So, just picked up a 70-200 2.8 II, and...hmm...

Hey all.

All my new lenses i run through a simple sharpness test using a bob atkins chart. I picked up the 70-200 2.8 IS II on the current Adorama deal (after agonizing on that wallet hit- OUCH!!) However experience has showed me that canons quality control generally sucks..the only lens that i've bought that wasn't exchanged was the 50 1.8, and the 85 1.8. all my others have been exchanged for 'better' copies.

So to get to the point- this lens is renowned for its sharpness, and i always refer to the iso charts here as a guide of how a lens should perform. My xsi is my reference body- as my 7d is schizophrenic and exhibits too much noise (off to the shop, sigh).

I use a bob atkins chart against a wall. The lens shots so far look good, contrasty, consistent..but not what i would call really sharp. no hard edges on any of the lines. Ive taken several sets of shots at 2.8 through 70, 100, 135, and 200mm. they look ok. I put my trusty 24-105 back on the chart, and it is waaaay obvious that the sharpness is immaculate. every line and edge is 'hard', not a hint of soft anywhere. The 70-200 was meant to be shot at 2.8- and if its not going to be great here, then that affects the rest of the zoom range. Thats not good.

So of course, with the price of these lenses, im not settling for anything less than heaven. Ill run some more test, and ive compared my results against the results from the iso charts here, and the tests here looks sharper for sure.

Time to switch to Nikon :)

Yes the test shots here are sharper because they were shot using a 1Ds Mark III. I suggest you retest your lens and pay attention to your testing method to rule out errors. Post some test shots so other can see too. Welcome to the forum.

Welcome!

As stated, some examples would help.

How are you focusing your test shots of the chart?

My first thought was "focusing issue." Definitely rule that out before returning the lens.

here are two wide open, 70mm screenshots at 1:1 from lightroom 3.5. I saw that if i open the files directly in raw--> photoshop, the processing alters of the results, even at 300dpi. I would post the raw file if my upload wasnt so slow- let me look into that.


Attachment 300Attachment 301

My method is decidedly simple because my reference body is 'tuned' ... it produces extremely sharp photos. all of my lenses produce consistent performance on this body. i dont change the settings on it, so the results should always be predictable. Taking away the body, its rather easy for me to implicate the lens. This way has worked wonders for me for the last year.

As well, using this sites comparison tool, the 24-105 is definitly sharer at certain focal lengths wide open, but my question is do you think the shot of the 70-200 looks sharp to you. It does not to me, and ive done these chart tests a million times.

I focused the 70-200 by using the appropriate option (1.2 m- infinite /2.5m- infinite) based on the distance to the wall- which is less than 6 feet. No IS, tripod based, hood attached. I set the focus to infinity, then use autofocus to set the shots individually, on a 2 second timer. I dont use mirror lockup, as i wont be using that in the field so its a wash.

I should have been more specific. Did you use autofocus for the shot? (Sounds like yes, from your description.). Assuming you did, most likely you need to do an AF microadjustment.

To test lens sharpness, use manual focus with Live View at 10x. If it's sharp like that, not with AF, that confirms it.

AFMA tips: http://www.the-digital-picture.com/P...ment-Tips.aspx

i did use autofocus for th shots. microadjustment isnt available on the rebel bodies, and i cant trust my 7d for consistency :( if its good on the rebels, its good everywhere (or should be- then you can blame the bodies and commence with the MA). I had a horrible time with the 7d's micro adjust...testing that is NOT fun.

Ill try the manual focus, but ive never found it to be terrible spot on.

Attach the lens to your 7D. Set the 7D up on a tripod. Light the test area well. Use the 7D's Live View with x10 magnification to manually focus the lens. Once you do that, take the picture (or even better, activate the shutter remotely using a wireless remote or a corded one). Then analyze your results.

Without doing the steps above, you are more or less testing focus accuracy more than lens sharpness. Isolate the variables.

ashmadux

If you use the method Sean describes you will be able to tell if the lens is sharp or not. I think you will find that the lens is sharp using this method.

Your logic that your rebel body is not the problem is not sound. No doubt the camera may perform flawlessly with other lenses, but that doesn't mean it will play nice with all lenses. Bodies seem to vary from one to the other in how they interact with certain lenses.

When I bought my 70-200mm I set up tests to check it. After several thousand test pictures using the lensalign pro, on my 5D II, 1D IV and 7D I found something out. The 70-200mm has a focus curve from one end of the rage (70mm) to the long end (200mm). It will front focus slightly at 70mm and back focus slightly at 200mm. The lenses focus falls within the acceptable area through the whole range. This phenomena was hardly noticeable when using the 5D II, using the crop bodies it became more noticeable.

