I have previously posted a topic where I asked 24-70mm F2.8L owners to post some reference images.
I have also been reading some reviews and it seems that 28-70mm F2.8L is a great lense.
And so I ventured to a local 2nd hand camera store to have a look.

All shots are tested at opposite focal ends, ie 28mm and 70mm at F2.8, F5.6 and F8.
Tests done on 2 different lense copies. EXIF is in the images. Focus area is indicated by the red circle.
F8 shots are taken at a rather low shutter speed, however store attendant kindly provided a tripod for the test. Other shots are handheld while resting elbows on a table. Shutter speed was adequate enough to minimise the motion blur.
I tried to take the shots as consistently as I could.

The images in the post are 640x480 resizes of the centre 3x3 with the link to 1:1 under each image.
No post processing except CaptureOne default noise reduction setting.


First up, 28-70mm F2.8L Copy Nr.1

F2.8 @ 70mm

you got a dud (maybe 2?)

my 28-70 is tack sharp wide open throughout the zoom range.

it is touted as sharper than it's predecessor (the 24-70)

anyone else

greg

My 28-70mm F2.8L is sharp. I can't compare it to the 24-70, but I think that isn't the issue here. I think your testing is not good enough to compare these two fine lenses. Using images from hand held test photos is not good enough, in my humble estimation.

I've posted about mine before. Loved it on my old 350D but on my new 450D and now 40D its not just soft wide open at the long end its really awful. I hope its something that can be sorted in a service but if not its going and I'm going to look at the new Sigma (and take that crap shot gamble) or have to bite the bullet and drop the stop for the 24-105 as they are a good price used where as the 24-70 is just out of my price range which was the reason I got the 28-70L in the first place

What method of AF were you using? these should be way sharper than whats on show, and how close were you cause if your real close movements of only 2-3cm will casue an image to become OOF.

take a lens you know performs very well for you. go back to the store and perform the same test shots you did for the 28-70s. process same way and compare. just a thought by way of a more objective basis for assessment, given the test regimen.

1. Shoot outside in natural light.
2. Shoot at greater distances.

The center should be very sharp at all FLs, with trypical IQ deterioration towards the edges (worse at 28 mm).

EBH

Yep.. my 28-70L is tack sharp as well.. you should send it for calibration or if you got a 5DII, to calibrate it yourself... It could happened with any lens.. By the way, your 16-35L seems to have the same problem as well.. My 16-35L II is sharp too...
BE CAREFUL, for sharpness testing, use the 'one shot' mode selection. When using Ai servo, the quality is bad .. It happened to me when testing my new 35L.. It was horribly soft on my initial tests, since I realized I was in Ai servo mode. When I changed it for One shot, got it sharp again...

Thank you for your replies guys.
Forgot to mention that the camera used was 20D.

Andre: Indeed, I was gutted when I just got the 16-35mm. I went to Canon recently, but they wanted about £80 to calibrate, so I skipped for now. Shooting was "One Shot". I also noticed the AI servo gives soft results. Funny enough, it seems as if 16-35mm is sharper at 2.8 than 5.6..

evan9: I was thinking the same thing about testing conditions. F8 and 1/50s shutter is not ideal setting, but the camera was steady on the tripod. As you can see at F5.6 and 2.8 the shutters varied from 1/160 to 1/640, which should be good enough handheld considering the focal length. My hands were resting on the table and movement was minimal.

Correct me if I am wrong here:
Imho the lense should perform adequately in most conditions. If the lense requires ISO100 and shutters in excess of 1/1000 in order to produce good results, it renders the lense quite useless.

I just don't have a clue as wtf is wrong. Most likely it seems to be my lack of skill.
Prior to 20D I was shooting 350D and in case of 16-35mm the results were exactly the same. Only about 30% of my shots are keepers.
Other than these I got SMC Tak 135mm, 50mm F1.4 and Sigma 150mm f2.8 macro. Can be hard to compare those to a zoom.

Besides, recently I really struggle to get a sharp shot with any of the lenses...

Best Regards

My 28-70L is a good lens.
My 24-70L is a better lens.......a bit sharper for one.

One really has to shoot those two lenses side-to-side in order to be able to compare more reliably.

Obviously, I can not say how representative my 28-70L is of the whole population of that model. All I can say is that I bought the lens in a true LN condition and the focus calibration is right on the money.

