Intermittent soft focus

[Raskill]

Ahoy all,

I've run into a bit of a problem, which is causing me some concern.

I keep having an intermittent soft focus. I don't think it's me either

It's more pronounced using the flash, however it's also happened in bright day light. I thought initially it may have been a problem with the sigma 70-200mm, but tonight I tried my sigma 20mm f1.8 and it had a problem also.

Very few images of moving vehicles are coming out in focus, which makes me think it's something to do with the cameras AF.

Has anyone else out there ever had this problem? Any advice of AF settings (currently they are centre weighted, continous).

Any help would be great.

Ras.

[gstark]

Can you show us some images?

And please also supply the exif data with the images ....

[Raskill]

Well I figured it out.

My Sigma 70-200mm F/2.8 is soft at 200mm.

Most shots at 200mm are quite soft, some unuseable. Thankfully I got many shots not at this distnace, so I'm not overly worried about it.

My question is, where in Australia can I send my Sigma to get repaired (following the A1 race of course ). Does any one know???

Thanks.

[gstark]

Well I figured it out.

My Sigma 70-200mm F/2.8 is soft at 200mm.

Gee .... a Sigma lens being soft. I'm shocked beyond belief!

Changing the subject, have you had a chance to review this thread as yet?

[Raskill]

I've read it now, as you can see.....

[Raskill]

I've finally got around to posting some 'soft' images. These were taken on an overcast raining day, however the results were the same regardless of weather conditions.



Exit data:
FStop 2.8
Spot metering
Focal length: 200mm
Shutter speed: 250
Curve: sreala 2
In camera sharpening on
No flash

Closing speed of the vehicles was about 100 kmph, however again, this seems to of made little difference, with slower images the same. Focal point was the centre of the bonnett/windscreen.





These ones are the worst so any advice might help me.

I can't see any focal point in the images at all.

I don't want to go and buy a new lens and find oout this ones okay and it's a camera setting

[JordanP]

was your focusing set to AFC or AFS? It needs to be AFC for this sort of shooting as the camera/lens continue to focus as the shot is being taken. After checking your first post it seems this is the case, but always worth checking. Also the shutter speed @ 1/250 is a bit slow with the car coming straight for you and the f2.8 setting at 200mm is also a very shallow depth of field.

You could also be experiencing camera shake at that sutter speed when shooting at 200mm focal length. Were you set up on a tripod or monopod?

If you were panning as the car comes past your shutter would be fine but I'd still try to have a higher aperture and definitely AFC. If you want to really test the lens in terms of focus then I would suggest a still object as these samples could be soft for a number of reasons.

Cheers,

[Glen]

Alan, like Craig, I would suggest putting it on a tripod and see how it looks then at 2.8. Maybe take a nice high contrast item to make it easy to identify softness.

[sirhc55]

Alan - any shot at f/2.8 in the pouring rain is going to be soft The rain acts as a diffuser that will make most shots look as if they have no contrast. I would follow Glen’s lead and do a controlled experiment before sending the lens in

[Raskill]

focus was set to AF-C, single area. Needed the aperture wide open due to extremely poor light. Camera mounted on a monopd and also hand held, with same results.

Just looking at a couple of other images, evening panning in good light, using same aperture, focal length 105mm has produced blurry images.

I'd forgotten that I was using Sreala curve, could this be an issue???

The reason I ask is that if you crop to 100% or larger you can see some bleeding of colours into one and other. Sreala is good for contrasty shots, so perhaps this is making the problem seem worse?



The white stickers on the window are 'bleeding' through shades of blue onto the window, as is the 'corolla' wording on the car body.

Once again folks, your help is much appreciated

[JordanP]

I'd be surprised if a curve is the issue, although I'm no expert with them. Others on this board can offer advice there.

While you had poor light conditions leading you to f2.8, I would have compromised ISO before aperture. Also just as an indication of how tricky it would be photographing an on coming car. I calculated that a car doing 150km/hr would cover about 20cm in 1/200 of a second. Photographing a vehicle at this speed head on not only wants a very high sutter speed but with a shallow depth of field you are asking AFC focusing to respond at 1/500th of a sec or faster.

By testing the lens in a controlled environment you can rule it out as the issue. Then if it comes up clean then you need to see what the camera can successfully do with a depth of field as narrow as 2.8 @ 200mm. What speeds can it handle. Also test at higher apertures when conditions are more favourable.

Best to test one variable at a time. The shots and exif data you have given us have many possible variables.

Cheers,

[Raskill]

JordanP, thanks for the advice. I've done some quick tests that show the lens is definitely softer at 200mm F/2.8 than at 200mm F/5. Even a 200% crop of the F/5 test still showed considerable sharpness, so it doesn't appear that I have lens sharpness issue.

I willl have to run more tests, but thanks for pointing me in that direction.

Thanks again, and thanks to the rest of you also.

[gstark]

JordanP, thanks for the advice. I've done some quick tests that show the lens is definitely softer at 200mm F/2.8 than at 200mm F/5.

This would be normal perfvormance; my expectation is that there would be a sweet spot somewhere, perhaps in the f8-f11 range.

Still, and this is very disconcerting - in looking at the first two images that you've posted, there doesn't seem to be anywhere within the entire photo (for each of them) that appears to be sharp.

While I note and agree with Chris's points about the rain acting as a diffuser, I'd still be expecting better results than this.

Yes, even from a Sigma.

What you are going to have to do is exactly as suggested - shoot some test shots under controlled conditions.

You'll need to do this for a variety of focal lengths, and at a variety of apertures, and also at a variety of focus distances too.

Stick a full page of the Sydney Morning Herald onto a fence. Try to keep it as flat as possible.

Set the camera up on a good tripod, just a bit further away from the target than the lens's minimum focussing distance.

And make sure that it's a very good tripod, and use your remote to fire the camera - you don't want camera shake here.

Shoot a series of images at each of 70, 105, 130, 150 and 200 mm focal lengths, and for each of those focal lengths, at apertures f2.8, 4, 5.6, 8, 11, 16, and 22.

Try that using auto focus, and then repeat using manual focus, noting that once you have the image focussed, it should not need altering.

When you've done all that, move the camera and tripod maybe 20 feet further away from your target, and then repeat the whole lot.

When that's done, move the camera a further 20 feet away from the target, and again, repeat the whole bloody lot!

By now any novelty that a new lens might have had will be well and truly worn off, but you will have a good series of images that you should be able to pixel peep on, and then work out exactly what is going on.

And, into the bargain, you should also learn quite a bit about your lens, including finding its sweet spots.

[Raskill]

Gary thanks for the test details. I'll have to get it done asap, as I'll be attending the A1 GP this week and need the images to be sharp.

Thanks to all, this really is a great community!