Nikkor 58mm f/1.4

FreezeAction

This image available from Nikon has me leaning for the 58 f1.4. After downloading I immediately saved it as a tif file. 58mm_F14G_03_04_2238_WEB. The file can be found on the 58mm 1.4 information page. I carefully interpolated it to a 24x36" poster size at 360ppi and made a print to evaluate for my self. You can count eye lashes and brows. The transition from sharp to out of focus is as good as I've ever seen. The only real question I have about the lens now is how fast will it lock and hold focus in AF-C mode? It is worth the cost of a rental to evaluate and see. A full body shot sharp in motion as a moving ballet dancer lens? Just maybe.

Golf007sd

Here is another one for comparison. I'm truly pleased with the performance of this lens.



D800 @ f/2.0

Larger size of image

Lastly, here is a review worth reading by photographer Sam Hurd.

shawnino

Am I the only person (@PitchBlack maybe?) who is going to make the purchase decision on this lens almost solely on what I can do with it wide-bleeping-open?

I have a 50 1.4D. At f/2, f/4. f/8, f/11, f/14, it's OK.
Is this lens better than my 50 at all those apertures? Very likely. But my 50's OK for my needs and doesn't require a net $1500 upgrade..
Where my 50 displeases me is at f/1.4-1.8. And that's where I'll be looking for performance that justifies the expense.

FreezeAction

Here is another one for comparison. I'm truly pleased with the performance of this lens.



D800 @ f/2.0

Larger size of image

Lastly, here is a review worth reading by photographer Sam Hurd.

I also enlarged and printed this one. Same results. I can see using this lens wide open and stopped down for whole body action photos when more dof is desired.

NSXTypeR

DPReview's review of the 58mm is out.

I wonder if it changes anyone's opinion on it.

http://www.dpreview.com/lensreviews/nikon-58mm-f1-4g

It certainly didn't change mine because I sure as hell can't afford it.

donaldejose

Yes, the dpreview article does make a difference in my opinion of the lens. If I had one I would shoot it at f2 to f4 so that I get both the center sharpness and the edge bokeh. I think that will be the "sweet spot" in shooting the lens and I did happen to notice when Nikon announced it with sample images many of those images were shot at f2 rather than at f1.4. Shooting at f1.4 is fine if you are willing to accept center softness in your image which isn't so bad and what you may want in many images. Like any tool it has strengths and weaknesses and one needs to know them so you can use the lens to take advantage of the strengths and avoid the weaknesses. I never wanted this lens before; now I do.

WestEndBoy

The sharpness graph on page 4 (Test Rusults FX) of the Dpreview article is really interesting.

Slide the aperture to F 2.8. It is sharpest in the very centre but still very sharp in the centre third and slowly declining until the edge of the centre two-thirds. The outer one third then flattens out.

Now imagine that you are shooting a portrait of a person at 2.0 or 2.8 and assume, like Dpreview suggests, that this sharpness is due to field curvature which is almost certain to be the case. The centre of the frame, assuming a person’s face, will be in perfect focus while the background on the outer one third will be more out of focus than usual.

Would that not accentuate the “Bokeh effect”?

This seems to apply at F 2.0 as well.

It is obvious what the intended use of this lens is. Shooting portraits and stars. Everything about this lens, shouts that. It is good for shooting stars because there is little or no coma and the field curvature does not seem to be a problem when the lens is focused at or new infinity (See the “Sharpness Compared” section on page 5). It is good for shooting portraits because everything about the lens seems designed for that plus, THE ONE THING THAT THE PIXEL COUNTING SHARPNESS ENTHUSIASTS WILL PICK ON actually seems to be a likely benefit when shooting portraits where you want the centre sharp and edges out of focus and fuzzy. Even the fact that the lens is soft in the centre at 1.4 is a possible benefit, as people often prefer a slightly soft look to obscure blemishes.

Nikon is Brilliant! The first time I shot this lens I shot a portrait of my wife on a DF, and she looked at the LCD screen and said, “Wow!”. She really “popped” and the colours looked fantastic.

