Nikon D5100 Review – Page 3
Welcome back to part 3 of this relaxed journey with the new Nikon D5100. Yesterday, I didn’t post anything as it was easter sunday, time with the kids and family. Weather was great and friends came for barbeque. Later, when all went to bed I started to get to downtown Vienna for some night shots with the D5100. Started around midnight and came home at 3am – I’ll share some images from that trip and some more information where the entry level class cameras have still some limitations for night photography, despite significant progress in sensors sensitivity. More on this later.
This morning, as promised to some commenters, I mounted the 2 kit lenses on the Nikon D5100, the AFS 18-55mm and the AFS 18-105mm VR while walking the dog with my wife.
Summing up this post I’ll share some noise related information and a quick (but hopefully representative) sharpness overviews with some usual lenses in the D5100 realm. The AFS 18-55mm, AFS 55-200mm, AFS 18-200mm VR II, AFS 55-300mm enhanced with some shots with the AFS 14-24mm, AFS 50mm/1.4 and AFS 104mm/2.8 Micro.
Experiences with the Kit lenses (AFS 18-55mm and AFS 18-105mm VR)
I’ll start with the kit lens shots from this morning. To put the proper context around the way how they were produced. Image walking with your wife, engaged in a chat, stopping for a second when something interesting is in your view. Taking images without great preparation. “Tripod” you might ask? Not here. Carefully balancing the light? Nope. Asking your partner to carry some stuff? Are you kidding? Point and shoot at its best. BTW, the LowePro OffTrail 1 waist bag has the perfect size to keep the D5100 with the AFS 18-105mm AND the sun shade. This accelerates significantly the speed to be able to get an image – you don’t have to fiddle around with the sun hood. For those who don’t think about and don’t like the sun shade. There is no simpler and cheaper way to improve the image quality of your precious camera than using this el cheapo accessoir.
Honestly, I never use the AFS 18-55mm for my personal photography. It came along with the D40 in the past, so I have it. Not that it is a bad lens, but it is so tiny, so light that 2 important things are missing for me. A bigger lens allows my left hand to fully support the lens tube and more weight and increased inertia tremendously helps with handhold photography – like VR, more weight gives you another stop or two. Weird as it sounds but that’s my personal approach to this.
It comes at no suprise that you should stay away from the edge cases with this lens (like all others as well). DOn’t expect that the lens performs at its best when at the wide or long end, or shot completely open. Staying 1 step away from these extremes helps any lens and improves your enjoyment with your gear. Before I forget, there is another limit you will to be more careful vs previous generations of D-SLR cameras with lower resolution. There are other things to consider as well, but diffraction limit with a 16,2 MP DX sensor starts at around f8 (based on the usual airy-disc size). So, for maximum sharpness, be careful when you get into f11 and f16 territory – it won’t work with your D5100. This is not a limitation of this particular lens, but a physics phenomenon. No the end of the world, but you should be aware.
With most of the overview images, clicking on the photo get you a larger version, usually 1920x1200mm. Most images of this part of the Nikon D5100 review can be downloaded in full size for personal use from here.
All resized and cropped images used in this part are located here.
Nikon D5100 & AFS 18-55mm @f8 (DSC_0923)
Getting closer to this wooden horse. Nikon D5100 & AFS 18-55mm (@ 55mm and f5.6) – as written above it is a kind of worst case for image sharpness for this lens (DSC_0921)
There are some limitations, what a f5.6 (@55mm) stop lens is able to produce with respect to shallow depth of field (DOF). Depending on your background, bokeh is not too bad either. You might not be happy with leaves and branches in the background, they are notorious hard to come by as blurred background.
Nikon D5100 & AFS 18-55mm (@55mm and f5.6) (DSC_0908)
Last example. Some leaves in the morning sun. Nikon D5100 & AFS 18-55mm (@ 55mm and f5.6) (DSC_0882)
Switching to the AFS 18-105mm VR
This is potentially the most common kit lens people might choose for the D5100. It has a better range then the other kit lens (the AFS 18-55mm) avoid lens changes vs. 2 lens setup. It is cheaper vs. the AFS 18-200mm VR II. There is “more” lens in your hand vs. the AFS 18-55mm. From an IQ perspective, the lens is often underestimated for its sharpness, but quite often people with higher aspirations complained about a high level of distortion. For most others, just switch auto distortion correction on and forget about this aspect in most circumstances. If you intend to move into tabletop, repro and architectural photography, then it is time to move on – lens wise.
