree angle and focus in the middle of the ruler ( 6 inches or 15 centimetres take your pick ) see where the focus lies and use the micro adjust accordingly or not depending on the lens
i then repeat the exercise with the ruler vertical at 45 degrees
the camera will be on a tripod with IS switched off if the lens has it

when the images are sharp at the focus point the lens / camera is pretty much set up
if i am still getting soft images in real world situations after setting up i know it's not the camera / lens combination it's me

can you not try the lens on a different body just to see
 
Looking at the LCD shot, it is large JPEG but as I shoot raw I cannot recall the 'order' that the picture style info is indicating i.e. whether no he has left default or the 3 is an increase to sharpening setting.

Whichever......when I shot jpeg I left all such settings at 0 because I always pp'ed on computer and all needed to a lesser or greater degree some sharpening!

Edit ~ I wonder if the 70D at ISO 800 is introducing noise making the edge definition soft.

I note the OP is using a natural subject to express the problem but says he has done appropriate empirical tests.......I would rather see those controlled condition images with full exif and descriptions of setup etc..........the only way to identify potential pilot error whilst handholding.

He's got a sharpness of 3 set. It's sharpness, contrast, saturation, colour tone from left to right.
 
Looking at the LCD shot, it is large JPEG but as I shoot raw I cannot recall the 'order' that the picture style info is indicating i.e. whether no he has left default or the 3 is an increase to sharpening setting.

Whichever......when I shot jpeg I left all such settings at 0 because I always pp'ed on computer and all needed to a lesser or greater degree some sharpening!

Edit ~ I wonder if the 70D at ISO 800 is introducing noise making the edge definition soft.

I note the OP is using a natural subject to express the problem but says he has done appropriate empirical tests.......I would rather see those controlled condition images with full exif and descriptions of setup etc..........the only way to identify potential pilot error whilst handholding.


I havn't got any test shots to upload as I've alreay deleted them but I'll tell you what I've just done 15 minutes ago..

Focused of the house opposite, it's about 20 meters away and nicely lit. My target was the flat brick wall which is a darkish brown brick with light mortar, so you can easily see the detail. Camera was on a tripod with a cable release, manual mode, manual focus, IS off, 400mm, 1/500, f5.6, ISO floating between 100 and about 250.

I took a control shot using the setting above and got it as sharp as I could, and it was beautifully sharp!

I then very carefully switched it to auto-focus, gently de-focused and used to shutter release cable to focus and then take a few shots. De-focusing and re-focusing between each shot. I was super careful to eliminate any vibration.

Results were mixed.. non were quite as sharp as my manually focused control, but some were close. Some were what I'd call sharp and some were not!

It makes it hard to accurately perform a micro focus adjustment because the rustles are so all over the place.
 
Okay, interesting.. when I use a flash (430exii set to ETTL) the photos are perfectly sharp! Only had a quick play around with this in the house over a distance of about 7 or 8 meters but yes, very sharp. The flash didn't send out a focus-assist flash or IR beam either.

Thoughts?
 
Okay, interesting.. when I use a flash (430exii set to ETTL) the photos are perfectly sharp! Only had a quick play around with this in the house over a distance of about 7 or 8 meters but yes, very sharp. The flash didn't send out a focus-assist flash or IR beam either.

Thoughts?

A thought that occurs to me if the flash is improving the situation it may be motion blur.

If the only effective light is coming from the flash gun the shutter speed is being replaced by the flash duration. The flash duration is much shorter (1/4 power on a 403exii is around 1/3000 of a second) so you will have no motion blur.


Some one with more knowledge than me may have a better idea. :confused:
 
A thought that occurs to me if the flash is improving the situation it may be motion blur.

If the only effective light is coming from the flash gun the shutter speed is being replaced by the flash duration. The flash duration is much shorter (1/4 power on a 403exii is around 1/3000 of a second) so you will have no motion blur.


Some one with more knowledge than me may have a better idea. :confused:

Exactly what I was thinking.. But why wouldn't 1/500 or faster with the lenses worldclass image stabilisation be sufficient to stop motion blur? I know it's effectively 640mm on a crop sensor but from what I've heard about the IS, shutter speeds of what I've been using shouldn't be a problem. Had no problem with my old Tamron at 300mm while using shutter speeds of 1/80 and sometimes even slower.
 
Perhaps a fault with the IS and it's actually moving the lens group continuously.....try switching the IS off and use a shutter speed that you can confidently control.

Bob
 
Try using shutter speeds at or faster than 1/(FL x 1.6).

Is it sharp at the short end?
 
Try using shutter speeds at or faster than 1/(FL x 1.6).

Is it sharp at the short end?

Faster shutter speeds help a small amount but I wouldn't have thought I would need to pick it that high with the IS on. Yes very sharp from 100 to about 300.
 
Upon doing more tests I have came to the conclusion that it's most likely not the lens itself. It is capable of achieving sharp images at 400mm and f5.6 and after all, it is a modern L series lens. I think it's most likely the focus system on my 70d.

Results just seem so random though, I can go out side on a nice bright day and it'll fail to capture a sharp image of a subject and setting that you would think was easy. But then, I can be sitting in the house of an evening and I'll take a picture of the other side of the room at settings such as 1/80, f5.6, ISO 1600 and the image is as good as you could hope for!

Just so confusing!
 
Upon doing more tests I have came to the conclusion that it's most likely not the lens itself. It is capable of achieving sharp images at 400mm and f5.6 and after all, it is a modern L series lens. I think it's most likely the focus system on my 70d.

Results just seem so random though, I can go out side on a nice bright day and it'll fail to capture a sharp image of a subject and setting that you would think was easy. But then, I can be sitting in the house of an evening and I'll take a picture of the other side of the room at settings such as 1/80, f5.6, ISO 1600 and the image is as good as you could hope for!

Just so confusing!

Surely you know someone who shoots Canon that you can try the lens on their body? If not, do you have a dealer that you could go to and try it on another crop body. The 100-400 is a very sharp lens even wide open and, from people I know who have one, performs extremely well straight out of the box on a variety of cameras.
 
Would this have any effect?

No. All this is doing is applying sharpening to the jpeg that the camera is displaying on the LCD screen. If you set it to zero then the image will look a bit softer. It won't impact on the ability of the camera or lens to focus
 
Surely you know someone who shoots Canon that you can try the lens on their body? If not, do you have a dealer that you could go to and try it on another crop body. The 100-400 is a very sharp lens even wide open and, from people I know who have one, performs extremely well straight out of the box on a variety of cameras.

I do know a few people but time is so hard to come by at the moment! I might take both lens and body to a shop I know when I get a chance though. See what they think.

Another thing I have noticed when shooting birds in trees and bushes is that when I get an image where the target subject is soft, or had that motion blur look, there is often a random twig or branch near the edge of the frame that looks sharper. This was using centre focus point only and on both one shot and servo. I currently have the tele end micro focus adjusted to -6 and I do think things have improved a bit. Could I just need to MFA the combo better? Results are random though so it's difficult.
 
You could send it back to Canon to get them to calibrate lens to body, you would need to send both obviously. Not too expensive.
Matt
 
I think it's about £80 at Elstree includes postage (secure) both ways, give them a call, they are usually excellent.
 
I have thought about this! Any idea of cost?
If its new then I ask for it free or replaced if it is at fault.
Take a photo of a brick wall make sure your 100% square on and also at each end of the zoom. on a good tripod and fast shutter try different apertures.
 
Back
Top