I have a D300 and recently got my 24-70mm/2.8. Life should be good. Well, not really.
I have never been someone who spends a lot of time testing equipments. I guess I have been lucky. But then I recently noticed something wrong with my new toys: when I took some portrait photos with the D300/24-70 combo, if I set my D300's focus point to the top of the image, it always came out a bit soft. If I focus at the center, it is incredibly sharp. After a lot of tests (tripod, flash, cable release, try other lens, etc), I sent the lens to Nikon for repair, and got it back today. Alas, the problem is still there.
So to illustrate my point: here are two images:
center focus: http://www.flickr.com/photos/stevexfu/2734673254/sizes/l/
focus area at 100% of above image: http://www.flickr.com/photos/stevexfu/2733884283/sizes/l
right focus: http://www.flickr.com/photos/stevexfu/2733843185/sizes/l/
focus area at 100% of above image: http://www.flickr.com/photos/stevexfu/2734717104/sizes/l/
You can see when setting the focus point to the far right, the image is noticeably softer. I notice similar behavior with other lens, but much less degree than the 24-70mm/2.8. Certainly more so at f/2.8 than at smaller aperture.
Now I am not sure if it is my D300's problem, or if I have unrealistic expectations. Is that kind of softness expected at the edge of the lens (at f/2.8)? Or should I now send my D300 for Nikon to check it?
Thanks for reading.
#1. "RE: D300/24-70mm focus problem" | In response to Reply # 0lovemy8514 Registered since 05th Oct 2007Tue 05-Aug-08 04:12 PM
It looks like the focus point is actually a bit to the left of the red, boxed area on the second shots.
I look forward to replies from knowledgable individuals!
J a m e s
Using his camera as a pen, it is the photographer's job to tell a story: Each page authored in frozen moments of time.
All of my work is dedicated to my father, Terry Lee Geib (1943-2009)
Visit my Nikonians gallery.
Visit my Nikonians gallery.
#23. "RE: D300/24-70mm focus problem" | In response to Reply # 1Tue 26-Aug-08 03:38 PM
I finally got an answer that appears to be reasonable from another Nikon tech about this:
Thank you for clarifying. The NEF files you sent displayed the issues you were stating. The other images appeared as depth of field issues. I believe the reason you are seeing this has to do with the fact this is not flat glass elements but curved. For example our 105 mm macro and 60 mm are designed to correct this because of the applications in which they are used, close-up macro photography.
The 24-70mm is not designed to be used as a macro lens because of the curvature in the lens can cause slight variations in not focus but distortion. This lens is meant for more general usage, like portraits, landscape etc. Not close-up.
The center point is going to be sharper in appearance than the side focal points. The aspherical lens elements do help correct this but there is still going to be some slight difference. I think if you shot with the lens how it was intended, you will not see these issues.
#26. "RE: D300/24-70mm focus problem" | In response to Reply # 25Wed 27-Aug-08 12:55 PM
Not yet, I have a big party coming up that I'll be bringing the camera and I'll compare the results. That's probably a better test. If it works ok in real life shooting, then I'm ok with this.
#27. "RE: D300/24-70mm focus problem" | In response to Reply # 26Mon 01-Sep-08 01:46 PM
I got my lens and camera back from Nikon. As expected the problem is still there, even though they did a list of adjustments, checkings, and cleanups. And this is the second repair attempt. I have no reason to believe a third repair attempt would make any difference. This is definitely a design problem/limitation.
atempest's last comment from Nikon kind of makes sense. I re-tried my tests. It appears the problem is only visible when focusing targets are within closer range, like 3 feet or less. If I stay longer distance from the targets, the problem went away.
I find it a bit lame to call 2-3 feet range a close-up or "macro". But I guess that is the designed parameter space I have to live with.
#28. "RE: D300/24-70mm focus problem" | In response to Reply # 27TEITZY Registered since 14th Mar 2007Tue 02-Sep-08 10:14 AM
>I find it a bit lame to call 2-3 feet range a close-up or
>"macro". But I guess that is the designed parameter
>space I have to live with.