Ok, that’s it for the kit lens section. Don’t forget to get to the download page for full size JPEGs if you are interested (link is above)). The files are taken from unaltered NEF files, just converted into JPEGs. For me, it is really impressive what the bar is today for an eadvanced entry level camera.
Sun behind the leaves, D5100 & AFS 18-105mm VR (click for enlarging)
Why the bare 16.2mp sensor is great for night shots, but not the filters.
I need to be careful with this section. The internet has proven to be a place, where single attributes of increasingly more complex consumer technology is often taken out of context and rage wars start if it is this or that, with endless discussions, sometimes quite heated, in between.
I love night photography. Not that I am particulary good at it, but the night is usually the only time where I can enjoy my hobby with ease. During the day, my job during the week and the family over the weekend are somehow a “barrier” to be 100% focussed on a hobby. It is what it is. A great pasttime, a way to express yourself (not on this blog – which is rather focussed on technical aspects) and an excellent way to relax. So, I relax in the night.
Before I zoom in on the issues, I’d like to share my view on the current status of sensor technology and capabilities of cameras. When “real” D-SLR photography started with the revolutionary D1 started in 1999, the most important metric for most photographers quickly became sensor resolution. The more the better. With the advent of consumer cameras, it quickly became the familiar marketing game to fish for more customers with sometimes ridiculus high MP numbers. Betting on the fact that most users didn’t differentiate between Megapixel and real captured information, which is dependent on the whole image pipeline (Lens & camera & and shooting conditions). When the D3 was introduced back in 2007, a new super metric became en-vogue (it was there before as well, biut the D3 raised it to prime time). Leading to the often heard key question with new cameras: How high can its ISO go? Life would be simple, if the world would be simple. There are so many things to be considered, things which need to be done correctly to get an excellent overall technical image quality. It starts with lenses, where besides the omnious sharpness, other aspects are tremendously important. To name a few: Sharpness (center, border, corners), reflections, flare, ghosting, field cuvcature, contrast & micro contrast, spectral transmission, precision of focus mechanism, AF speed, zoom range, price point in the market and many more. This is the reason why it continues to be a formidable challenge for lens designers – it needs to be and usually is a compromise to design a lens.
What is logical and often discussed in internet forums about lenses is less often done about the optical system in front of the sensor – mostly the Bayer filter and the anti aliasing (AA) filter. While it is obvious why they are in the optical pipeline, it’s side effects are less often discussed. Either people are not aware, or they don’t care – I don’t know. Back to our little gem, the Nikon D5100. Similar with previous consumer cameras, they often get the same sensor as their higher priced silblings. A natural consquence of the merits of mass production in silicon technology which benefits from higher production runs. But the key thing is that the other components of the image pipeline within the camera can be optimized for the target market by the manufacturer. Depending on technolog evolution, some components will be better (benefiting from a later market introductio) or worse (because the unit price of this component didn’t fit the price enevelope the manufacturer has on his mind).
How often have you seen and heard that the D90 has a better IQ than the D300? Benefiting from its later market introduction (about 1 year)
But there are use cases, where the D300 benefits from its higher quality components (AA-filter and Bayer filter). I am not talking about the better AF Module, higher speed fps, etc. just the key components around the sensor.
Like the D7000, the Nikon D5100 has this new sensor with excellent high ISO capabilities, really excellent for a DX camera. It can not be overemphasized – it is fantastic. Less fantastic is that Nikon hasn’t found an innovative way to to improve the filters in front of the sensor accordingly. They know how to do it, as seen with the D3-series or the D300/D300s, but the problem is either one of finding a more cost effective way, how it can be used with the D5100/D7000 or it might be a feature serving a market separation function. Differentiating the higher priced bodies from the lower end. Ok, so much for the background.
Let’s view some images. I used the D5100 with 2 different lenses. The AFS 35mm/1.8G and the AFS 35mm/1.4G. For comparison, some shots where also taken with a D7000 and the current Nikon reference camera in low light photography, the Nikon D3s. The scenes aren’t the most compelling ones. They just serve the purpose to illustrate some of the points mentioned below, they are handhold with high ISO to visualize what is possible without using a tripod – opening up a huge amount of new photo opps to those interested in it.
