24-70mm backfocusing and sharpness UX copy

it seems that my copy seems to be backfocusing on 24mm wide open. just recently purchased this from a member here. cool guy. i dont think he know but who knows. im just glad its a UX copy so i can send it in to canon if i have to. here are some test i took today. they arent like any hardcore test or anything so if u guys have better suggestions pls let me know. i specifically center focus on the word "perfectly" in that sentence on both images. also if u guys can comment on its sharpness. seems its decently sharp at 70mm but 24mm seems a little soft.

70mm @2.8

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by FLICKR

Someone else with more knowledge can chime in, but from what I've read about the lens it seems this is a somewhat common occurrence. Good thing is it also sounds like those that have had the problem were able to get it fixed by canon. Good luck.

Check out this thread for a discussion of it: https://photography-on-the.net …?t=743396&highl​ight=24-70. I think it was toward the end of the thread.

yea i understand that it is also but it seems like canon is either not caring about their QC or they dont want to fix the issues with the method they are making these lenses. this is a UX copy so i figured they figured it out by now but i guess not.

Unfortunately, with the way you positioned the focus test chart, this is not a very valid test. You really need to make sure the focus test chart is resting on a flat surface, your camera needs to be on a heavy enough tripod, MLU has to be enabled, and a remote switch (or the timer, set to 10 sec) is required to activate the camera. Furthermore, you should enable the centre AF point only for this test.

Furthermore, you need to make sure th ecamera is aligned perfectly with the test chart according to the instructions that come with it. This includes focusing on the fat black line with "Focus here" inscribed. The text you focused on is just not good enough fo this test, considering the size of the AF sensors.

Regardless of all this, and having doen a zillion of these tests (ok, slight hyperbole, but quite a lot regardless ), I would say your lens is fine, and a proper test will prove that. Unless you notice problems when shooting real life photographs, don't even bother with this test, I'd suggest.

A final word: this test was not designed for a Canon AF system. According to Canon literature you need a vertical rather than a horizontal focus bar for correct focusing assessments. Furthermore, the focusing point shoudl really be at approximately 50X the focal length of the lens used, i.e., 1.2 m (4 ft) at 24 mm, and 3.5 m (11 ft 8 inches) at 70 mm.

HTH, kind regards, Wim

sdsviet wrote in post #8540596
yea i understand that it is also but it seems like canon is either not caring about their QC or they dont want to fix the issues with the method they are making these lenses. this is a UX copy so i figured they figured it out by now but i guess not.

I reckon that before you make such comments, you better learn how to carry out tests properly first of all.

This is pure user error. See my other post.

Kind regards, Wim

wimg wrote in post #8540620
Unfortunately, with the way you positioned the focus test chart, this is not a very valid test. You really need to make sure the focus test chart is resting on a flat surface, your camera needs to be on a heavy enough tripod, MLU has to be enabled, and a remote switch (or the timer, set to 10 sec) is required to activate the camera. Furthermore, you should enable the centre AF point only for this test.

Furthermore, you need to make sure th ecamera is aligned perfectly with the test chart according to the instructions that come with it. This includes focusing on the fat black line with "Focus here" inscribed. The text you focused on is just not good enough fo this test, considering the size of the AF sensors.

Regardless of all this, and having doen a zillion of these tests (ok, slight hyperbole, but quite a lot regardless ), I would say your lens is fine, and a proper test will prove that. Unless you notice problems when shooting real life photographs, don't even bother with this test, I'd suggest.

A final word: this test was not designed for a Canon AF system. According to Canon literature you need a vertical rather than a horizontal focus bar for correct focusing assessments. Furthermore, the focusing point shoudl really be at approximately 50X the focal length of the lens used, i.e., 1.2 m (4 ft) at 24 mm, and 3.5 m (11 ft 8 inches) at 70 mm.

HTH, kind regards, Wim

i totally understand what you are saying. i just take a couple with a tripod on a table at the right angle but i did not use a shutter release. the results are pretty similar. i just took these really quick because that area was where i had the most natural light. pretty gloomy day today. i will probably take a more complete test later this evening. but yes i havnt actually gotten a chance to use in real life situations yet but i will and let u guys know.

sdsviet wrote in post #8540660
i totally understand what you are saying. i just take a couple with a tripod on a table at the right angle but i did not use a shutter release. the results are pretty similar. i just took these really quick because that area was where i had the most natural light. pretty gloomy day today. i will probably take a more complete test later this evening. but yes i havnt actually gotten a chance to use in real life situations yet but i will and let u guys know.