I've had 2 copies and both were very soft in the corners at 50 cm (just less than 2 feet) on the D300 & D70s. Even after calibration with my body the original copy is still very soft on the edges and ordinary in the centre at the wide end even shooting close to infinity at f8! Funnily enough in the 50-70mm range the lens is spectacular even in the corners at normal shooting distances (1-3m). Edge sharpness is not critical for me at the long end but it's definitely a must @ 24mm. Any way the lens and body are back at Nikon at the moment and I'll let you know what the outcome is.
#3. "RE: My guidance is try another test target." | In response to Reply # 0
Thanks for detailed information - much more helpful than queries like "my camera does not AF - full stop"
For some subjects an outer D300 "single line" sensor is less likely to focus accurate (or even at all) than one of the 15 central cross type sensors.
What I cannot tell despite your 2 examples is whether the problem is expecting too much of an outer D300 AF sensor (which is not a camera or lens fault) or something else which needs attention.
To understand what might be going on locate a framed picture on a white wall.
The white wall should have insufficient detail for AF to lock on.
Using a central sensor aimed at the white wall above the picture take and hold first pressure on the shutter and the AF hunts - keep first pressure on the shutter and move the camera down. When the AF point reaches the border of the picture frame the camera should focus when held in landscape or portrait position.
Repeat using an outer AF point and AF locks with the camera in landscape mode but is unlikely to lock with the camera in portrait mode.
What this clarifies is a central cross type AF point reads horizontal as well as vertical detail whereas an outer point (with camera in landscape position) reads only vertical detail.
Now for a D3, D700 and D300 "fault".
the AF indicators imply the camera always reads left/right (camera in landscape position) which is not true. With outer points the camera reads top to bottom - the opposite of the AF indicator in your outer point AF example
There is a possible issue with your AF target.
It is aligned just off center with your outer AF area example.
The narrow line likely to be read by the AF sensor probably ran just left of center along the similar to a vertical black line which contains very little readable contrast difference. If it did run along the black "line" there would be AF difficulty for a vertical reading AF outer line.
In the central AF area example the accurately lined up extra horizontal AF reading line runs through some good contrast and should read your target very well - which it has done.
One issue I have found with the D3 and D300 AF is with outer sensors and some AF targets shifting the viewfinder a very small amount makes the difference between accurate and soft AF lock on.
If you can find another test target with good readable detail in the vertical direction and exactly center the outer AF point you should have a better idea of whether you were expecting too much of some aspects of D300 AF ability, or if there is some other problem. One possibly is something like the alignment of the outer AF sensors in the mirror box.
Photography is a bit like archery. A technically better camera, lens or arrow may not hit the target as often as it could if the photographer or archer does not practice enough.
#4. "RE: My guidance is try another test target." | In response to Reply # 3Thu 07-Aug-08 07:03 PM
I just received my D700 and a new 24-70mm f/2.8 earlier this week. I'm seeing this exact same problem. I can use the center focus AF point in up, center, or down. But when I use any left or right focus AF point, I loose sharpness if I crop in at 50% or more. This problem only occurs at F/5.6 and higher opening such as F/2.8. F/5.6 and smaller seems to be fine. I also tested the 50mm f/1.4 with the same result. All of my tests are with D700 set on a tripod in daylight good light. I also tried the Liveview test and the result is the same. Tried this with Single Servo, Continuos Servo, 3d Matrix, Spot metering, dynamic AF, single AF, and various different ISO. I tried using the AF fine tuning as well, no help.
I finally called Nikon tech support and send them the pictures and they told me that my AF system in the camera needs to be tuned. So I'm going to be sending them my D700 and the lens.
#5. "RE: My guidance is try another test target." | In response to Reply # 4Thu 07-Aug-08 07:06 PM
Here're my sample photos:
Test shot with my dog (no tripod, low light)
Center AF pic (tripod)
Right AF pic (tripod)
Center AF pic 100% cropped (tripod)
Right AF pic 100% cropped (tripod)
#6. "RE: My guidance is try another test target." | In response to Reply # 5Thu 07-Aug-08 11:40 PM
Looks like I am not the only one who saw this problem.
I went into a camera shop today and tried their D300/24-70mm and D700/24-70mm. In both cases I found the same problem! I am quite discouraged by this and thought this might just be a design problem of the focusing system.
