Saturday, 30 August 2014

Wide angle prime lenses compared

The classic 28mm prime lens is a natural part of any lens lineup. Most kit zoom lenses include the 28mm equivalent in the wide end, but you can also get prime (non-zoom) lenses with the same field of view.

Below are three such lenses, for three different interchangeable lens systems. How do they compare?


From left to right: Sigma 19mm f/2.8 (old style) mounted to a Sony NEX-3N, Lumix G 14mm f/2.5 mounted to a Lumix GM1, and Nikon 1 10mm f/2.8 mounted to a Nikon 1 V3 with a user optional EVF.



LensSigma 19mm f/2.8Lumix G 14mm f/2.5Nikon 1 10mm f/2.8
AnnouncedJan 10th, 2012Sept 21st, 2010Sept 21st, 2011
System crop factor1.522.7
Equivalent focal length29mm28mm27mm
Maximum aperturef/2.8f/2.5f/2.8
Equivalent max aperture, in terms of DoFf/4.2f/5f/7.6
Weight140g55g77g
Diameter61mm56mm56mm
Length46mm21mm22mm
Filter thread46mm46mm40.5mm
Minimum focus0.20m0.18m0.20m
Lens elements/groups8/66/56/5
Hood includedYesNoNo
Focus ringYesYesNo




The lenses are laid out here:


Sigma 19mm f/2.8


I have the first version of the lens, the "EX DN". It has a nice plastic finish, with a ribbed base, good for mounting the lens. Also, the focus ring is ribbed for a good grip. The second version of the lens is called simply "DN", and is part of their Art-line of lenses. It has a smooth metal finish on both the base of the lens and the focus ring, making it harder to grip it properly, in my opinion.

The Sigma lens is a generic lens, made to be used on a number of mirrorless systems. I have tested the Micro Four Thirds version of the lens, and found it to be not quite as good as the Lumix G 20mm f/1.7.

This is not a pancake lens, in fact, it is rather large. It is often liked for the good image quality and the low price point.

Sony also has a similar lens, the Sony 16mm f/2.8 pancake. However, it is often perceived as being somewhat dull optically.

Lumix G 14mm f/2.5


This lens was originally launched as the kit lens for the ultra compact Lumix GF3 camera. It is still the smallest Micro Four Thirds lens, excluding the body cap "toy" lenses.

As this lens has often been sold cheaply on Ebay, coming from split GF3 kits, the perceived value of the lens has been small from time to time. However, I think this is a good lens, for the compact size. I made a comparison with the Lumix G 20mm pancake here.

In the picture above, I'm using a 46mm to 37mm step down ring as a simple hood.

Nikon 1 10mm f/2.8


This lens is different from the others in the sense that it doesn't have a focus ring. There is a ribbed ring around the front of the lens, but it is only for mounting the lens more easily. To be honest, I don't dislike this solution. I seldom use manual focus anyway, and the ribbed edge is great for getting a safe grip on the lens when handling it.

In the picture above, I'm using a 40.5mm to 37mm step down ring as a simple hood.

Here is a comparison picture of the rear side of the lenses:


The Sigma lens has an exit pupil which is recessed into the lens. This may look strange, but I am guessing it is for compatibility with the Samsung NX system, which has a much longer register distance than the Sony E-mount. So probably, the lens was designed to work with Samsung NX, even if they haven't launched a lens for Samsung NX yet.

The Nikon lens has the smallest exit pupil, by far. A large exit pupil is often seen as a mark of quality. We shall see if this matters in the comparisons in this article.

Sharpness


To evaluate the image quality, I have taken a series of images with all three lenses at various apertures. All the images were taken on a tripod, at base ISO, and with a shutter delay to avoid camera shake.


All the lenses do very well in the image centre, even wide open. So I don't see much point in going into details there.

To better find the differences between the lenses, I am looking a the performance in the top left corner, which tends to be more challenging to render for a lens.

