[SammyC]

Just bought a second hand Canon 28-105 f3.5-4.5 USM off Ebay and I'm not convinced its as sharp as it should be. Firstly I get the feeling it's focus is off and I'm also not convinced its as sharp as it should be.

Any idea how I can go about checking it or getting it checked/serviced?

[Steep]

Google got me this http://www.photo.net/learn/focustest/ should help you I think.

[CT]

A quick check I find is to mount the camera on a tripod and set it up facing a brick wall outside where there's plenty of light. There's lots of fine texture in the bricks and mortar, and it's all on one plane. Make sure the camera is square to the wall. Take a series of shots at different zoom settings and apertures. f8 tends to be as sharp as it gets as after that you get diffraction effects fom the light bouncing off the diaghragm blades.

Make sure you release the shutter with the delayed timer - preferably a cable release. Use mirror lock-up if you have it.

Not conclusive or scientific but should put your mind at rest ... or not!

[SammyC]

Ok, well I've used the focus test sheet for a comparison of my four lenses (I appologise now for the size of these crops):

25-105 at 63mm and f11


50mm prime at f11


Kit lens at 55mm and f11


Sigma at 168mm and f11



This one is the 28-105 at f4.5 and shows it front focussing. Very weird as it corrects this as you narrow the apperture!

[SammyC]

So, it seems the lens is just a bit less than linear when it comes to the apperture and focusing.
Anyone else experienced this with a lens?

I also think a problem I have is that my kit lens is actually a bloody good one and tends to show up other lenses that should be better. Oh, woe is me

BTW, a mod might want to mark this thread as not 56k friendly now.

[Steep]

Do all the test shots at wide open (e.g. do the 50mm prime at f1.8) testing at f11 will get you nowhere. since the depth of field will overcome most discrepancies.

That said your kit lens at f11 seems to show some serious back focusing.

[SammyC]

Agree about the kit lens there steep.

Taking the shots at wide open will just show the lenses at their worst which isn't what I'm trying to do, I want to show their best to make sure its good enough.

BTW, the last one is the 28-105 at full open.

[Steep]

Well in that case just use it at F8 all the time

Did the tests on my 50mm, 70-300mm and 10-20mm today, The only one that shows a problem is the 70-300 which back focuses at little bit but only in macro mode.

[SammyC]

Oh, right, I see what you're saying.

Well I will get around to doing a full test on my lens to determine its behaviour but I was originally worried that it was broken but it just seems to front focus when wide open.

[Gandhi]

I just hope you're all mindfull of hyperfocal type stuff when you're conducting these tests. It may just be that the shot fo the prime lens at f11 is showing the perfect hyperfocal distance for that fstop and focus range.

Just a thought?

Or am I speaking rectalise?

[SammyC]

WTF! Speak English boy!

[CT]

LOL @Sammy

[Steep]

http://www.vividlight.com/articles/3513.htm

[CT]

Hyperfocal distance is the distance at which you need to focus a lens at any given aperture to achieve maximum depth of field from near to infinity. (Not Here To Eternity )

If you photograph a landscape with mountains in the far distance and little snowdrops in the foreground 3 feet from the camera, and you want both to be in focus, then you select a very small aperture for maximum depth of field. It's no good obviously, focusing on either the mountains or the snowdrops. DOF extends both forwards (towards you) and rearwards from the point at which you focus, and is greater the further you focus away from the camera.

Too bring the mountains and snowdrops both into acceptable focus we need to focus at a point in the mid distance such that when we take the shot and the lens stops down to the selected aperture, DOF extends from the snowdrops to the mountains. The distance we've focused on to achieve this maximum DOF is called the hyperfocal distance.

There used to be very handy depth of field scales on lens barrels to aid with setting hyperfocal distance. Sadly they've now disappeared and we have to rely on the depth of field preview button to stop the lens down before we take the shot and see the effect of the selected aperture on DOF. The downside is because we look through the lens, the viewfinder becomes much darker when using the DOF preview button.

Because Sammy has finely focused on a given point on his test chart, that point should be rendered sharply regardless of the aperture used so I can't see that the hyperfocal distance comes into it at all. It's a bit early in the morning for this, but thanks Spencer!

[SammyC]

SO is it 'normal' for lenses for vary in their back/front focusing as zooms and apps change? I would have though the focusing was done by the body rather than the lens?

[CT]

Originally Posted by SammyC

SO is it 'normal' for lenses for vary in their back/front focusing as zooms and apps change? I would have though the focusing was done by the body rather than the lens?

I dunno if it's normal Sammy - I doubt it. It's certainly not desirable and not the designer's intention. When you think of all those elements in a zoom lens moving backwards and forwards in relation to each other - the effective aperture changing as you zoom - and then add in the fact that it all has to interact with the gearing of the focusing system, there must be a fair amount of potential for mismatches, in fact the less you know about it the better!

Brilliant as modern zoom lenses are in their design, they'll always be compromised to some extent when compared to a quality fixed (prime) lens, although it may be a very fine difference with some quality zooms.

If I were you mate I'd get out at the earliest opportunity and take some 'real' pics with your lens and judge it on that basis.

[SammyC]

Well I bought it to use skiing this coming weekend so I'll see how we get on with it then.