Please update us with your result from Nikon. Maybe they have some way to fix it...
#8. "RE: Yes but (again)" | In response to Reply # 6Fri 08-Aug-08 08:32 AM
>Looks like I am not the only one who saw this problem.
- with some AF targets the AF sensor needs very careful placement to get precise AF - and Nikon give guidance on AF target selection in all camera instruction books.
In way AF is a bit like spot metering - it occasionally needs some user input as to what it is aimed at to get a good result.
The chances of a new D3 and a new D700 both having defective AF are extremely low so maybe you need a little practice at learning how AF target selection has changed with the D300/3/300 in much the same way as D2x/D200 had to learn when changing from a D1/100.
Photography is a bit like archery. A technically better camera, lens or arrow may not hit the target as often as it could if the photographer or archer does not practice enough.
#7. "RE: Yes but" | In response to Reply # 5Fri 08-Aug-08 08:20 AM
>Here're my sample photos:
- without your showing where the focus point was aimed (a tools option in NX) it is imossible to know whether the AF was aimed at a good AF target - or one where Nikon (in your camera instructions) say AF can produce unreliable results.
Sorry but without more information it is impossible to distinguish between whether the results are poor AF target selection or a lens/camera issue.
#9. "RE: Yes but" | In response to Reply # 7Fri 08-Aug-08 12:41 PM
I don't believe this is a problem with focus target. Below are the test targets I used in the camera shop (all with the 24-70mm/2.8). I took some screen shot from NX at 100% to show the focus point.
D700 center focus:
D700 right focus:
D300 center focus:
D300 right focus:
When the focus point is set to right, the result is not acceptable in my opinion. I believe my test target is as good as it gets...
Can someone with the D300/24-70 or D700/24-70 please do a similar test and report perfect focus result? At least in that case I know it is not a design issue and I have some hope to ask Nikon fix the problem.
I sent the lens once. Their "repair" did not really fix the problem. Now I am thinking sending both camera and lens to Nikon. But I really don't like the idea of out of camera for 2-3 weeks in summer.
#10. "RE: Yes but" | In response to Reply # 9Fri 08-Aug-08 01:40 PM
My D700 test shows exactly that. I used the focus point in Capture NX 2 to show where the focus point to make sure I was using the AF point correctly and then crop in at 100%. I focused at the same point whether it is center AF or right or left AF point and at Single, Continus Servo. I didn't try M servo or manual focus at all. So in the same condition, tripod, lighting, and etc. Center AF point shows an amazing sharp picture, but at left or right focus AF, I get blurring picture at the same distance and everything. I'm not good at Capture NX 2, so I'm not sure how to save the the AF point shown in Capture NX 2 into a sample image. But if anyone know how to, please let me know and I will update the image again.
I'm bringing it in to a Nikon service center today or Monday, so I'll post an update here when I get a response back from Nikon. I'm bringing all my lens and the D700 in. But in center focus, I must say that D700 and the 24-70 is an amazing combo. I was in a restaurant last night testing this again, center focus picture even at ISO 6400 shows barely any noise at no NR at all. But of course anything to the left or right is not sharp. I love not having to use a flash in a restaurant now, the days that everyone looking at me with my flash going off on the SB600 on the D200 because they were probably annoyed by my flash are gone!!!
#11. "RE: Yes but" | In response to Reply # 10JDMils Registered since 23rd Jul 2008Mon 11-Aug-08 01:00 AM | edited Mon 11-Aug-08 01:08 AM by JDMils
From a complete NOOB, I could not see what focal length your pics were taken at, so i will assume they were taken at 24mm.
Could it be that at such a large aperture and small focal length that the distance from the subject to the outer rim of the lens is so much more than to the centre of the lens such that the DOF has something to do with this issue? Thus, this would account for why this issue is not apparent on lens' with smaller apertures.
I suppose I'm assuming that the left & right AF points are only valid when used with specific focal/aperture spec'ed lens.
PS. ATempest- use PrintScreen to capture the current monitor image to the clipboard or ALT-PrintScreen to capture the current active window to the clipboard. Now you need to paste the clipboard to an image app to save as a file and you will have a picture of the AF point from NX.