If you are shooting flat surfaces (architecture and some macro) and you want perfect sharpness everywhere, buy a PC or macro lens. If you are shooting portraits, stars (and landscapes at 58mm as they are close to infinity and the field curvature won’t be an issue plus you are probably shooting at F8 on a tripod) and other real world subjects that tend to be in the centre of the frame surrounded by a background, but this lens. I have just committed to adding this to my list of lens to purchase (right after the 200mm F4 Macro before Nikon discontinues it and replaces it with an inferior lens).

If you want a “general all round” lens, buy the 50mm 1.4G, which I own and like (just wish there was a professional grade 50mm).

For more on field curvature, read:
http://tashley1.zenfolio.com/blog/2013/5/field-curvature---a-practical-guide

A well known to us contributor to this article said:
Now the truth is that the majority of lenses these days exhibit very mild field curvature, if any. But there are certainly a few that do and those few tend to be superb wide-aperture prime lenses. If you have one or two of those, taking a few minutes to evaluate the pattern of its field curvature (and I still prefer the term ‘area of best focus’) can improve your photography.
© Roger Cicala 2013

"SUPERB wide-aperture prime lenses." Does anybody here know of any?

Jeff

Paperman

Eyelashes at f1.4 ! ! @ 1:1 magnification ...Will we ever see that much mag in any print/magazine ?

The single hair of an eyelash maybe no more than a few pixels wide but the AF system's AF point maybe covering/reading from probably around 50-100 pixels at that distance ???

The DOF should also be taken account in such magnification. It is 20mm DOF at 1 m, 100 mm at 2 m but can't figure how to calculate when you magnify this much ( take the distance as what it looks magnified - 10 cm or take focal length as 400-500mm ? Take a different CoC ? ) . May work out to be not more than the gap between hairs :-)

Just some thoughts rushing into my mind .... It is true that the lens is overpriced but isn't the expectation by many too high for any lens ?

Paperman

Why would it matter as the center will be out of focus/outside DOF range anyway?

Maybe you are trying to say something else ? :-?

Great shot btw

WestEndBoy

Yes, great shot. I am sure you can do more with a fuzzy lens than I can with a sharp.

You bring up a great point, but I have the same question as Paperman.

However, if I read the Dpreview review correctly, you still might be soft on the edge where your subject is. I suppose that I might try pulling back a bit as 58 is wide for a portrait lens. Then crop the right. But then the extra distance would cost us in sharpness to.

I am looking forward to finding out how you deal with that.

WestEndBoy

@Paperman. Hmm... I may be confused because I'm not a tech guy, but my understanding is that the softness away from the center is caused by field curvature, right? But if you place the point of focus away from the center on one of the outer focus points, does that outer point suddenly become sharp and the field curvature causes the center to be soft? I literally have no idea.

And I don't think I'm being too demanding for a $1700. The shot above was shot with Nikon's 85/1.4 and it's very sharp at f1.4.

I don't think you are being too demanding either. It will be interesting to compare the 58 to the 85. Both have no coma and great bokeh. I wonder if the colours pop a little more on the 58. If Nikon had to give up on wide open sharpness to achieve that, then maybe it was a decent tradeoff. Otherwise, couldn't they have made an 85 1.4 in 58?

According to Dpreview, the softness in the centre is caused by field curvature. This means that if you place the point of focus on the edge, it will become sharp and the centre will become softer.

I think the important question to ask and not answered by the Dpreview review is "how sharp will it be on the edge if I focus on the edge". If field curvature is the issue, then it should be sharp. If it is entirely something else, then it will not be sharp at all. If field curvature is part of the issue, then it will be sharper but not as sharp as the centre.

For insights into the other possible issues, the link I posted earlier will provide insights into what the other issues might be and suggests an approach for testing them:

http://tashley1.zenfolio.com/blog/2013/5/field-curvature---a-practical-guide

Pitchblack, your full photo shoot, especially if you do it after reading the above article and throw in some shots suggested by the article, will be very telling and interesting.

Jeff

Paperman

I am stepping into unknown territory here but I guess concepts are being mixed . Field of curvature originates from the curved shape of glass and it should have nothing to do with DOF not being enough to cover center and corners at the same time ( due to being at different distances )

Field Curvature, also known as “curvature of field” or “Petzval field curvature”, is a common optical problem that causes a flat object to appear sharp only in a certain part(s) of the frame, instead of being uniformly sharp across the frame. This happens due to the curved nature of optical elements, which project the image in a curved manner, rather than flat. And since all digital camera sensors are flat, they cannot capture the entire image in perfect focus, as shown in the below illustration ..... The effect of field curvature can be very pronounced, especially with older lenses.Read more: http://photographylife.com/what-is-field-curvature#ixzz2p5tHsTka:

It is said to change from lens to lens - suggesting it is not simply a matter of distance/DOF.

