First 28mm f1.8G review up

harold1968

Veteran Member
Messages
4,604
Reaction score
434
Location
London, UK
I am guessing that the perceived focus issue is with the D800's known to be defective AF system and not with the lens. It was probably inappropriate for the reviewer to test this lens with a known to be problematic brand new camera. Ridiculous.
 
I am guessing that the perceived focus issue is with the D800's known to be defective AF system and not with the lens. It was probably inappropriate for the reviewer to test this lens with a known to be problematic brand new camera. Ridiculous.
Are we sure though? Focus shift isn't uncommon, but this lens appears to have more of an issue with it.

I'm not convinced about the need for f1.8 at 28mm myself either, and the price is a bit heavy too. I think I'll pass on this one.
 
I am guessing that the perceived focus issue is with the D800's known to be defective AF system and not with the lens. It was probably inappropriate for the reviewer to test this lens with a known to be problematic brand new camera. Ridiculous.
No, this shouldn't have anything to do with it.
 
I just double checked what the reviewer wrote about focus shift. I tested on a D90 instead of my D800 just to be sure.

Yes very strong focus shift to the back. :( :(
 
thanks for that

which aperture range do you see it over ?
+ how do you find corner to corner sharpness on f8 (on the D800) ?

thanks
I just double checked what the reviewer wrote about focus shift. I tested on a D90 instead of my D800 just to be sure.

Yes very strong focus shift to the back. :( :(
--
http://www.haroldmiller.me
 
Focus shift is a common feature in uncorrected lenses. Photographers learn to deal with it.

The "D800's known to be defective AF system" is actually a "known to be defective extreme left AF point only in a few bodies and not others, and can be fixed by Nikon if you send it back, and if not you can send it back again or get a replacement" AF system.

Lens focus shift has nothing to do with a body's AF system.

To be precise.
I am guessing that the perceived focus issue is with the D800's known to be defective AF system and not with the lens. It was probably inappropriate for the reviewer to test this lens with a known to be problematic brand new camera. Ridiculous.
--
------------------------------------------------------------------

"The job of the photographer is not to record indisputable fact but to try to be coherent about intuition and hope."
-- Robert Adams
 
thanks for that

which aperture range do you see it over ?
+ how do you find corner to corner sharpness on f8 (on the D800) ?

thanks

--
http://www.haroldmiller.me
I only tested on the D90 as my D800 is off to nikon because of the left af point issue.

The focus shift was clearly visible at F4 and strongest at F5.6. At F8 it seemed to be a bit better, but could be simply due to a deeper DOF. On the D90 corner sharpness is good already at F5.6, excellent a F8. Corner sharpness at F8 is better then my 16-35 at F8 on the D90.

harold,

did you check if your lens is decentered. Mine is ever so slightly decentered according to the tests they recommend at lensrental.
 
Focus shift is a common feature in uncorrected lenses. Photographers learn to deal with it.
Yes but on this lens the shift is big.
Yes, apparently. I would personally avoid the lens because I don't want to have to deal with it. Others may choose differently.

--
------------------------------------------------------------------

"The job of the photographer is not to record indisputable fact but to try to be coherent about intuition and hope."
-- Robert Adams
 
The focus shift is bad news indeed.
I rated the reviewer's ability as "poor" when I noticed part of the review appears to involve a long discredited 45 degree narrow black line target :(

--
Leonard Shepherd

99% of photo equipment is good or very good. The main thing to do with it is to learn to use it well.
 
Is focus shift taken care of by the AF electronics - it knows the focus distance and the aperture to be used for the exposure so can it compensate?

I've seen some sharp shots from this lens wide open - would these not suffer from focus shift it the AF system did not compensate for it?
Thanks for the link.

The focus shift is bad news indeed.
 
Is focus shift taken care of by the AF electronics - it knows the focus distance and the aperture to be used for the exposure so can it compensate?
No, the normal phase detect autofocus establishes the correct focus always when the lens is wide open. As far as I know there is no compensation for focus shifts with specific lenses, but you are right, this would be therotecially possible.