+-- AF-S DX NIKKOR 18-55mm f/3.5-5.6G VR
+-- AF-S NIKKOR 50mm f/1.4G
+-- HOYA 52mm Clear PRO1 DIGITAL Protector Filter DMC LPF (on the 18-55)
+-- HOYA 58mm Clear SKYLIGHT Protector Filter (on the 50)
+-- Navman forums at http://firstname.lastname@example.org
#12. "RE: Yes but" | In response to Reply # 11Mon 11-Aug-08 03:24 AM
My photos were taken at 70mm. But I don't think that matters in this test.
Let's examine theset two focusing issues:
1. Focus your lens at the center of your view finder at f/2.8, and examine the edge sharpness of the resulted photo.
2. Focus your lens at the edge of your view finder at f/2.8, and examine the edge (where you focus on) sharpness of the resulted photo.
In the first case, I would expect some degree of softness on the edge, although these $1500+ lenses are supposedly to produce edge to edge sharpness if the subject is completely flat. There are enough reports for 24-70mm/f2.8 for its field curvature and not able to produce edge-to-edge sharpness at 24mm. Although I am disappointed at this, I think I can live with it in real world usage. DOF certainly is at play here.
In the second case, if the result is not perfect, practically we are saying: this lens limits my creativity in that I can never make off-center part of my image be perfectly sharp. If I take a portrait of a girl, and focus on her eyes/hair. Her eyes/hair will be soft in my photo! That is quite a fatal defect in my opinion. DOF should not be a problem here. I have got great results at f/1.4 with other lenses.
I guess I will have to send both the lens and camera back to Nikon to try my luck... even though no other lenses I own has such serious problem.
#13. "RE: Yes but" | In response to Reply # 11Mon 11-Aug-08 05:48 PM
I was hoping there was a way for the Capture NX 2 to save the file as is instead of of the Alt-PrintScrn method. This is because the camera meta data will not be saved. Oh well, but back to your question similar to Steve's mine pictures were taken at around 50 or 60mm and the apeture is wide open at 2.8. If I set the apeture to 5.6 or more like toward the f/8 stop, then I would get sharp result. But that completely defeated the purpose of getting a 2.8 lens. Anyhow, my lens and camera is at the Nikon service center, I dropped it off on Friday. I'll update everyone here on their response. I'm thinking this is more of a lens problem than the camera.
#18. "RE: Yes but (to Len)" | In response to Reply # 13Tue 19-Aug-08 10:47 PM
Got my camera and lenses back and Nikon checked the AF system and the lens and all they did was cleaning really. It seems the problem is still there. Maybe Len is correct, the outer AF will only focus vertically which by the way is not documented in either my D200 or the new D700 manual unless I'm not seeing something in the manual. But I'll send this question to Nikon's online support to confirm.
But one thing that still bothers me is why is F5.6 and smaller opening have a better result than a wide open F2.8 or even 1.4. Since the problem is more notable at 50% or more cropped, I guess I can live with this.
#19. "RE: Yes but (to Len)" | In response to Reply # 18Wed 20-Aug-08 09:53 AM
I sent my camera and lens to Nikon and am waiting for them. I fear this is a design problem they have, and not fixable. I am now fully expecting a disappointing result.
I don't believe this is a focus taget problem. I have seen good results with other lenses such as 17-55mm/2.8 or 85mm/1.8 wide opened on the same focus targets.
I hate to say this, but I believe: 24-70mm/2.8 does not auto focus well with left/right focus point on D3/D300/D700 at f/2.8. Period! I would love to see it if someone proves me wrong with a perfect test result with _any_ test target. I wonder if I can get full refund from Nikon (I know it is wishful thinking.) This expensive lens is just completely not useful for my application.
#20. "RE: Yes but (to Len)" | In response to Reply # 19Wed 20-Aug-08 12:31 PM
I also did the focus test chart test, no help. My 24-70 was perfectly focused on center. I'll have to test out other lens like the 50mm some more and see what the result is. I also did the test that Len suggested regarding putting the camera in portrait mode, same result. I also then put it to the 51 AF points and test it point by point from left to right, same thing, as soon as it leaves two AF point off the center or so, sharpness goes away.
