-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
First 28mm f1.8G review up
- Thread starter harold1968
- Start date
Interesting. I agree, there is no focus shift visible in your test (the D700 has big pixels though).
Perhaps there is really copy variation. However, I would have thought that residual spherical aberration is a pretty fundamental characteristic of the lens design, not likely to vary much. - Joseph may be able to answer this.
Yes, I only have a D700 to hand (waiting for a D800). Also, DPReview has downsampled the image (a 100% crop) from 1400px tall to 1200px tall. Nevertheless, for me this test is sufficient to demonstrate that my copy of the lens exhibits no significant focus shift at the focus distance I want to use it at.
Please post the results of your test.
I have the D800 and just received an email from Amazon that my 28/1.8 should be here next Wednesday.
DuxX
Senior Member
Do you think it is possible that the Nikon made a lens which misses properly focused eye at the f5.6? Ahm sorry, I can't believe in that.
One specific lens can be faulty - of course.
Regards
--
Dusko Jovic / DuxX /
http://www.duxx-photography.com
I do hope it's only 1 or 2 specific lenese.
How good is your optical theory?
Every lens is optically a compromise. Nikons newest generation of 1.8G lenses are spherically under compensated. There are several reason they could be doing this, let me name 2. Firstly, they are cheaper and easier to produce. Secondly and more importantly under compensating for spherical aberrations produces a smoother and creamier bokeh. I assume that the second reason is the primary motivation. In the mean time i'm on my second 28 1.8G. It has like the first the typical signs of a spherically under compensated lens. Creamy bokeh, a tad soft, and focus shift. The second one seams better as far as focus shift is concerned. But i'll be testing that tomorrow. btw: the 35 1.8G is the opposite. Bitingly sharp, no discernible focus shift and a harsh bokeh.
Now that i am on my second 28 1.8G which is probably 2 more then most of those discussing this lens have: Yes i do believe Nikon ( btw I have a lot of nikon lens of several different generations) deliberately made this lens with these characteristics/compromises. I do think that production tolerances will have some slipping beyond what is OK as far as focus shift is concerned.
How good is your optical theory?
Every lens is optically a compromise. Nikons newest generation of 1.8G lenses are spherically under compensated. There are several reason they could be doing this, let me name 2. Firstly, they are cheaper and easier to produce. Secondly and more importantly under compensating for spherical aberrations produces a smoother and creamier bokeh. I assume that the second reason is the primary motivation. In the mean time i'm on my second 28 1.8G. It has like the first the typical signs of a spherically under compensated lens. Creamy bokeh, a tad soft, and focus shift. The second one seams better as far as focus shift is concerned. But i'll be testing that tomorrow. btw: the 35 1.8G is the opposite. Bitingly sharp, no discernible focus shift and a harsh bokeh.
Now that i am on my second 28 1.8G which is probably 2 more then most of those discussing this lens have: Yes i do believe Nikon ( btw I have a lot of nikon lens of several different generations) deliberately made this lens with these characteristics/compromises. I do think that production tolerances will have some slipping beyond what is OK as far as focus shift is concerned.
Someone please correct me if I'm wrong, but I believe your test is avoiding this issue.
It's all about AF. AF will use "phase detection", which will lock focus at the wide open aperture, only stopping down to user selected aperture with the shutter press.
Live View will use "contrast detection" side stepping the focus shift issue.
Manual focus will also side step the focus shift issue.
I'm very disappointed to hear of the very significant focus shift issue. I may cancel my pre-order and stick with my 35 f1,4G, while waiting for more tests. I was really hoping to switch them out and put some cash in my pocket, like with the amazing 85 1.8G.
It's all about AF. AF will use "phase detection", which will lock focus at the wide open aperture, only stopping down to user selected aperture with the shutter press.
Live View will use "contrast detection" side stepping the focus shift issue.
Manual focus will also side step the focus shift issue.
