70-200 f/4L IS — not always great

Pindy

I'm curious, if any you more experienced types think these are unsharp, suffering from motion blur or just OOF (all 100% crops). I've never been totally happy with this lens and as quick as I am to blame myself, I find I don't have this kind of love/hate thing with any other lens. in addition, I find the lens exhibits the worst contrast of any lens I own, making my RAW processing work harder than with the others. Why do I feel like I should get more keepers?

here they are:

Pindy

These are two different 100% crops of the same shot, thinking maybe the focus point locked onto a different part of the frame. Notice the ghosting on the top part of the feathers? This bird was static, there's no panning or any movement whatever. I feel like I see this ghosting a lot, even with IS turned on which it almost always is:

Pindy

Then again, this lens was also capable of decent sharpness and focus (or so I thought), which confuses me further:

Manfr3d

The first batch doesn't look like motion blur, there is no ghosting
on any edge. My guess is that the focus just wasn't right. Is there
a difference in focus mode in the images? Which camera did you use?

Pindy

These are all using a 30D more than likely in one shot mode, though the BIFs are AI SERVO.

I'm surprised to see what I think is ghosting with IS and 1/1600s, like the seal shot in the first batch. If that's not ghosting, it's a particularly ugly oof look.

Manfr3d

Pindy wrote:

These are all using a 30D more than likely in one shot mode, though the BIFs are AI SERVO.

I'm surprised to see what I think is ghosting with IS and 1/1600s, like the seal shot in the first batch. If that's not ghosting, it's a particularly ugly oof look.

It's not easy to blur an image unintentionaly at 1/1600s. The majority
of the unsharp images seem to have high shutter speeds. Maybe the IS
is acting funny?! Do you have these problems without IS too? If I was you
I would shoot a test row at different shutter speeds with IS and without IS
on the same subject. Handheld and maybe also with tripod.

If you can find the cause of problem, I'm shure you can get the lens repaired
or exchanged.

Pindy

Manfr3d wrote:

It's not easy to blur an image unintentionaly at 1/1600s. The majority
of the unsharp images seem to have high shutter speeds. Maybe the IS
is acting funny?! Do you have these problems without IS too? If I was you
I would shoot a test row at different shutter speeds with IS and without IS
on the same subject.

I was wondering the same thing about the IS, yet it does appear to steady my viewfinder. It doesn't make sense with the higher shutter speeds.

Okay, a test to come. My 5D arrives tomorrow so I will do it then, which will also introduce a new factor. I wonder if it will magically clear up?

If not, it's off to Irvine with it.

Thanks

ziggy53

Manfr3d wrote:

It's not easy to blur an image unintentionaly at 1/1600s. The majority
of the unsharp images seem to have high shutter speeds. ...

That's a very interesting, and astute, observation. I think it deserves more tests specific to shutter speed and use of IS.

Pindy

ziggy53 wrote:

That's a very interesting, and astute, observation. I think it deserves more tests specific to shutter speed and use of IS.

Ziggy, I intend to do much more. Are we in agreement, though, that these look pretty woeful? Part of me was wondering if I had unrealistic expectations.

ziggy53

Pindy wrote:

Ziggy, I intend to do much more. Are we in agreement, though, that these look pretty woeful? Part of me was wondering if I had unrealistic expectations.

I would expect better consistancy, but the "different camera" test should be additionally telling.

Pindy

ziggy53 wrote:

I would expect better consistancy, but the "different camera" test should be additionally telling.

5D Arrived today. Test shots as soon as work allows.

pyry

How much did you crop these shots?

They look oof to me, but I kinda thought (pictures being of birds and all that) that maybe it's over cropping. That would start to show the de-bayering softness from the sensor.

Pindy

pyry wrote:

How much did you crop these shots?

They look oof to me, but I kinda thought (pictures being of birds and all that) that maybe it's over cropping. That would start to show the de-bayering softness from the sensor.

Not sure how you'd define "over" cropping. These are 100% crops from a 30D, so 8MP. I would say they are about 20-25% of the overall size.

Took some pics today, but not to my satisfaction. I need to better document which ones have IS—wish that read in the EXIF. The 5D is such a different animal, being FF. I don't know what will come of the test photos, but already I can see focus accuracy in the viewfinder soooo much better than previously.

pyry

Pindy wrote:

Not sure how you'd define "over" cropping. These are 100% crops from a 30D, so 8MP. I would say they are about 20-25% of the overall size.

Took some pics today, but not to my satisfaction. I need to better document which ones have IS—wish that read in the EXIF. The 5D is such a different animal, being FF. I don't know what will come of the test photos, but already I can see focus accuracy in the viewfinder soooo much better than previously.

I'd say 100% is the limit
Should be sharper than that then.

Have you checked if it's focusing correctly (front/back focus)? That should be an easy fix for Canon's service.