However, contrast detect autofocus (in Live View) focuses at the specific aperture selected, at least with the Nikon D800. This should thus compensate for any focus shift. My Canon 5D2 is actually different. Contrast detect AF is also always wide open, thus even here focus shift can be a problem. With my Canon EF 70-200/2.8II the only way to get critically sharp focus at f/4 is to focus manually... :-(
I've seen some sharp shots from this lens wide open - would these not suffer from focus shift it the AF system did not compensate for it?
No, if the lens is correctly calibrated shots wide open will focus correctly. The problem is that shots stopped down a little (f/4-f/5.6) will be misfocused. If stopped down beyond f/8 the misfocus will be hidden because of the deeper depth of field.
 
You might want to rethink this part...
  • This poses quite a challenge for people relying on normal AF-operation because AF grabs focus at fully opened aperture (f1.8) and does not compensate for any focus shift.
The AF "grabs focus" at either f2.8 (central AF zones, when an f2.8 or faster lens is used) or f5.6 (outer AF zones, or any lens slower than f2.8). It uses two pretty tightly focused portions of the exit pupil at +/- 5 degrees from perpendicular (f5.6) or +/- 10 degrees (f2.8).

It then does "compensate" for focus shift by applying the lens's default "focus tuning" to get a focus offset for wider apertures, and any correction the user has dialed in, so, depending on how Nikon set the default and how the user adjusted focus, you could be correct at f1.4 and off at 5.6, or off at 1.4 and correct at 5.6. The former is more likely, as Nikon calibrates fast lenses to be correct wide open.

In any case, there will be a focus shift between wide open and stopped down on pretty much every lens. Which brings us to...
  • I've never seen a similar behavior before.
That sort of blows your credibility as a reviewer, because it's a well-documented characteristic of many lenses, seen pretty strongly in all double Guass variations, such as the Nikon 50mm f1.4 AF-D.

--
Rahon Klavanian 1912-2008.

Armenian genocide survivor, amazing cook, scrabble master, and loving grandmother. You will be missed.

Ciao! Joseph

http://www.swissarmyfork.com
 
I only tested on the D90 as my D800 is off to nikon because of the left af point issue.

The focus shift was clearly visible at F4 and strongest at F5.6. At F8 it seemed to be a bit better, but could be simply due to a deeper DOF.
You should send your 28/1.8 back to Nikon too - it clearly suffers from the same fault as the lens Camera Labs reviewed. My 28/1.8 performs properly - i.e. it sharpens up as I stop down from 1.8, to 2 to 2.8. to 4 to 5.6. A lens that gets softer as it is stopped down is unusable.
 
The AF "grabs focus" at either f2.8 (central AF zones, when an f2.8 or faster lens is used) or f5.6 (outer AF zones, or any lens slower than f2.8). It uses two pretty tightly focused portions of the exit pupil at +/- 5 degrees from perpendicular (f5.6) or +/- 10 degrees (f2.8).

It then does "compensate" for focus shift by applying the lens's default "focus tuning" to get a focus offset for wider apertures, and any correction the user has dialed in, so, depending on how Nikon set the default and how the user adjusted focus, you could be correct at f1.4 and off at 5.6, or off at 1.4 and correct at 5.6. The former is more likely, as Nikon calibrates fast lenses to be correct wide open.
Thank you for the explanation.

So, fast lenses are usually calibrated for correct focus wide open and the ones with big focus shifts will then be off when shot at f/4 or f/5.6. Why can the built-in focus offset not change with different apertures?
In any case, there will be a focus shift between wide open and stopped down on pretty much every lens. Which brings us to...
  • I've never seen a similar behavior before.
That sort of blows your credibility as a reviewer, because it's a well-documented characteristic of many lenses, seen pretty strongly in all double Guass variations, such as the Nikon 50mm f1.4 AF-D.
I think what the reviewer meant is that the lens is worse in this regard than other lenses.