Looking at 100% crops makes it easier to compare the image quality (click for a larger image):


What we see here, is that the Lumix G 14mm f/2.5 has a very good performance in the corner, even wide open.

The Nikon 10mm lens improves when stopping down, and achieves a very respectable performance at f/5.6.

The Sigma lens is the least good of the three, and is more dull even at f/5.6. This is consistent with my test of the Micro Four Thirds version of the lens, in which it did not impress.

Bokeh


In the top table, I quote the "equivalent max aperture, in terms of depth of focus". This is the max aperture multiplied with the crop factor. The smaller the number, the more selective focus the lens/camera system is capable of.

As the Sony system has the largest sensor of the three formats, the Sigma lens, when used on a Sony E-mount camera, can take pictures with the most selective focus. This means that the background can be the most blurred when you are using the largest aperture.

Here is an overview of the relative sensor sizes (from Wikipedia):


The Sony camera uses the APS-C sensor size, and the Lumix camera uses the Four Thirds size. Finally, the Nikon 1 camera uses the 1'' sensor size. The 1'' sensor size is also used by the popular premium Sony RX100 compact camera series.

Here is a series of images, taken with a similar focus distance of about 0.3m, which illustrates the differences in depth of focus and bokeh:


Looking at 100% crops, we see that it is in fact the Nikon lens which has the nicest looking bokeh, even if it has the least selective focus.


And another example taken in daylight with the same short focus distance (click to enlarge):


From the 100% crops, we see clearly that the Sony lens is the most capable of achieving a blurred background:


The conclusion here is that if bokeh is important for you, then the Nikon 1 system may not be your system. If you want to achieve the most selective focus, then go for a large sensor system, like the full frame Sony A7.

From my point of view, I think the smaller sensor of the Nikon 1 system makes good sense. I get more depth of focus, which is, in my use, often a good thing.

Geometric distortion correction


Most mirrorless lenses feature some in camera distortion correction. To examine the geometric distortion characteristics, I have photographed a square tiled wall, and then overlaid the out of camera JPEG (in black) with the uncorrected image (in red). I used the third party RAW converter software UFraw to assess the uncorrected image.

Sigma 19mm f/2.8 (0%)Lumix G 14mm f/2.5 (-16%)Nikon 1 10mm f/2.8 (-4%)

The percentage in brackets is the relative distortion correction applied in The Gimp image processing software to get a rectilinear image. This is a way to compare the relative distortion between the lenses.

The Sigma 19mm f/2.8 does not use any in-camera distortion correction at all. However, it is not completely rectilinear, and you can see some barrel distortion.

The Lumix G 14mm f/2.5 requires the most in-camera distortion correction of the lenses. And at a short focus distance, you often get some residual barrel distortion. At infinity focus, this is not an issue.

Last, the Nikon 1 10mm f/2.8 needs some minor geometric distortion correction. However, after the adjustment, I find that the images are very nicely rectilinear, the most of the three lenses.

Conclusion


It is perhaps odd to compare these lenses, as they are for different systems, and very few will choose between them. But my conclusion is, again, that the Lumix G 14mm f/2.5 is a very under appreciated lens. It is generally seen as a "cheap and sub par" lens, but in my comparisons, it often comes out like a star performer. I like the lens a lot for the consistently good image quality, the fast autofocus, and the very small size.

The Sigma lens does not get good scores here. It has the least sharpness, and is the largest of the three. On the other hand, it is also the cheapest.

The Nikon lens performs very respectably. I think it has the best out of focus rendering of highlights (night time), and the sharpness is quite good. It is small, and focuses quickly. It also appears to have the least vignetting when wide open.

LensSigma 19mm f/2.8Lumix G 14mm f/2.5Nikon 1 10mm f/2.8
Pros
Can be used with larger sensor cameras. Cheap.Small and sharp lens.Good image quality.
Cons
Some barrel distortion. Not the sharpest lens.Some barrel distortion at close distances. Expensive.Little potential for selective focus, due to small image circle.



No comments:

Post a Comment