Snowgirl

One thing I've learned with IS is that if you use a tripod or monopod, you should turn the IS off. When there is no potential camera movement (or very little) the IS goes looking for movement and actually creates it! Have had that problem with some shots but if I turned the IS off when using manual stabilization - problem over. Maybe try that?

Scott_Quier

Another IS lesson - give it a chance to stabilize before pressing the shutter. If you don't give the IS a chance to do it's job, you can end up with OOF effects in the photos - lord knows I have a gotten my fair share of these.

Pindy

Thanks for your replies. Every one of the above examples is handheld. I typically turn IS off when mounted. Also, most of these involved me holding the button down to engage the IS AND AF for more than a second or two (sometimes for ages in the case of people and wildlife!) then squeezing the other half of the way to release the shutter. I'm aware of the IS-priming thing, but thanks for bringing it up. Hopefully today I can take better test shots.

claudermilk

Snowgirl wrote:

One thing I've learned with IS is that if you use a tripod or monopod, you should turn the IS off. When there is no potential camera movement (or very little) the IS goes looking for movement and actually creates it! Have had that problem with some shots but if I turned the IS off when using manual stabilization - problem over. Maybe try that?

AFAIK that's an issue on the older IS lenses (e.g., 100-400), but the 70-200/4L IS is supposed to be able to detect that it's on a tripod & adjust its behavior accordingly.

Tee Why

To me, it looks like it's either soft or OOF.
If you are getting sharp and soft images with similar settings, I'd be inclined to think this is a focusing issue.

Pindy

Tee Why wrote:

To me, it looks like it's either soft or OOF.
If you are getting sharp and soft images with similar settings, I'd be inclined to think this is a focusing issue.

Vis-à-vis what you just said, woudl you consider the third batch of "sharp" photos I posted to be sharp and well-focused, for 100%?

Pindy

Also I noticed last night that this is the one I lens with a "cheap" UV filter on—not the usual B+W MRCs I typically spring for. I'm trying this lens without a filter to remove any doubts.

cmason

you might also check your photos in zoombrowser to make sure of the focal pt, it will show you which focal pt was used for the photo. this way you can be sure that the crop area you are considering is the area the camera was focused on.

Pindy

cmason wrote:

you might also check your photos in zoombrowser to make sure of the focal pt, it will show you which focal pt was used for the photo. this way you can be sure that the crop area you are considering is the area the camera was focused on.

good point!

Pindy

I haven't been allowed outside before 2AM this week so doing any high-shutter speed shots isn't going to happen for a couple days. I think, on the new 5D at least, that the lens is behaving itself at low SS with IS on. See what you think:

Pindy

I haven't tracked anything moving, or in broad daylight (what's that?) but now that the Hoya filter is off the lens there should be no handicaps. Memorial Day should be a good opportunity.

harjtt

Pindy wrote:

Ziggy, I intend to do much more. Are we in agreement, though, that these look pretty woeful? Part of me was wondering if I had unrealistic expectations.

Hi Pindy

Just spotted your post and I can see why the lens would be driving you nuts. The first batch of shots were pretty woeful and I had always thought that it was a very good lens - like the 2nd batch of pics you posted.

Don;t know what else I can add as I'm a 4/3rd shooter with regards to the lens - other than whats already been suggested.

Cheers

Harj

ziggy53

I now have the Canon EF 70-200mm, f2.8L IS USM.

I'll be running some tests next week.

Manfr3d

ziggy53 wrote:

I now have the Canon EF 70-200mm, f2.8L IS USM.

I'll be running some tests next week.

Let us know how it compares to your 70-200/2.8L non-IS as well

Cameron

Pindy wrote:

Also I noticed last night that this is the one I lens with a "cheap" UV filter on—not the usual B+W MRCs I typically spring for. I'm trying this lens without a filter to remove any doubts.

That could certainly be contributing. A crummy filter could also result in poor contrast as well.

Pindy

Congrats Ziggy!

Okay, got out this morning with the boys. Apologies for the size of these downloads—100% crops. Note, I haven't WB-corrected some of these and IS was turned on in all shots. Note the mostly fast shutter speeds. Please assume the middle focus point on the shots that have no red square, it's because I use the custom function to focus with the * button, so I would have taken finger off the button a moment before making the exposure:

Pindy

So....

It would seem from the above that I have nothing to complain about on the 5D at least, in terms of sharpness and focus. The lens seems to behave much better than it did on the 30D and AI Servo is faster and more reliable too. The following, however, I think might be evidence of back-focusing. Tell me if I'm being unrealistic, blind or overly paranoid:

The hair to the left and below the target point is more in focus.



Again, focus on hair. Eyelashes seem in focus.



Look at the deadbolt. Nice and sharp. A good foot behind the target point.