Please do, before worrying about it.

Do note that at 24 mm it will always resolve less detail (hence may potentially seem or feel softer), for the simple reason that the AoV is much wider and hence that less pixels are available for the same part of the picture as there is at 70 mm. The sensor surface area covered by the same object at 70 mm vs that object at 24 mm from the same shooting point is approx. 9X bigger at 70 mm than at 24 mm, and thus, at 70 mm, will (apparently) show quite a bit more detail and potentially look or feel sharper.

Kind regards, Wim

as I understand, DoF is usually 1/3 front and 2/3 back. so it looks like your lens is operating properly

wimg wrote in post #8540620
Furthermore, the focusing point shoudl really be at approximately 50X the focal length of the lens used, i.e., 1.2 m (4 ft) at 24 mm, and 3.5 m (11 ft 8 inches) at 70 mm.

HTH, kind regards, Wim

Hi Wim,

This focusing distance you are talking about. Are we measuring from camera sensor plane, mid lens length or a certain element within the lens?

Also, what is the best way you've come up with to accurately and consistently setup this distance?

Thanks

MS-18E wrote in post #8541125
Hi Wim,

This focusing distance you are talking about. Are we measuring from camera sensor plane, mid lens length or a certain element within the lens?

Also, what is the best way you've come up with to accurately and consistently setup this distance?

Thanks

That's 51X the FL from the mark on the camera body next to the flash shoe lower down on the body (the zero with long strike through it), which indicates the position of the film or sensor plane.

It doesn't really have to be exact, approximate is good enough. The reason I indicated this is that many people shoot very close to the focus test chart, and this isn't really the correct way to do it, certainly not according to Canon.

The way to set this up as correct as possible is to align the camera body along a perfectly straight 2X4 or something similar, the focus chart along another one, equal in length (rulers will do too), and measure the distance to the 2X4s (or rulers) at the ends. They should be the same. Alternatively, an electronic measuring device should do the trick too, as long as you make sure the sensor is perfectly parallel to the focusing bar on the above focus test chart.

The distance from the centre of the lens to the focus test chart focusing bar should be positioned in such a way that it is approx. 50X the FL, or the sensor plane mark about 51X.

As long as you don't do MTF-charts, approximate is good enough .

Kind regards, Wim

garycoleman wrote in post #8541064
as I understand, DoF is usually 1/3 front and 2/3 back. so it looks like your lens is operating properly

That's really only true for focusing distance further away, say 100X FL and further. Nearby it is approx. 1/2 - 1/2. However, for F/2.8 lenses it is only accurate for about 1/3 of DoF, so 1/6 - 5/6 either way is within spec, and for slower lenses as long as it is within DoF it is within spec.

You'll find that most lens/body configurations are more accurate than that however, provided the AF-sensor is placed well on a good contrast transition.

Kind regards, Wim

can u suggest a better method to do this test chart or a link to it. from the instructions i'm reading it was asked of me to take up as much of the test sheet as possible resulting in being really close to the sheet during the 70mm shot.

I tested a 24-70 today on my XS, and the result is opposite of yours. It is focus properly on the 24mm, but it is a bit back focus at 70mm, very very similar to what you have posted.

Another strange thing, I tested it on a 50D, it focus accurately at all FLs except 50mm! And again, the result is very similar to what you see, it's just it happens on 50mm.

Can anyone explain why does a lens behave differently on different camera body?

I used that chart just for kicks when I got mine.. I got pretty much identical results as you.. and I used a tripod, MLU, timer, flat 45 degree plane and focused on the correct bar.

So then I did a whole lot of battery tests, which while people drill the battery testers.. is a more effective real world test if done right. Out of 20 or so shots from 24-50mm after defocusing in between each shot, every single one was excellent. Proper battery was always on focus and the others were always blurred the same amount.

Mine has never missed focus in actual real world use, except shooting dark nightclub stuff. I will just stay away from shooting charts and I should be fine. They are boring to shoot anyway.

this is so sad....nit picking instead of going out there and using it for what it's intended....sure it's easy to say equipment failure rather than camera user's inability...so sad....so so sad....i guess you can post your chart and we can admire how sharp that piece of sheet is.....