I'm starting to feel that this is a serious issue with fast prime lenses. Perhaps it makes indeed sense to use those almost always wide open (or stopped down to at least f/8). Slower (zoom) lenses may have the unexpected advantage of less focus shift, producing better results at f/2.8-f/5.6. BTW, my only Nikon lens with no detectable focus shift at all is the 60/2.8 macro.
 
My 28/1.8 performs properly - i.e. it sharpens up as I stop down from 1.8, to 2 to 2.8. to 4 to 5.6. A lens that gets softer as it is stopped down is unusable.
Are you sure your lens does not front focus a little bit wide open?

This would explain the "sharpening up".
 
The AF "grabs focus" at either f2.8 (central AF zones, when an f2.8 or faster lens is used) or f5.6 (outer AF zones, or any lens slower than f2.8). It uses two pretty tightly focused portions of the exit pupil at +/- 5 degrees from perpendicular (f5.6) or +/- 10 degrees (f2.8).

It then does "compensate" for focus shift by applying the lens's default "focus tuning" to get a focus offset for wider apertures, and any correction the user has dialed in, so, depending on how Nikon set the default and how the user adjusted focus, you could be correct at f1.4 and off at 5.6, or off at 1.4 and correct at 5.6. The former is more likely, as Nikon calibrates fast lenses to be correct wide open.
Thank you for the explanation.
You're welcome.
So, fast lenses are usually calibrated for correct focus wide open and the ones with big focus shifts will then be off when shot at f/4 or f/5.6. Why can the built-in focus offset not change with different apertures?
I've often wondered that, myself.

I have a theory: that they do it to make the camera less confusing. You can't compensate for the focus shift until you know the shooting aperture, so that means you'd be restricted to shooting in A and M mode, and I don't think Nikon could handle the support calls from people who can get sharp pictures when they put the camera in A (or "amateur") mode, but not in P (or "professional") mode.
In any case, there will be a focus shift between wide open and stopped down on pretty much every lens. Which brings us to...
  • I've never seen a similar behavior before.
That sort of blows your credibility as a reviewer, because it's a well-documented characteristic of many lenses, seen pretty strongly in all double Guass variations, such as the Nikon 50mm f1.4 AF-D.
I think what the reviewer meant is that the lens is worse in this regard than other lenses.
I have trouble picturing any lens from the 2010s being worse than a 50mm f1.4 from the 1950s.
I'm starting to feel that this is a serious issue with fast prime lenses.
It is. Has been for years.
Perhaps it makes indeed sense to use those almost always wide open (or stopped down to at least f/8). Slower (zoom) lenses may have the unexpected advantage of less focus shift, producing better results at f/2.8-f/5.6. BTW, my only Nikon lens with no detectable focus shift at all is the 60/2.8 macro.
;)

--
Rahon Klavanian 1912-2008.

Armenian genocide survivor, amazing cook, scrabble master, and loving grandmother. You will be missed.

Ciao! Joseph

http://www.swissarmyfork.com
 
I only tested on the D90 as my D800 is off to nikon because of the left af point issue.

The focus shift was clearly visible at F4 and strongest at F5.6. At F8 it seemed to be a bit better, but could be simply due to a deeper DOF.
You should send your 28/1.8 back to Nikon too - it clearly suffers from the same fault as the lens Camera Labs reviewed. My 28/1.8 performs properly - i.e. it sharpens up as I stop down from 1.8, to 2 to 2.8. to 4 to 5.6. A lens that gets softer as it is stopped down is unusable.
it deosn't 'soften up'. The plane of focus shifts backwards. btw: a typical symptom of uncorrected spherical aberrations. In this case i have the suspicion it's deliberate then another 'symptom' is a very smooth creamy bokeh, which the 28 does have.
 

Keyboard shortcuts

Back
Top