Even if your new 70-200mm passes Seans test for sharpness, it will not correct your problems. If you are not getting sharp focus using AF you have three options (2 in the case of the rebel), Do an AF Microadjustment, send the lens and bodies to Canon and have them do an adjustment or send the lens back for another copy and hope it plays well with all your camera bodies.

Quote:

---

Originally Posted by ashmadux

if its good on the rebels, its good everywhere.

---

Sorry, but Rick is correct - this logic is unsound. My 17-55mm f/2.8 IS was perfect on my T1i/500D, and front-focuses on my 7D. My 100mm f/2.8L Macro IS back focused on my T1i (which was a significant factor in my decision to get a 7D, for AFMA). On the 7D, the 100L is perfect, but on the 5DII it front focuses. Each lens-body combination is somewhat unique. Take a pair of lenses that both front focus on one body, put them on another and one may still front focus while the other back focuses.

Also, an f/4 lens isn't the best 'gold standard'. The deeper DoF at that comparatively narrow aperture will mask AF errors that would be evident with faster lenses.

I hear you guys- cant argue with your reasoning. Ill give it a try. I may not even have the time for much testing since its the holidays...more than likely its going back for an exchange within 6 days.

the 7d is going back to the (canon)shop, already boxed up, but i may pop that open to try this tonight. The 7d has issues...way too much iso noise, and softness fluctuates from shot to shot. i compared shots on a t2i rental from months ago to any/all of my 7d shots..7 loses. As of now, i dont trust a 7d, and i may toss it in favor of a 60d.

Quote:

---

Originally Posted by HDNitehawk

ashmadux

If you use the method Sean describes you will be able to tell if the lens is sharp or not. I think you will find that the lens is sharp using this method.

Your logic that your rebel body is not the problem is not sound. No doubt the camera may perform flawlessly with other lenses, but that doesn't mean it will play nice with all lenses. Bodies seem to vary from one to the other in how they interact with certain lenses.

When I bought my 70-200mm I set up tests to check it. After several thousand test pictures using the lensalign pro, on my 5D II, 1D IV and 7D I found something out. The 70-200mm has a focus curve from one end of the rage (70mm) to the long end (200mm). It will front focus slightly at 70mm and back focus slightly at 200mm. The lenses focus falls within the acceptable area through the whole range. This phenomena was hardly noticeable when using the 5D II, using the crop bodies it became more noticeable.

Even if your new 70-200mm passes Seans test for sharpness, it will not correct your problems. If you are not getting sharp focus using AF you have three options (2 in the case of the rebel), Do an AF Microadjustment, send the lens and bodies to Canon and have them do an adjustment or send the lens back for another copy and hope it plays well with all your camera bodies.

---

Spot on (no pun intended)

After performing an initial test and comparing images side 2 side (MF/AF) its just like you said- sharper in MF- the AF is producing a softer picture throughout the zoom range- though the lens seems soft at 200 mm using either AF method. Personally, I dont want to have to deal with glass that is inconsistent from body to body, as impractical as that may sound.

Ill try the MA tonight and see how that goes, but its leaning towards the AF not producing the best it can. Ah well, it will likely go back to the shop tomorrow. Crossing my fingers for a second version.

ashmadux

Let me qualify my statement a bit. The 70-200mm is the best zoom canon lens I have owned. If your version isn't working with your AF system, sending it back is what I would do as well. It has always been my opinion that if you spend $2,000 for something it should work as intended.

What a camera needs and does not have is the ability to do a AF manual adjustment at both ends of the zoom range.

When I did my testing I did over 1000 shots. The focus curve I spoke about seems to be S shaped, meaning in the center ranges it is less noticeable, and feathers off abruptly when your at the widest and longest. I suspect that it is just the nature of this lens and just how the mechanics work. Even though it works like this, in real life situations it was not significant enough that I ever notice. When I say it is off I mean it has a bias for the DOF to shift toward that direction. You should be getting super sharp results, I do with mine. I wouldn't discount the lens because it works this way. Each lens has its own unique properties, and if you know those properties you can work the lens to its potential.

You should know also by looking at the charts that the 70-200mm is at its sharpest at 200mm, and less so at 70mm.

I had a 135mm lens that had an autofocus issue. Back focus. Sent it to Canon with my body, they calibrated the lens to my body (because my other lenses worked fine), and it came back behaving superbly. Easier that than taking a chance to roll the dice.

@Collin what, if anything does Canon charge for that? If it's in warranty is it covered?

It was in warranty, and it was free, at least the first time, but it hasn't needed a second.