AFMA'ed the 24-70 mk.ii ... not sure I am loving this copy.

[LetTheRightLensIn]

24mm should be super sharp, it's resolution scores were off the charts and the general consensus was that 24mm was the sharpest spot.  I compared it to the 24LII and while the 24 had less distortion the 24-70II was noticeably sharper.

24 1.4 II is crazy sharp (on photozone didn't get truly insane numbers there center frame? I forget but I think so, my 24 1.4 II is nuts in the center, it drove 5D2 video to constant moire way more than even my 70-200 f/4 IS or almost anything else), but it seemed hard to tell apart from 24-70 II, both crazy sharp at 24mm IMO, the prime maybe a bit more consistent at corners though.

The Photozone review of the 24 L MkII is a bit misleading, as they mark it down due to the soft corners when wider than f/2. However, if you look at their charts, it shows how good it is at f/2.8 and narrower, even in comparison to the Zeiss 21mm and the 24mm TS/E L MkII. If the 24-70 MkII is as sharp as the prime, then it should be razor sharp.

Edited, because of this stupid touchpad .

Not nearly as misleading as my fogeting the word "it" I meant to type "on photozone didn't IT get truly insane numbers there center frame? " a rather different meaning. 
So yeah we agree, PZ did say it ws super sharp there, my point.

[Dylan777]

Ok gents, Bought focal and set it all up. Before Focal I was +5/+6 for w/t

test 1 W with focal said set w=0
test 2 T said set T= -1
test 3 W said set W = -1 (down 1 form last test)
test 4 T said set T = +4 (up 5 from last test)

Should I keep going in endless loops of T +W tests? I know the W setting affects the T setting so when should I call it quits?

K-amps - I ran my 16-35 II and 24-70 II  thtrough FoCal Pro x3 per lens. The results I got are same.

A tripod plays major role in this test, hope you have a decent one. Also, I'm running on window(5D III), therefore, I was required to change the AFMA values manually in camera and small movement could changes the value(s).

[K-amps]

Ok gents, Bought focal and set it all up. Before Focal I was +5/+6 for w/t

test 1 W with focal said set w=0
test 2 T said set T= -1
test 3 W said set W = -1 (down 1 form last test)
test 4 T said set T = +4 (up 5 from last test)

Should I keep going in endless loops of T +W tests? I know the W setting affects the T setting so when should I call it quits?

K-amps - I ran my 16-35 II and 24-70 II  thtrough FoCal Pro x3 per lens. The results I got are same.

A tripod plays major role in this test, hope you have a decent one. Also, I'm running on window(5D III), therefore, I was required to change the AFMA values manually in camera and small movement could changes the value(s).

Dylan... I might need to re-test. I had the 5d3 placed on a table on a rubber mat... I don't think it moved, but with 10+ menu changes per test, I cannot be sure.

Will need to reset the test system to a tripod next... 

[Radiating]

This lens is supposed to be one of the sharpest lenses ever made at 24mm and is supposed to be much sharper at 24mm than it is at 70mm.

http://www.the-digital-picture.com/Reviews/ISO-12233-Sample-Crops.aspx?Lens=486&Camera=453&Sample=0&FLI=0&API=0&LensComp=787&CameraComp=453&SampleComp=3&FLIComp=0&APIComp=1

If the lens is not delivering mind blowing sharpness at 24mm you have a deffective copy. This lens much like the Mark I version appears to have extremely uneven copy varaiation. Early tests are showing 50%-66% of lenses delivering results I would personally be unsatisfied with.

Simply put you should return it and get another copy. +5 MA isn't so good either.

[cliffwang]

Ok gents, Bought focal and set it all up. Before Focal I was +5/+6 for w/t

test 1 W with focal said set w=0
test 2 T said set T= -1
test 3 W said set W = -1 (down 1 form last test)
test 4 T said set T = +4 (up 5 from last test)

Should I keep going in endless loops of T +W tests? I know the W setting affects the T setting so when should I call it quits?

K-amps - I ran my 16-35 II and 24-70 II  thtrough FoCal Pro x3 per lens. The results I got are same.

