Bobu wrote:
jvarszegi wrote:
Your 24-70L is clearly defective, or misfocused in these tests. My old 24-70L was far sharper at f/2.8, and yours isn't sharp by f/5.6 in the posted pictures. The TS looks like a winner (and of course the 14-24 is no slouch).

EDIT: Even this doesn't look as good as my copy did, but still obviously better than the photos posted in the test here: http://tinyurl.com/madlso


A missfocus is highly unlikely with liveview at 10x magnification. It could be that the sensor-plane was not exact parallel to the wall. But if this was the case then each of the three lenses should suffer under this uncontrolled parameter in the same way, because the focus-point was the same for each lense.

It could be that my 24-70 is defective, but this is in my mind also pretty unlikely. I tested my 24-70 against another 24-70 from a friend. His lens had better contrast, but the sharpness was nearly the same. I also tested my 24-70 against my 16-35 (Mark I). My 24-70 was better at all apertures at the whole range 24-35 (after this test I sold my 16-35). And, as I allready wrote, Canon has tested this lens a few weeks ago and I have a writen statement from Canon, that the sharpness of this lens was tested is within spec.

Best regards,

Boris

i'm not the tiniest bit surprised the 24-70 would look like that in the corners, I am a little surprised it would be like that in the center though

i have found that sharpness tests are insanely, ridiculously sensitive to focusing variation and even using 10x zoom LV I need to give it at least 3 separate goes to really start feeling good about the fairness of a comparison.

seanbolton wrote:
Interesting tests and I would agree your 24-70 is not right.
I'll third that.

The 24~70 is the lens that seems to be plagued the worst by sample variation. Should be better than your tests.

YOU'RE FRIKEN AMAZING DUDE.

I was debating whether to get a 14-24/16-9 adaptor, 21 distagon, 24TSE II or a 24/1.4 II stopped down.

The answer is now the 24TSE II. Saving up for one right now.

ILOVECANONL wrote:
YOU'RE FRIKEN AMAZING DUDE.

I was debating whether to get a 14-24/16-9 adaptor, 21 distagon, 24TSE II or a 24/1.4 II stopped down.

The answer is now the 24TSE II. Saving up for one right now.

Don't rule out the 24L II either, although not reflected in the tests conducted by the OP.
It is sharp as heck all the way from 1.4 to stopped down....

here's one from yesterday

epuja, stop. i had made my decision. it's the 24TSE II. stop making me think otherwise argh

Your pic is epic though. It's not full resolution but wow.. The picture must of had a bit of USM and smart resize/bicubic sharper?

ILOVECANONL wrote:
epuja, stop. i had made my decision. it's the 24TSE II. stop making me think otherwise argh

Your pic is epic though. It's not full resolution but wow.. The picture must of had a bit of USM and smart resize/bicubic sharper?

Thanks ILOVECANONL - yes -- i used photoshop unsharp mask in this picture... but regardless, the 24L II is SHARP - and other than being sharp has other very desirable characteristics such as very sharp at f/1.4, autofocus, awesome colors, controlled CA, etc etc etc..

I only wish it had 9 bladed aperture...but then Canon seems to have a thing with 8 bladed....

The 24L II is sharp wide open, I believe you. I was actually going to get it when it was released, until I saw the digital pictures first review of it (he had his noise reduction set automatically so his first 100% crops were soft). That made me get the 35L instead and I regret it because I needed the wider angle.

As for the 24L II, I would imagine it to be just as sharp as the 35L wide open but with less CA. Still deciding on the 24LII with AF or 24TSEII.

As I recall, 24mm is the weak end of the Nikon zoom, so not too shabby, and by the time you get to landscape apertures, there's not a tremendous difference, even at 24mm. The 24 is definitely sharper, however, and it has flexibility that the 14-24 does not, namely tilt and shift. Good to see Canon giving Nikon a run for their money on the wide end of lenses in general.

erichard wrote:
As I recall, 24mm is the weak end of the Nikon zoom, so not too shabby, and by the time you get to landscape apertures, there's not a tremendous difference, even at 24mm. The 24 is definitely sharper, however, and it has flexibility that the 14-24 does not, namely tilt and shift. Good to see Canon giving Nikon a run for their money on the wide end of lenses in general.

