Why do I get blur with a telephoto lens even with mirror lock-up?

Question

Trying to fix the blur I see when shooting via a 75-300mm telephoto lens (this one) on a tripod, I enabled mirror lock-up in my Canon EOS 500D, and set a 2-second self-timer, so that after raising the mirror the camera made actual shot after a 2-second delay.

But whatever I try — releasing shutter from LiveView, doing shots without LiveView — I invariably get blurry photos. All this despite seeing a much sharper image in LiveView.

Below is an example of the shots I get with different focus points (which I adjusted using focus single-stepping commands by gPhoto). I shot at 300 mm focal length with aperture f/5.6 and exposure of 1/25 s. Test target was at 3.8 m away from the lens (this is the test environment at home, the problem reproduces with 50 m away objects outdoors). I get basically the same results with and without MLU.

Here's the sharpest version I get in the LiveView:

Below is the comparison of a shot with exposure of 1/25 s and aperture f/8 (blurry frame) vs a shot with 1/60 s exposure with aperture f/5.6 (well-focused frame). Both were taken with MLU enabled.

What could be the reason for this inability to get a sharp image when shooting? Is there any other tool than MLU to help me make exposures of 1/25 s usable?

Answer 1

The mirror isn't the only source of vibration, the shutter causes vibration also. A more solid tripod will help, as well as weights hung on the tripod to minimize the vibration.

What tripod are you using? Post a picture of your setup.

I did some tests years ago, unfortunately, none with mirror lock-up.

http://www.suncoastcameraclub.org/Tips/Tripods.pdf

Answer 2

• Your first example shows the limits of optical viewfinder based PDAF using a dedicated AF sensor array independent of the main imaging sensor combined with the limits of the AF motor of your lens. The system is a so-called "open loop" one. The "ideal" focus position appears to have been somewhere between the third and fourth of the seven frame sequence.
• This is also combined with a slow exposure time (1/25) that also leaves the images susceptible to camera vibration, which can have quite significant impact on images at longer focal lengths such as 300mm.
• Keep in mind that with mirror lockup you may reduce or eliminate the influence of vibrations caused by the mirror movement, but you are not reducing the influence of vibrations due to shutter movement or camera movement caused by an unstable mount/tripod.
• Also note that with viewfinder based PDAF and mirror lockup, AF is done before the mirror swings up. Any camera movement caused by mirror movement will be done after AF has been locked but before the image is captured. If the tripod is also unstable, then the camera position, even if all motion has subsided by the time of the exposure, may be different than the camera position when AF was performed.
• Your second example shows the higher accuracy imaging sensor based CDAF is capable of when operating in a "closed loop" mode. There's also less time between AF and exposure, so less chance the camera is moving as far between AF and exposure.
• Your third example shows higher amounts of motion blur when longer exposure times are used. Either the camera or your target was moving during the exposures. The smart money is on your camera/tripod not being rock-solid stable.

Answer 3

Your last image sequence (comparing two shots at 1/25s ƒ/8 vs. 1/60s ƒ/5.6) does appear to show motion blur in the 1/25s image. Because you have already said you are using MLU, we can probably eliminate that as a vibration source (unless you can test with maybe a longer delay such as 5 seconds). If your tests have tripod fully extended, with an extended center column, try performing the tests with the center column retracted fully in the down position, and collapse the leg sections. Or better yet, set the camera on a pillow or folded blanket or towels on a table and retry the tests (still using MLU). That will completely eliminate vibration issues.

But I don't see much vibration in the first image sequence. I see general softness, lack of focus, and chromatic aberration. I think this is indicative of the lens, not vibration induced by the tripod.

The Canon EF 75-300mm f/4-5.6 III USM is known to be a mediocre lens, at best. From Bryan Carnathan's (the-digital-picture.com) review of the lens (emphasis is his):

[it is a] very cheap lens for this focal length range.

And the 75-300 III really is cheap. You usually get what you pay for - unfortunately this is the case with the 75-300 III.
...
Image quality from the Canon EF 75-300mm f/4-5.6 III USM Lens is also mediocre.

The 75-300 III is soft wide open at all focal lengths. Sharpness decreases as it zooms from 70mm to 300mm. However, the corners are softest at 70 and getting better by 300.

Corner and center softness decrease as the 75-300 III is stopped down.
...
The 75-300 III suffers from CA (Chromatic Aberration). Pincushioning is visible at 300mm.
...
If you care about great image quality and sharp photos, the Canon EF 75-300mm f/4-5.6 III USM Lens is not for you. And Photoshop cannot enhance details that are not there.