Is my Sigma 10-20 lens crap - opinions please?

wez130

wez130

Steak,wedges and a pint
Suspended / Banned
Messages
2,255
Edit My Images
Yes
Not sure I'm seeing what's wrong wez, they look sharp enough to me.

Last one might be a bit borderline but I'd guess that's a combination of refraction and maybe even a wee bit of shake :shrug:
 
the tree does look a bit soft maybe, was it windy?
 
I think I see what you mean, and it's always possible you've got a duff one - but plenty here sing the praises of this lens

Just a thought... what shutter speeds were these taken at? The trees to me do look a bit soft, but did they perhaps just blow in the wind a bit, i.e. they moved and hence look less sharp?
 
The problem at f8 shows up on the house window to the right at well so maybe a tiny bit of camera shake?

Try another test to see, also post a wider aperture shot if you can, I'm assuming these are all at 10mm from the edge distortion?
 
yes they are all at 10mm but it's the 10mm i bought the lens for, it's not just the tree thats soft, it's everything on that side of the image yet the other side isn't too bad, but i think it also looks a tad noisy for ISO160 too.

Maybe i'm just being too picky, i dunno
 
It does look a bit soft to me, but I can't see the exif to see the shutter speeds. I got one at home I can test later if you don't have any response by then.
 
the 1st one (F/8) = 1/320th second so no chance of camera shake.
the 2nd one (F/13) = 1/125th second
the 3rd one (F/22) = 1/100th second
 
I don't have mine anymore and I don't remember it being that poor but I used to avoid close up subjects at wider angles. Do a search for Sigma 10-20 samples and see what shows up.
 
As far as the noise goes the exif shows some variation in the exposure

f8 * 1/320s
f13 * 1/125s
f22 * 1/100s

The shot at f22 shows the most noise and I suspect Adobe Camera raw has auto corrected the under-exposure and introduced the noise in the process. If the f8 shot was about correct then the f22 is around 1.5 stops under exposed - it should have been around 1/40s based on the shot at f8.

The softening does look worse on the left, I'd try some tests shots of a brick wall so that the distance is more constant - I doubt it's camera shake with 10mm at 1/320s unless you were very drunk!
 
I'm not blown away by the quality, true - ideally it'd be good to see some out of camera JPEGS perhaps, to stop Adobe Camera RAW having its say.

I don't like the edge to edge distortion - the car and the brick wall look horribly stretched and distorted. I guess if you were doing landscape where it's grass/hills/trees you might not notice it, but anything more recogniseable does look 'odd' to the eye.

Cheers,
James
 
ok, give me 5 minutes and i'll go take some of wall at F/4, F/8 and F/13 at 10mm and 20mm at ISO 100
 
I would be interested to see those test shots as I was thinking of buying one of these. Also are you using a tripod? I guess it shouldn't matter at those shutter speeds although might help?
 
Great minds, Tom B, great minds!!

I see you have one of these lenses, do you have any shots we could look at to compare? (don't know if Nikon Vs Canon would affect anything but I would hope not)
 
Moos3h said:
I'm not blown away by the quality, true - ideally it'd be good to see some out of camera JPEGS perhaps, to stop Adobe Camera RAW having its say.

I don't like the edge to edge distortion - the car and the brick wall look horribly stretched and distorted. I guess if you were doing landscape where it's grass/hills/trees you might not notice it, but anything more recogniseable does look 'odd' to the eye.

Cheers,
James
Click to expand...

Anything that is as close to the lens as the wall, at the 10mm end will suffer some distortion relative to the rest of the image. You have to remember that this is only a step away from a fish eye, and as such there is a trade off between getting that extra width and some acceptable barrel distortion. I have found that a lot can be straightened up using skew tools, plugins or whatever without too much loss in quality if is bothering you. Oh and you are right, you notice much less when the objects are more distant, or equally, all are much closer.

Wez, I have had the odd soft image off mine used on a Nikon, but as they have been here and there, I suspect 'user error' in most cases. :nuts:

However, I know someone here, Spence I think, got a Sigma lens recently[cant recall without a search which lens] and was having similar problems, so did get an exchange. But you do need to do some more 'balanced' test shots as has already been suggested.
 
TomB said:
Use a tripod and a remote release/timer if possible just so we can rule out shake.
Click to expand...

I was going to suggest the same. Get it mounted on a tripod with a remote shutter, pointing at some newspaper print and test all combinations of aperture and focal length.
 
