Sigma 17-70: soft?? needs calibration?? faulty??

[TG V1P3R]

Hi all,

I was bought a Sigma 17-70 for Christmas as a replacement to the 400D Kit Lens (non IS version), and initially I was very impressed with the whole package.

However, it's started to strike me recently that I seem to be getting a lot of soft/mushy/out of focus shots from the combination. I've just been testing it now, and even on high ISO like 800/1600, with a decent aperture of 6.3/7.1 etc, fast shutter speed, AF lights red and beeps, yet the resultant picture looks out of focus....there's nothing in the shot that strikes me as even sharp, never mind tack sharp.

So, I'm wondering whether lenses can often be 'pups' from the minute they leave the factory, or they can be supplied 'out of calibration' or something?? Mine hasn't had that much use since December last year, and has recieved zero knocks/drops etc, and been stored in my Lowepro bag very securely.

I remember when I was researching my first new lens I saw some illustration shots with the Sigma 17-70 and I was really impressed by the increase in PQ, but at the moment I just don't feel that's what I'm getting. Does it need to go back to the retailer?? Does it need a visit to Sigma??

Any suggestions are most welcome.

[stevelup]

The focussing is a closed loop system and sharp focus is determined by the camera. In other words, all the camera does is adjust the focus motor on the lens until the camera thinks the image is in focus.

I'd say it was more likely to be an issue with the camera than the lens. Do you still have your kit lens to do a comparison?

[TG V1P3R]

Originally Posted by stevelup

The focussing is a closed loop system and sharp focus is determined by the camera. In other words, all the camera does is adjust the focus motor on the lens until the camera thinks the image is in focus.

I'd say it was more likely to be an issue with the camera than the lens. Do you still have your kit lens to do a comparison?

Unfortunately not - sold it pretty much straight away. It's such a strange one, some lengths/settings come back really good, but other times it looks really mushy I'm no pro, but I'm pretty confident I've a sufficient understanding of exposure settings to ensure that my shots are sharp, so I'm just a bit confused really...

[springtide]

1) Put the camera on a tripod

2a) focus manually and do a series of shots at different apertures.
2b) Repeat the test using Auto Focus.

And compare the results.

You probably want to repeat the tests at different distances.

To add to this...
If you get sharp results with manual focus but not with AF - it's probably a calibration issue with the lens, which means going back to Sigma - it's not a big task, more of a pain than anything else.
It obviously could also be the camera, but less likely.
When testing, use your Base ISO setting (ISO100 or 200) for the best quality.

[chrisbaker42]

Two points.

1. You don't mention using a tripod which is essential if you want to check sharpness of a lens.
2. High ISO should also be avoided as this will also lead to less than perfect shots due to noise.

[danburbridge]

Sigma are somewhat notorious for having rather varied quality control - some lenses are perfect and others can be dogs - sounds like you have a bad copy....

[stevelup]

Some good advice there. It may be that the lens is inherently soft and the camera is focussing it as sharply as it can.

If you can get sharp results with manual focus, but blurry AF shots, then I'd err towards suspecting the camera. If you can't get sharp results at all, then it's going to be the lens.

As suggested earlier, make sure it's on a tripod and I'd advise using the self timer or a remote release for the test shots just so you can be absolutely 100% sure that external factors are not influencing the tests.

[themoron]

I've got a similar problem on a sigma 70-300 APO on a canon 400d. I'm still learning the ins and outs of SLR photography but I only rarely get a nice crisp image. Could be me, the camera or the lens.

I'll have to do as you suggested but the below shots were taken on a tripod, one is sharp to my eyes, one isn't.

http://www.flickr.com/photos/27285539@N08/2574991159/
http://www.flickr.com/photos/27285539@N08/2575814412/

P.S. I hope it doesn't seem like I'm hijacking this thread, just in a similar situation.

[tryingtimes]

Originally Posted by themoron

I've got a similar problem on a sigma 70-300 APO on a canon 400d. I'm still learning the ins and outs of SLR photography but I only rarely get a nice crisp image. Could be me, the camera or the lens.

I'll have to do as you suggested but the below shots were taken on a tripod, one is sharp to my eyes, one isn't.

http://www.flickr.com/photos/27285539@N08/2574991159/
http://www.flickr.com/photos/27285539@N08/2575814412/

P.S. I hope it doesn't seem like I'm hijacking this thread, just in a similar situation.

I think this is probably a different issue to the OP.
They both look sharp to me - apart from I don't think that the intended subject is sharp in either of them. They both have a shallow depth of field which makes this trickier, and in the first, the log bark looks sharper than the dial, and on the second it's focussed on the near petals, not the stamen.
This is easily remedied by using only one of the focussing points (the centre one would have worked for the above images). Or by focussing manually (easier with live view magnified).
HTH
tt

[Sorj]

Originally Posted by tryingtimes

I think this is probably a different issue to the OP.
They both look sharp to me - apart from I don't think that the intended subject is sharp in either of them. They both have a shallow depth of field which makes this trickier, and in the first, the log bark looks sharper than the dial, and on the second it's focussed on the near petals, not the stamen.
This is easily remedied by using only one of the focussing points (the centre one would have worked for the above images). Or by focussing manually (easier with live view magnified).
HTH
tt

First of all i can see that both pics are not evenly tested. Different ISOs and focal distance, of course it will be slightly different. You need to do a same test with AF and manual at same distance ISO settings shutter etc etc. The only thing that should change is the focusing.

I agree also that both pic are quite sharp but one is on 300 with ISO 200 and the other at 70 with ISO 400. Although this might not be too obivious it does change the picture sharpness and speed of the shutter. Higher ISO means it will be more sensative to colours to the point where some pictures being to haze!

[TG V1P3R]

[pics removed as they didn't highlight the problem, and were oversized for the forums as well].

[springtide]

They are not really high enough resolution and/or the picture doesn't show enough specific detail.

What you need to do (sorry should ave said earlier), is to phtograph a book or magazine cover, make sure it has good lighting (the DVD's in the pics might be OK, but they look a little dark), and show 100% crops of the images of a specific part that shows detail (e.g. lettering).

The problem at the moment is that we (I should I say 'I') can't really tell much difference between any of the shots.

Also, (almost forgot this part) - you need to spot focus on the area of the picture that you plan to crop. i.e. not all of the objects in the frame will be the same distance away, so you need to pick one.


Make sense?

BTW, is that Pepper Pig?

[maverick52]

did you shoot in raw or jpeg any pp

try some in raw then try some sharpening

[TG V1P3R]

Thanks Springtide - I'll do that more specific test later today hopefully, and post back the results.

And yes, Jr Sp4rt4n was watching Peppa Pig at the time

[TG V1P3R]

Haven't got round to making some more test images, but I'm wondering the following:

Is it possible that a lens could start off as performing 'correctly' and then get progressively worse?? The reason I ask is that I've just been looking back through some of my saved images, and some from a few months back look really really sharp (though not particularly photogenic, ducks and other random stuff)....is it possible that a Sigma lens could start out as correct and then deteriorate??

All this is leading me to believe that the fault lies with the [out of warranty by a month] camera

EDIT: Mod comment about 300k embedded images duly noted.