Yes - u are correct, 24mm is the 'weak' end of the Nikon zoom - Unfortunately it was the 24mm end that I used 90% of the time

The Nikon zoom is an amazing lens. My only complaints with it are the size...which kept me from using it as much as I'd like to.

Since we are already talking about the Nikon 14-24, I suspect that at least for me, it may find itself on the Buy & Sell soon. The combination of 17 TS-E and 24 f/1.4 II seem to cover my ultra wide angle needs really well (though only time will tell for sure). The 14-24 will still be useful when I need AF in the full range, but I'm finding that its rare.

thedigitalbean wrote:
Since we are already talking about the Nikon 14-24, I suspect that at least for me, it may find itself on the Buy & Sell soon. The combination of 17 TS-E and 24 f/1.4 II seem to cover my ultra wide angle needs really well (though only time will tell for sure). The 14-24 will still be useful when I need AF in the full range, but I'm finding that its rare.

man u have a TON of gear

jvarszegi wrote:
Your 24-70L is clearly defective, or misfocused in these tests. My old 24-70L was far sharper at f/2.8, and yours isn't sharp by f/5.6 in the posted pictures. The TS looks like a winner (and of course the 14-24 is no slouch).


Agreed, that 24-70 doesn't look all that good. While mine isn't perfect at f/2.8 & 24 MM, it is closer at f/2.8 to the performance of the OP's 24-70 at f/5.6.

That 24/TSE II looks excellent. Better than the original 24/TSE.

The Nikon 14-24 performs very well also. Pretty impressive.

The 24TSE is looking amazingly sharp! Just called my store to check my place in the Que and they had some cancellations and I'm getting the first one they will receive

What I find interesting is that it holds up so well at f22. I am not at all surprised at the poor performance of the 24-70, I had one briefly which was about like this, traded it for a 24-105 which was just as bad, and later sold it. The 17-40 also looks like this, but we are told that corner sharpness is not important, or shown images with no corner detail, etc etc.

My big problem is between this and the 24f1.4 mk2. Zooms need not apply.

Tom_W wrote:
Agreed, that 24-70 doesn't look all that good. While mine isn't perfect at f/2.8 & 24 MM, it is closer at f/2.8 to the performance of the OP's 24-70 at f/5.6.


Well, I just did some test shots with my 24-70 on the 5D II. It's not quite as good as I thought - the 5D2 exploits the corners a bit more than my 5D does. It's performance appears to be better than that of the OP, but my f/2.8 appears to be more like his f/4 shots, not his 5.6.

I've been a strong defender of this lens, but it appears to be no match for the fine new 24/TSE II nor for the 14-24 Nikon.

fraga wrote:
Holy crap...
That 24 TSE is sharp beyond belief...
across the whole frame...

No joke!

Nice test but let's see some "real" photos with the 24 TSE II

The shots are very interesting but I'd really like to see how the new 24 TS-E stands against the 24 PC-E. Apples to apples, you know….

Happy shooting,
Yakim.

Yakim Peled wrote:
The shots are very interesting but I'd really like to see how the new 24 TS-E stands against the 24 PC-E. Apples to apples, you know….

Happy shooting,
Yakim.



I'm afraid I can't help you with this, because I don't have the 24 PC-E.

Best regards,

Boris

Bobu wrote:
...

And, as I allready wrote, Canon has tested this lens a few weeks ago and I have a writen statement from Canon, that the sharpness of this lens was tested is within spec.

Best regards,

Boris

Thanks for doing these tests, Boris!

Looking at the 24-70 pictures I think there might be some element de-centered. This is easy to test.

If you do a test shoot of the same wall, but turn the camera 180 degrees between two shots (most easily done by having the camera vertical-left on a ball-head to start with, then turning the camera to vertical-right, and using good pano technique to keep the lens in the exact same lateral location), and and compare the images.

If the colour finge in the shadow of the window-frame seen at the center of the image changes at all between the two images, you can be 100% certain there is at least one lens element that is not properly centered.

I don´t know about the 24-70, but many canon lenses have internal adjustment screws to allow the service to adjust centering and tilt of the lens groups.