I'm very disappointed to hear of the very significant focus shift issue. I may cancel my pre-order and stick with my 35 f1,4G, while waiting for more tests. I was really hoping to switch them out and put some cash in my pocket, like with the amazing 85 1.8G.
@Kali108,
you are absolutely correct. Liveview will bypass the issue. I hadn't caught that in Elliotn's message. To see the issue one has to use phase detect focus.
B
Barry Fitzgerald
Guest
Your getting mixed up here. They're not under corrected for spherical aberrations the 35mm f1.8 has an Aspherical lens element.
That stops the softness and soft focus effect you can get on some lenses.
What they have is more distortion and under corrected CA
snooked123
Forum Enthusiast
Tried to do what photozone does not sure if that is the best way to go about it. In any case the focus shift if any is not worse than 24 1.4 or 35 1.4 (as shown on photozone.de).
Used phase detect focus wide open to focus on "28".
100% crops: Left f1.8, Right f4. What do you guys think?
Used phase detect focus wide open to focus on "28".
100% crops: Left f1.8, Right f4. What do you guys think?
You're wrong. I only focused once, at the beginning of the test. (I assumed that was a given, sorry for not spelling it out.)
The lens has no focus shift issue, as demonstrated by my test.
Barry,
I'm not saying they are NOT corrected. They are corrected but deliberately 'UNDER'-corrected. 'Under' or 'Over' correction is a choice made by the lens desgners to influence the characteristics. For instance: 'Under' gives you a creamier smoother bokeh.
BTW: the 28 1.8G also has an aspherical element. At least it says so right here on my copy
I'm not using Live View for each exposure (i.e. as I stop down the aperture). I'm using it once, at the start of the test, at f1.8, to ensure that I have accurately acquired focus on the desired focus park. This is a 'focus shift' test, not an autofocus test.
Hmm, another test that illustrates that there is no focus shift problem. Must be an error with your test procedure! I'm sure Goosel will put you right
You claim to have run tests on two 28/1.8 lenses, and both tests have shown that the lens suffers from focus shift.
You've also made the outlandish speculation that Nikon made a design decision to sacrifice focus accuracy at f4 and f5.6 for 'creamy bokeh'.
To have any credibility please post the results of your tests. Otherwise it is so much hot air.
The second one tests better then the first. The focus shift is there but the '0' point stays within the DOF range. This one is a bit softer at 1.8 with a touch of front focus.
So enough of using the lens calibration tool.
I then tried out a series of random snaps, close middle and far distances. about 1/2 of the images weren't sharp. I've never missed focus so often. Here a series in close range just to illustrate.
At the moment i don't know what to make of it.
Full size jpgs no shrapening except what std profile of cnx2 provides. First focus on drums, second focus on clock, third focus on metronome.
On the first pic the clock is sharper in focus then on the second.
So enough of using the lens calibration tool.
I then tried out a series of random snaps, close middle and far distances. about 1/2 of the images weren't sharp. I've never missed focus so often. Here a series in close range just to illustrate.
At the moment i don't know what to make of it.
Full size jpgs no shrapening except what std profile of cnx2 provides. First focus on drums, second focus on clock, third focus on metronome.
On the first pic the clock is sharper in focus then on the second.
Yes i am on my second 28 1.8G. No need to call me a liar.
btw: Zeiss has some interesting literature about deliberate decisions made during lens deisgn pertaining to compensation of spherical aberation.
Thanks for posting some images. Your clearly have a serious focus problem.
I just tried a similar test with some objects on my kitchen table. My D700 + 28/1.8 nails focus every time (tested at f4 and f5.6, using regular AF). No evidence of back focus.
I don't know why your D90 + 28/1.8 is not focusing accurately. But I am sure that Nikon has not designed the lens to perform in this way.
You're right to be concerned. I assume you have confidence in the D90's autofocus system?
Edit:
Just to be clear, you're saying that you get accurate focus when shooting wide open (f1.8), but when you stop down to f5.6 you get the severe back focus illustratred in these images?
Keyboard shortcuts
- f
- Forum