Did not fully read the Tim Ashley article, which seems to be saying the same thing as Jshickele.

Maybe they are both the same - one causing the other but my mind is too blurred at the moment to take anything in :-) with a few errands in mind to go out and do on this Dec 31st ... Would love to hear opinions from those with clear minds...

donaldejose

PitchBlack: If you focus on a point near the edge of the frame I would think you would have a "donut of sharpness" with a soft center; which I think is also what you suspect. However, that assumes you are photographing a brick wall. In the image you posted every part of the "donut" (except the models shoulder) is so far out of the DOF zone of sharpness that it should all be soft anyway. In certain compositions it might be a problem and cause some sharp object elsewhere in the "donut of sharp focus" to draw your eye away for your intended focal point.

WestEndBoy

I agree with Paperman and Donaldejose.

Paperman, one point, DOF is not related to field curvature, but a deep depth of field will mask it. Technically speaking, it will mask all of the field curvature if the depth of field is larger than the field curvature (not entirely accurate, but good enough for less than 500 words and no drawings).

Practically, this means that three things will mask field curvature:
1.
Stop down as this increases the depth of field.
2.
Shorten the focal length, as depth of field becomes deeper. Wide angles, if uncorrected, have massive field curvature, but the massive depth of field offsets and makes it easier for lens designers to deal with field curvature as they don't need to totally eliminate it. As an aside, the 14-24 2.8 is a legend in the making because Nikon's lens designers have nearly eliminated field curvature and made this lens sharp in the corners.
3.
For a subject that does not occupy all of the frame, focus on the subject.

Pitchblack, the 85 is my go to lens too. I love it. The 58 is on my list because many spaces are too tight for an 85 and the 85 is too long for street photography. If I have enough space, out comes the 85.

Paperman

Paperman, one point, DOF is not related to field curvature

I did not imply that ( or at least I tried ); I am trying to say the opposite. But I really don't know if there is any short way of explaining what I think.

Very briefly, I will say what I understand from Field Curvature is that it is a lens fault - changes from lens to lens and also distance to distance ( see link in prev post ). Stopping down corrects it but not because of the deeper DOF that comes with it but because transmitted light is limited to the centre of the lens. The linked article mentions even stopping down to diffraction level to correct Field of Curvature, suggesting it is not the DOF curing it ( just a stop or 2 would have corrected if it were distance related - which I will get into below )

The term field of curvature is also used to express the case where one focuses in centre and corners are therefore a bit out of focus because the distance to corner is longer ( = hypotenuse ) . Since this is directly a matter of distance, it is related to DOF and can be corrected by stopping down - simply because smaller apertures provide enough DOF to cover both the shortest & longest distances in frame ( like the diagram in Tim Ashley article )

So either this OR I am really lost .... Field of Curvature is just "one" and I am unable to see the link between the two .... :-? :-?

WestEndBoy

No Paperman. Nothing that we said until that post contradicts each other, though the way I worded my comment above may sound like I think we do. Sorry if I caused a misunderstanding.

I would like to comment on your recent post though. You said "Stopping down corrects it but not because of the deeper DOF that comes with it but because transmitted light is limited to the centre of the lens."

We are getting into optics theory here that goes beyond all the articles. I am not 100% sure, but after thinking about it, my gut says that this explains other aberrations such as vigneting and spherical aberration, but not field curvature. Field curvature, uncorrected, is independent of aperture. Stopping down increases depth of field, which masks it. However, I am not 100% sure.

If you think about your second to last paragraph, "The term.......Ashley article)", I agree with that and the articles do speak to that.

And look on the bright side, we both agree and understand enough to advise Pitchblack what his likely challenges for testing the 58mm 1.4G will be and what to watch for. Namely, shoot a brick wall and fool around with the focus to see what is focusing and what is not combined with practically shooting it. If he can put a subject on the side, focus on it, and it comes out sharp, then he likely doesn't have to worry about anything.