When checking a complex lens professionally in an optics lab, a rotating projector is used. On that device the lens under test has a reticle with a test pattern behind it (where the image sensor normally is) and the lens is being continously rotated about its optical center axis.

The projected pattern is viewd at a white screen at a distance of typically 100x the focal length of the lens, giving a 2.4 x 3.6 m enlarged image of the test pattern.

As the lens is rotating, any change in the test pattern is observed. The amount of movement of the test pattern gives an indication of how good the centering is. The lens is then adjusted and re-tested until satisfactory.

BR,

David

Bobu wrote:
Yakim Peled wrote:
The shots are very interesting but I'd really like to see how the new 24 TS-E stands against the 24 PC-E. Apples to apples, you know….

Happy shooting,
Yakim.



I'm afraid I can't help you with this, because I don't have the 24 PC-E.

Best regards,

Boris

Yes, I could figure this up myself.

Happy shooting,
Yakim.

David Svensson wrote:
Bobu wrote:
...

And, as I allready wrote, Canon has tested this lens a few weeks ago and I have a writen statement from Canon, that the sharpness of this lens was tested is within spec.

Best regards,

Boris

Thanks for doing these tests, Boris!

Looking at the 24-70 pictures I think there might be some element de-centered. This is easy to test.

If you do a test shoot of the same wall, but turn the camera 180 degrees between two shots (most easily done by having the camera vertical-left on a ball-head to start with, then turning the camera to vertical-right, and using good pano technique to keep the lens in the exact same lateral location), and and compare the images.

If the colour finge in the shadow of the window-frame seen at the center of the image changes at all between the two images, you can be 100% certain there is at least one lens element that is not properly centered.

I don´t know about the 24-70, but many canon lenses have internal adjustment screws to allow the service to adjust centering and tilt of the lens groups.

When checking a complex lens professionally in an optics lab, a rotating projector is used. On that device the lens under test has a reticle with a test pattern behind it (where the image sensor normally is) and the lens is being continously rotated about its optical center axis.

The projected pattern is viewd at a white screen at a distance of typically 100x the focal length of the lens, giving a 2.4 x 3.6 m enlarged image of the test pattern.

As the lens is rotating, any change in the test pattern is observed. The amount of movement of the test pattern gives an indication of how good the centering is. The lens is then adjusted and re-tested until satisfactory.

BR,

David

Thanks a lot for this info. But this lens just returned from Canon and they said that they tested it and everything was alright. I don't know whether this is true or not, but I don't trust this lens anymore and will probably sell it.
If a lens is decentered shouldn't be there a significant difference in the sharpness between some of the 4 corners (which is not the case with my lens)?

Boris

Bobu wrote:


Thanks a lot for this info. But this lens just returned from Canon and they said that they tested it and everything was alright. I don't know whether this is true or not, but I don't trust this lens anymore and will probably sell it.
If a lens is decentered shouldn't be there a significant difference in the sharpness between some of the 4 corners (which is not the case with my lens)?

Boris



Well, I´ve had Canon US service say a 135/2 was fine, and then later on my local Canon Service Tech in Sweden managed to get it to perform excellently once a slight misalignment was fixed. It took the technician 20 minutes, I think, but then again he is very experienced.

Usually you would get different corner sharpness when there is a misalignment, but a combination effect of off-center and tilt, or two groups off-set in different directions etc can hide the difference between the corners, making all 4 corners bad, especially so on a lens that has already not-too-good sharpness in the corners wide-open to start with.

Anyway it is easy enough to test yourself with the turn-the-camera-180-deg method.

David

epuja wrote:
I only wish it had 9 bladed aperture...but then Canon seems to have a thing with 8 bladed....


Trying to remember, but didn't the mark1 have only 7 blades? I just got my 24LII last night, and looking forward to using it this week. I am also very seriously considering the 24LTS-EII (jeeze, too many 24Ls to keep straight), I just wish somebody with the means would compare both of the mark 2s "stop for stop" to determine a winner.

Chez Wimpy wrote: I just wish somebody with the means would compare both of the mark 2s "stop for stop" to determine a winner.

Mouse over chart to compare, change aperture to taste:

http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=486&Camera=453&Sample=0&FLI=0&API=2&LensComp=480&CameraComp=453&SampleComp=0&FLIComp=0&APIComp=6