#22. "RE: Yes but (to Len)" | In response to Reply # 19Wed 20-Aug-08 02:18 PM
Did some more testing this morning with the 50mm set at 2.8 to try to replicate the 24-70. I believe you are correct, the problem seems to be more visible with the 24-70. And it's more visible if you use the outer AF point. In other words, I have mine set up as 11 af point versus the 51, so I can have a center, mid right or left, and a far out right or left AF point. If I use the mid right or mid left, it seems to be better than the far right.
I'll be doing some more real life shooting with the lens in the weekend to make sure this problem does not affect my pictures at least from 50% crop or less. If it does, it's going back to Nikon again. If this is a design issue, I would hope Nikon would at least tell me about it. Nikon also noted that they did test images, so I've requested those images from them. I just want to know if this is the way it is then I would just avoid using the outer AF point once and for all.
Also, you can see this sharpness issue in their picture gallery for the D700 with 24-70 with a lady inside a church or something, this is posted in their website where you can download the original image and check the AF point and then crop in to check sharpness.
#14. "RE: Yes but" | In response to Reply # 7
I have read quite a few of your posts and I always seem to agree with the logic in your replies including this one. However I am bored tonight with nothing much else to do so I set up a curd test since I have this combo i.e D300/24-70 2.8 to see for myself.
I now see what the OP was talking about and can not explain the results I found. What do you think is going on here? Below are a center, left and right images all shot from a tripod, all I did is move the focus point for the three images. It clearly shows the right & left images out of focus. I also noted that when the focus point was at the right or left the center of the image was more in focus then where the focus point really was.
Attachment#1 ( file)
Attachment#2 (jpg file)
Attachment#3 (jpg file)
Attachment#4 (jpg file)
#16. "RE: 24-70 AF "issue" possibly explained"" | In response to Reply # 14Mon 18-Aug-08 05:04 AM
> What do you think is going on here?
I have just recreated your test target using the 24-70, D3 and SB 800 as illumination source to eliminate camera shake, but hand holding.
At 24mm where the AF sensor covered several more lines of my test target than yours AF was accurate with centre, left and right sensors indicating no AF issue at 24mm.
At 70mm with the camera in landscape mode relative to the lines I got similar line coverage to yours with centre AF good and either left or right out of focus - i.e. like your result.
As soon as I switched the camera through 90 degrees (portrait mode relative to the lines of print) AF was good centre, left and right.
As you probably know the centre AF points are cross type and read horizontal and vertical derail.
However the outer AF points read VERTICAL detail - despite the AF point implying they read horizontal detail.
My assumption as my camera and lens worked well at 70mm ONLY when in portrait mode is there is enough detail for AF to read and focus accurate with the camera in portrait mode (with the sensor reading along the lines of lettering) but not in landscape mode where the AF line is reading through vertical lines of lettering with spaces between.
Can you re test to see if your problem is resolved switching the camera 90 degree relative to your target?
If it is it is
1/ A D3/300/700 AF limitation with your type of AF target
2/ Even so it is misleading of Nikon to indicate in the AF mark the outer 36 AF points read horizontal detail when they actually read vertical detail.
If it is not resolved it could be either lens or camera so I would talk to Nikon.
#17. "RE: 24-70 AF "issue" possibly explained"" | In response to Reply # 16Tue 19-Aug-08 04:06 AM
Thanks Len for taking the time to investigate. While I was moving the focus point left & right I was paying attention to how the outer point would line up with a vertical line. I thought if I was able to center a letter within the single focus point all would be good. I will try here tomorrow to shoot it portrait style and see what happens. I don't get too strung out on test like this and take them with a grain of salt most of the time. I shot with the 24-70 sat & sun and the results were stunning. Even tried it for sports next to the 70-200, they are an awesome match.
#24. "RE: A happy ending" | In response to Reply # 17Tue 26-Aug-08 06:20 PM
>I shot with the 24-70 sat & sun and the results were stunning.
This implies, once you get used to when to carefully select an appropriate AF target, your 24-70 is as stunning as mine.
When you do the re test make sure the lens is out of focus each time you move sensors. The reason is AF needs to detect a minimum amount of out of focus difference before refocusing if it needs to because you have slightly changed the camera framing to locate a better AF target..