One thing is for sure, I would love to have a heart to heart with the lens designer. I am sure that would be very interesting to learn what his tradeoffs were.

Paperman

Well, I am 90% OK about the subject ; better not bore the rest trying to figure out the remaining 10% !

BUT ....

Can anyone please explain me how we can see 2750 lw/ph ( 5500 l/ph ) on a sensor that is 4912 pixels high ? ( Dpreview 58mm f1.4 test ( @f2.8 ) ?

Thought I had the "lines/pic height" subject figured out long time ago but this somehow beats me ... What am I missing ?

FreezeAction

@FreezeAction: I had a look at that image to make sure that it wasn't me missing the focus. Obviously if Nikon posts it on their site, it's going to be as near technically perfect as they can get it.

First of all, the image to which you are referring was not shot wide open. I checked the EXIF and t was dropped down a full stop to f2.0. Second, while you can see every eyelash, they aren't crisp. There is blurring along the edges of the lashes and there's a distinct loss of contrast. . Third, the eye looks rather hazy instead of sharp and there's very little detail showing in the iris. Same framing and aperture the Nikkor 85/1.4 is dead sharp. The lens has nice qualities, but

As far as speed of focusing. It was quite fast and accurate for an f1.4 lens. The AF is one of the lens' strong points.

This is the image FreezeAction was talking about:

You are right about the eye lashes not being crisp at 40x60 enlargement. Actually they look like newly hatched wooly worms. The eye brows look very good though. I went back to the original image, saved as a tif, resized to 360ppi without any interpolation taking place (the image size was reduced), and then used several different methods of interpolation to enlarge to 40x60. This is not a good image to enlarge. The model has split ends at that size and as mentioned the eye lashes are not crisp but would you really expect to see that if the printed image was displayed in a clothing store hanging over a display? The three strong points I was looking for are as mentioned a fast accurate AF, and according to DXOMark.com CA is absent (2), and great color. I wish Nikon had given us a fall color landscape to examine taken at an fstop suitable for a deep depth of field. Put the whole image in apparent focus.

I copied a 10x14 area of the 40x60 of the in focus eye. I see as good as I ever have of the transition to OOF moving over to the other eye. Next test will be mine with a beach landscape on a sunny day at ISO 100 @ f5.6-f8 to see if it will indeed do 40x60 landscape prints. Focus will be done with live view as Nikon suggests for absolute sharpness. After reviewing Sam Hurd's review there may be a place for a Sigma 50mm f1.4 for indoor dance studio photographs for whole body shots. I'd rather not go wide and put a few extra pounds on subjects. People get picky about that.....The other option if space permits is to use the Nikon 85 1.4. It's the need for whole body photographs of ballet dancers and perhaps the 85 from farther back would out perform the 58 f1.4 indoors. I normally don't shoot wide open unless poor lighting in an arena with action requires it. One other option for indoor dance students will be the http://www.bhphotovideo.com/c/product/845350-REG/Tamron_SP_24_70mm_f_2_8_DI.html I purchased just under the wire for this years taxes. After reading on the blog today that it was the highest rated zoom in that range for the D800 bodies I'm glad I bought it. I believe the 58 1.4 along with the N 851.4 will find a home in my bag. And if it doesn't cause a divorce on of these for what to me might just be the ultimate landscape lens: http://www.bhphotovideo.com/c/product/818358-REG/Schneider_06_1066457_PC_TS_Super_Angulon.html

While doing a weekend of testing at St. Augustine the Scheider will get a good workout along beside a N 58f1.4.

FreezeAction

@FreezeAction. I own the Sigma 50mm/1.4, and as I've stated before, I don't much care for it. I think it's a serviceable lens and I did my entire photoshoot with it last week (http://500px.com/MarkCrislip/stories/3860454/selva-en-la-selva)
compare with how luscious the 85/1.4 looks
(http://500px.com/MarkCrislip/stories/3862476/blacking-out-with-cami)

There are tons of shots with missed focus from the shoot I did with the Siggy 50. I blame the lens since I nailed 95% at 1.4 with my 85/1.4 with the Cami shoot, and the bokeh to me is quite rough and unattractive with the Siggy lens. It's 100% non quantifiable, but the pictures with the 58/1.4 are just prettier than the pictures with the Siggy 50/1.4. I don't have a full photoshoot to compare since I just got to use the lens for an hour, but to me there's no comparison, even though the 58 is softish.