A tripod plays major role in this test, hope you have a decent one. Also, I'm running on window(5D III), therefore, I was required to change the AFMA values manually in camera and small movement could changes the value(s).

Dylan... I might need to re-test. I had the 5d3 placed on a table on a rubber mat... I don't think it moved, but with 10+ menu changes per test, I cannot be sure.

Will need to reset the test system to a tripod next...

Mounting camera body on a tripod is better idea to test AF IMO.  I have two tripod, one is the Manfrotto carbon fiber one, and one is Manfrotto metal one(about 15 years old).  I use the metal one for AF test because it's more stable.  All my lenses were tested twice on FoCal to make sure the test results.  And all results are same except the cheap lens, Canon 50mm F/1.8 II.

[neuroanatomist]

Should I keep going in endless loops of T +W tests? I know the W setting affects the T setting so when should I call it quits?

W and T are independent, one does not affect the other.

Dylan... I might need to re-test. I had the 5d3 placed on a table on a rubber mat... I don't think it moved, but with 10+ menu changes per test, I cannot be sure.

Will need to reset the test system to a tripod next...

Might?      The two keys to consistency are stability and lighting. 

'On a table on a rubber mat' isn't what I'd call stable (with most tables - we do have analytical balances on rubber mats on tables in the lab, but those 'tables' are solid blocks of marble weighing hundreds of pounds).  Personally, I noticed some inconsistencies when testing on a stable tripod on the main floor of my house (hardwood flooring) - someone walking around, even in the next room, produces vibration.  Setting up the test in the basement (concrete slab) took care of that. 

Lighting should be constant and very bright.  Look over your test results - what are the Ev's you see?  Notice Mt. Spokane's are close to 12 Ev; mine are always in the 11-13 Ev range.  If you're at 10 Ev or less, you need more light.  Personally, I use three 150 W gooseneck lamps at a distance of ~14" from the target.  More light means a faster shutter speed, which further minimizes the impact of vibration.  I just tested my 600 II, outdoors in sunlight with the gooseneck lamps added, Ev's were around 15.

[K-amps]

Should I keep going in endless loops of T +W tests? I know the W setting affects the T setting so when should I call it quits?

W and T are independent, one does not affect the other.

Dylan... I might need to re-test. I had the 5d3 placed on a table on a rubber mat... I don't think it moved, but with 10+ menu changes per test, I cannot be sure.

Will need to reset the test system to a tripod next...

Might?      The two keys to consistency are stability and lighting. 

'On a table on a rubber mat' isn't what I'd call stable (with most tables - we do have analytical balances on rubber mats on tables in the lab, but those 'tables' are solid blocks of marble weighing hundreds of pounds).  Personally, I noticed some inconsistencies when testing on a stable tripod on the main floor of my house (hardwood flooring) - someone walking around, even in the next room, produces vibration.  Setting up the test in the basement (concrete slab) took care of that. 

Lighting should be constant and very bright.  Look over your test results - what are the Ev's you see?  Notice Mt. Spokane's are close to 12 Ev; mine are always in the 11-13 Ev range.  If you're at 10 Ev or less, you need more light.  Personally, I use three 150 W gooseneck lamps at a distance of ~14" from the target.  More light means a faster shutter speed, which further minimizes the impact of vibration.  I just tested my 600 II, outdoors in sunlight with the gooseneck lamps added, Ev's were around 15.

EV's were definitely falling as the test progressed. 5.6 to 6.3EV's. So this is another variable that requires attention. I can take the set up to the basement, but that would mean installing the software on my laptop... IS there any limitation on number of machines this can be set up with?

I am still a bit concerned since even with MF, I am not getting mind blowing sharpness... I guess i will run some more tests and then form an opinion.

[neuroanatomist]

EV's were definitely falling as the test progressed. 5.6 to 6.3EV's. So this is another variable that requires attention. I can take the set up to the basement, but that would mean installing the software on my laptop... IS there any limitation on number of machines this can be set up with?

Yep - stability and lighting.  One can compensate for the other to some extent (more stable means a slower shutter can work, more light means a faster shutter which mitigates vibration).  Looks like you may have issues with both.  I'd move it to the basement and add a lot of light. 