Right ok, here are the resluts of my test, instead of linking one by one, i'll link to the folder, they are all named like brickf4-10mm and brickf8-20mm etc so you should work out which is which.

http://www.wesleybrookes.co.uk/lenstest/

All handheld coz i couldn't be arsed with tripod, nor should i have needed one with shutter speeds between 1/100th second and faster. I was stood about 1 metre from the wall. The lens definitely seems better the closer it is to an object anyway.
 
The images don't look that bad to me, un-processed they are fine, Jpegs striaght from a RAW always look flat and need a little tweek. I have this lens and can say I have no problems with it. Waiting for the test shots of the wall as this will be a better indicator if the lens has a problem on the Left side.:thumbs:
 
Looks like a duff copy IMO, my 10-20 is no where near that bad, infact at 10mm f/4 edge sharpness is similar to the center. Either way i would class the above test shots as unacceptable, if your not happy get it swapped!

Edit: And thats some serious vignetting!
 
That's what you get from a sigma ultrawide really. I have the 12-24 and it's pretty much exactly the same and never impressed me.

What you see posted here by people that have them are tiny jpegs all sharpened and treated so they do look alot better. The shots out of the camera are pretty dire to say the least.

The other side of that is, once processed they are not too bad and when you think of all the shots you'l propbably want to take with it, they are likely to be more about the effect than the detail.
 
If you standing with the lens parallel to the wall, I definitely think there is something up with the left side of the lens. It is less sharper than the right side.
 
The earlier versions of this lens exhibited softness at one side of the image - was a major problem with them & a massive amount of the lenses were returned until it was remedied - as much as you may be advised about tripod use (which you will have to be arsed to use if you want any kind of real reassurance) if your version of the lens suffers from this then you will have to look into seeing what can be done.
If it was brand new or a recent model then I think that you can discount this as a possibility but if it is an older version then you may have to do some research.
Hope you get it sorted mate :thumbs:
 
and finally, as MK suggested, here are another 3 shots tripod mounted and timer used somuch slower shutter speeds, all at 10mm.

http://www.wesleybrookes.co.uk/lenstest/papertest10mmf4-1_30s.jpg

http://www.wesleybrookes.co.uk/lenstest/papertest10mmf8-1_8s.jpg

http://www.wesleybrookes.co.uk/lenstest/papertest10mm13-1_3s.jpg

Now these to me look perfectly acceptable and nice and sharp, which adds to my theory that close up it's a good lens, just at distance it loses detail and seems soft. I'm going to run one last set of tests, tripod mounted outside again, be back in 5 minutes :)
 
Did you buy it locally. If you did, take it back and try out another one to compare. Obviously more difficult if bought mail order or on line but if you are not happy and think the lens is faulty, contact the supplier and see about a replacement. Most reputable dealers should do this.
 
Looks fairly even side to side here. I still think that's about right for one of these.

I had the same worries but once you get out and just start using it as normal, you'll feel quite a bit better.

I did all the same test shots as you've done, convincing myself more and more there was a problem. Then I took it out to use it for real and got this shot, which is just as soft as all the tests but doesn't suffer for it at all.

12mmhill.jpg
 
that's a nice photo dazzajl, maybe i am being too fussy, the close ups seem fine to me, just looks like i'm losing detail in the distant pics. I haven't done the last set of tests yet, well i did, but they were all underexposed so need to do them again :D
 
Shouldn't worry. Mine's a lot worse than that down the rhs. I think I got one of the early ones and it was in use before I had chance to evaluate it properly. Also, I wasn't too worried at the time as it works perfectly for my interiors stuff. It wasn't until I started pixel peeping that I realised there may be a problem. Since then I've had some largish prints done of shots I've taken with it and they all look ok.

It only seems to be evident in some situations though, so I don't know what that's all about. Might be my shifty technique :lol:
 
maybe i am being too fussy
Click to expand...

You are and at the same time you're not. You want your new lens to offer a decent quality image but it's not really going to matter.

The shot above is every bit as soft as the tests you've done but even at a decent size, it really doesn't matter at all. Sure I could blow the quality out of the water with my mamiya kit or even just a decent canon lens.....

but if you want superwide, you have accept the downsides of the engineering probelms involved in making such a zoom. I'm sure a dedicated (non fisheye) 10 or 12mm prime could improve on that but until that's an option, we're stuck with what we've got.
 
You must log in or register to reply here.
Back
Top