I'll have photos from a full shoot with the 58 shortly.

Looking forward to them. Hope you have a few to compare the the 85 also.

On another note after thinking about it, the sharpness or lack there of on the eye lashes could be makeup residue. I had that happen once when I had a young lady pose for just an eye shot that I needed. She was wearing eye makeup and it ruined the chance for the needed photo for my own use. She had full grown wooly worms for eyelashes.

shawnino

Went back into the shop today. Did a careful re-test.
Bought the lens.
I pretty much agree with what everybody else said, here and elsewhere: will not win a contest for centre-sharpness. Maybe that's why it costs less than half what the Zeiss costs, I don't know. But sharpness isn't the only metric. Colours good, bokeh great, AF on D800E, when I need it, is certainly fast enough. (I have more D-glass than anything.) Looks like very little drama for coma/flare/vignetting/etc. in careful hands. $1700 all-in feels big, but I'm telling myself that's the new reality.

Mastering this lens, as I think I'll have a bit of a hill to climb, is now my winter project. If I fail miserably, I can sell it at a loss of a few hundred bucks and know I set myself a challenge and tried my level best. And start saving up for my next challenge (Otus 85, which remains vapourware for now, or the regular Zeiss 100/2).

Tomorrow I'm off to a snowswept beach. Not Antarctica, but it'll do.

Ade

Can anyone please explain me how we can see 2750 lw/ph ( 5500 l/ph ) on a sensor that is 4912 pixels high ? ( Dpreview 58mm f1.4 test ( @f2.8 ) ?

Thought I had the "lines/pic height" subject figured out long time ago but this somehow beats me ... What am I missing ?

Lens resolution is measured by analyzing a slightly slanted edge on an test chart, using an algorithm specified in ISO 12233 or similar.

Instead of examining the slanted edge one pixel at a time, the ISO 12233 algorithm has a clever "binning" technique which combines data from four pixels at a time. I.e., four pixels along the direction of the edge are taken together, followed by the next four, etc.

This method basically repeats ("oversamples") the edge measurement, by a factor of four. Oversampling is used to reduce aliasing effects (the jagged pixel effect on slanted lines). And, more directly to your question, this 4x oversampling is also the reason we can measure lens resolution above the Nyquist frequency of the camera's sensor (normally, 0.5 cycles per pixel max).

Paperman

Thanks for the info Ade ; did not have a clue about the standard used ...Dpreview just simply explains it as the physical count of lines visible ( on their test picture ) at a certain # of pixels width giving the % yield of sensor and then multiplying by total pixels ( height ) - like 18/26 x 4000 .. It is clearly not as simple as that then.

Ade

We don't count converging lines anymore. In fact on most modern charts, those lines are not even present.

For computerized measurement, all that's required is a slanted edge, e.g., from a black square or rectangle tilted around 5 degrees. So new charts will have various boxes or rectangles spread throughout the chart:


Ref: Imatest SFR Plus

The basic Imatest chart is shown above. This particular version has two features in addition to the slanted edges: a wavy pattern you see top center is a focus star (not used for resolution testing); and the series of patches bottom center is a step chart designed to measure dynamic range, noise, gamma, etc.

DPReview's 58mm f/1.4 test uses data provided by DxO. Below is an illustrative photograph from DxO's website, showing DxO's version of slanted edges:


Ref: DxO Testing Protocol

The 58mm's field curvature can be measured by first selecting an image where edges on the center square gives maximum sharpness. Then, using the same image, the measurement is repeated from the off-center and corner squares.

Paperman

They'd better update their website

http://www.dpreview.com/glossary/digital-imaging/resolution

http://www.dpreview.com/glossary/resources/images/cp8700_horizontal.jpg
http://www.dpreview.com/glossary/resources/images/Testchart_nikon_cp8700.jpg

Symphotic

I just picked up a book called "The Philosophy of the Nikon Df" published (in Japanese) by Asahi Camera. In the book they show two identical night sky photos, one with the 58 1.4G and one with the Noct taken with the Df. There is substantially more coma with the Noct.