The FoCal manual recommends at least EV 8 (IIRC, the earliest version of the manual recommend EV 10).  One reason I used the extra light even outdoors was that I was using a long, slow lens (f/4, f/5.6 with 1.4xIII), and I wanted shutter speeds as high as possible.  At 15 EV, I was at 1/2500 s with the bare lens and 1/1250 s with the TC.

I don't know that there's a limit.  I can tell you that I've got it installed on one Windows machine (a virtual machine) and two Macs, and I've run it from all of them.  Since you have to connect a camera with a serial number listed in the License Manager, I suspect that's their control mechanism.

I am still a bit concerned since even with MF, I am not getting mind blowing sharpness... I guess i will run some more tests and then form an opinion.

If you're not getting sharp images with contrast detect AF and a well-lit, high-contrast subject, you may want to exchange the lens...

[rpt]

While you retest on a stable tripod, don't raise the center column! I learnt the hard way about this no-no...

[Mt Spokane Photography]

While you retest on a stable tripod, don't raise the center column! I learnt the hard way about this no-no...

=1.  You can't be too stable for critical tests like this.  I get much better results in bright light and on concrete with my very heavy tripod and never extend the center column.  I could have added weight to the under hook, but my results were fine.
I'd not worry about a difference of 1 or 2 points unless the shape of the curve is very peaked.
Check the camera in good light tripod, and live AF.  If its not sharp, no amount of AFMA will fix that, the lens needs to be exchanged.

[K-amps]

While you retest on a stable tripod, don't raise the center column! I learnt the hard way about this no-no...

=1.  You can't be too stable for critical tests like this.  I get much better results in bright light and on concrete with my very heavy tripod and never extend the center column.  I could have added weight to the under hook, but my results were fine.
I'd not worry about a difference of 1 or 2 points unless the shape of the curve is very peaked.
Check the camera in good light tripod, and live AF.  If its not sharp, no amount of AFMA will fix that, the lens needs to be exchanged.

When you say Live AF ... do you mean Liveview/ contrast AF?

[Invertalon]

I am on my second copy of the 24-70 II... First had a decentered element (softness on right side of frame at 24mm).

Both copies were excellent at 24mm (minus right side on first one) and about even at 70mm... No real variation I see between the two. I would shoot a subject in good light outside on a tripod using live-view, then compare that image with the same image allowing the camera to AF. If they don't look identical, your MA is off.

My current copy is at -2 wide and +1 tele with my 5D3 (with some help from Focal). That -2 on the wide end made a BIG difference. I did not think with all the DOF it would matter so much, but it really is required for maximum sharpness. I see a huge difference if I shoot at 0 versus -2... The telephoto end, surprisingly is more forgiving. Strange, but just what I have noticed.

Shoot with live view though and see if the sharpness looks good to you then... If soft still, exchange it. Live view should draw maximum sharpness out of the lens for the most part, not relying on the camera AF system.

[rpt]

While you retest on a stable tripod, don't raise the center column! I learnt the hard way about this no-no...

=1.  You can't be too stable for critical tests like this.  I get much better results in bright light and on concrete with my very heavy tripod and never extend the center column.  I could have added weight to the under hook, but my results were fine.
I'd not worry about a difference of 1 or 2 points unless the shape of the curve is very peaked.
Check the camera in good light tripod, and live AF.  If its not sharp, no amount of AFMA will fix that, the lens needs to be exchanged.

When you say Live AF ... do you mean Liveview/ contrast AF?

Yes, that is what he means.

[K-amps]

When you say Live AF ... do you mean Liveview/ contrast AF?

Yes, that is what he means.
[/quote]

So contrast AF is more accurate than phase AF ?

[neuroanatomist]

So contrast AF is more accurate than phase AF ?

Yes, but slower.  Phase detect uses a separate sensor, and if that sensor or the optics in front of it are misaligned relative to the image sensor, focus will be off (and that why there's AFMA - to correct that misalignment).  Contrast detect uses the image sensor itself to determine focus - nothing to misalign.