I went though 3 copies of the 24-70 trying to find a sharp one. Sent the last to Canon for calibration, not happy with any. Gave up. This was not focus chart unsharp, but noticeable in print unsharp.
You will find two groups, those that praise it as the greatest lens in canon's line up, and those that expected much for for $1200. I don't know if that is because of differing standards for different people, or poor quality control.

this is not a function of the lens used, but you and your camera body.

>>Correct me if I am wrong here:
Imho the lense should perform adequately in most conditions. If the lense requires ISO100 and shutters in excess of 1/1000 in order to produce good results, it renders the lense quite useless.

28-70L is a fantastic lens...

go shoot in the real world not dim boxes in stores & try for yourself.

you can always get (the used lens) calibrated too.

but the fact you took in-store test shots makes me highly question what can satisfy you.

Petkal: I am still interested in acquiring one..=)

Cyrus: Agreed. There are many who praise the lense. Atm I tested one less copy compared to you, so I guess there is still a hope..=)

mh2000: Your comment makes sense to me, however in many threads people are advised to go for the better glas rather than new body.

Glen: I wish I could take the lense outside for some real world shots. I actually did this with my 16-35mm as soon as I got it. The fact that a £1000 was mariginally sharper than my plastic £100 24-70mm Sigma put me off buying any lense without testing.

I simply wanted to hear your opinions on the sharpness of the lenses tested.
Ever so often you see amazing photos taken at something like shutter 1/100-200 at f2.8 using the 28-70mm handheld. Perhaps I am a bit daft, but I fail to understand the condradiction that when the lens will be used for walkaround photography in various lighting conditions, "ideal" conditions are required to determine its sharpness.
Imho, if the lense is sharp - it is sharp (with the obvious exception of very low shutter speeds combined with excessive motion).

Here's a quick test for you guys. My copy is sharper than the 3 24-70L I have owned!

unless you have a lens with IS, the lens itself (other than FL) has nothing to do what *you* can hand hold for sharp photos.

>>mh2000: Your comment makes sense to me, however in many threads people are advised to go for the better glas rather than new body.

unless you have a lens with IS, the lens itself (other than FL) has nothing to do what ss *you* can hand hold for sharp photos.

>>mh2000: Your comment makes sense to me, however in many threads people are advised to go for the better glas rather than new body.

Charbel: This is truly sharp. Thank you for posting the shot!

Can't really make definitive assessments about lens sharpness unless it is manually focused

Dawei Ye wrote:
Can't really make definitive assessments about lens sharpness unless it is manually focused

Very true... otherwise you are just testing AF performance.

I been there recently trying different copies of the 24-70. Do remember that dof is quite shallow especially close range, 70mm, wide open. So shooting on a tripod would be preferable. That said I don´t think the examples are particular bad.

Dawei: Can be a tad challenging on a 20D without focusing screen.. I have to MF my Sigma 150mm as it hunts horrendously. At 150mm or 135mm it is easier to focus I suppose. When I tried MF on my 16-35mm I had a 10000:1 rate of keepers...

J. Andersen: Basically this is roughly what to expect from a Canon L zoom?

Can it be your 20D? Some of the comments made would make me wonder. I had 2 20D's at the same time and did some testing in the basement on various lenses such as the 28-70L and Tamron 28-75 F2.8. Two things came out of that testing, and since I had 2 20D's I was able to narrow it down a bit further. First thing was the 28-70 was sharper than the tamron on the long end while the tamron was shaper on the wide end. Also, the 28-70 takes sharpening better and was clearly the winner. The other thing (which is the reason I am posting) was the viewfinder was misaligned on one. So the AF Point I was choosing was not in the location shown in the viewfinder. I thought it was my problem but turned out to be equipment. Also during the testing, I found that the AF system/lens focusing was indeed affected by florescent lighting (in certains situations and with certain lenses) I could duplicate AF (one shot, tripod) issuses (back focus typically) with several L lenses. However, when using the Tamron in the same testing, there were no issues, so I speculated those new coatings they're puting on lenses designed for digital do make a difference. So, I agree with taking the 28-70 outside for the test and no rely upon one done under inside the store conditions.That's only from my own experience.

Jim: Thank you for the tip. I will probably rent a copy for a day and give it a good and proper testing.

I once owned the 28-70 and it didn`t do anything for me, I sold it. Maybe today and using live view to focus might have been different.