Schönbrunn palace (taken through the gate, it was closed)
D5100 & AFS 35mm/1.8
It is really great to see the sensor sensitivity with the Nikon D5100. Looking at the halo around the light poles in the palace you might attribute the light fields to imprefections to the AFS 35mm/1.8G lens. But is not lens dependent, as you will see later. (click on the image for larger version)
Nikon D5100 & AFS 35mm/1.8G (ISO 1600, f2, 1/50sec)
Here is a crop of the D5100 & AFS 35mm/1.8G. In 100% crop everything looks ugly. Don’t be scared, most of it can be removed in post processing and size reduction to viewable sizes. But this kind of halo is visible in resized images as well.
Here is a crop of the D5100 & AFS 35mm/1.4G, currently the best 35mm Nikkor lens. Doesn’t it look similar?
I thought that Nikon built the D5100 sensor with lower quality AA- and Bayer-filter components. Looking at the D7000 image with the AFS 35mm/1.8G displays the same behavior is evident for the D7000 camera.
Ok, lets compare this scenery with the D3s and the AFS 50mm/1.4 (not the best lens, but similar for its FOV to the 35mm DX lenses) (click to enlarge)
100% crop of the D3s image. See, what I mean by this difference? The halo is significant less pronounced.
This is another opportunity to show the benefit of heavier bodies with heavier lenses. All images above are shot at f2 and 1/50sec (with ISO 1600). With the D3s is it possible for me to shoot 1/15sec with the 50mm lens. It is significantly more error prone to try this with the lighter bodies. Gives me 2 f-stops in this scenery (click to enlarge)
Second scenery: Naschmarkt
D5100 & AFS 35mm/1.8G (ISO 1600, f1.8, 1/15 sec)
D5100 & AFS 35mm/1.4G (ISO 1600, f1.4, 1/25sec)
D7000 & AFS 35mm/1.8G (ISO 1600, f1.8, 1/25sec)
D3s & AFS 50mm/1.4G (ISO 1600, f1.4, 1/30sec)
purple fringing, a lens limitation, can easily be removed
Third scenery: Neuber’s Enkel
I won’t go into details here, but show straight hand hold photography in 2011. It is amazing and enjoyable.
D5100 & AFS 35mm/1.4G (ISO 800, f2.8, 1/200) (click to enlarge)
D3s & AFS 50mm/1.4G (ISO 800, f2.8, 1/125 sec)
Fourth scenery: Burgtheater
D5100 & AFS 35mm/1.8G (ISO 1600, f2, 1/50 sec)
D5100& AFS 35mm/1.4G
D7000 & AFS 35mm/1.4G
D3s & AFS 50mm/1.4G (ISO 1600, f2, 1/50 sec)
Reproducing and isolating the issue
As always, street photography introduces so many irregularities, that I’d like to demonstrate this issue in a more formal way at home in my test lab (just kidding, don’t have a lab. Just a room to play around)
This is the test setup. Besides the ISO 12233 chart, which in this context doesn’t serve any particular purpose besides being there for the next test, the real important thing is the small LED flashlight in the upper right corner. This flashlight points directly to the lens to imitate street lights, like in the images seen before.
Now, a series of 100% crops around the LED flashlight follows:
D5100 & AFS 35mm/1.8G
D5100 & AFS 35mm/1.4G – The lens cannot play to its strength, as the filter layer is limiting image quality
D90 & AFS 35mm/1.8G. Just to show, that the highly regarded D90 has the same quality level filter layer, in line with the D7000
D300s &´AFS 35mm/1.4G. Here comes the difference to the D90. Same sensor substrate, different filter. If this is correct, then the upcoming D400 will utilize the low light capabilities of the 16,2 MP sensor much more than the D7000 is currently able to deliver
D3s & AFS 50mm/1.4G
D3s & AFS 35mm/1.4G
I stay with my initial verdict. The D5100 and the D7000 are impressive low light cameras. There is room for improvement with some components which are less than optimal due to a variety of reasons. If history is an indication (D300 vs. D90) we can look forward to the D400 which, when issued with the same basic sensor, will probably top the D7000/D5100 in this regards. Looking forward to this introduction. Oh and yes. The D3s is good, very good in this regards as well.
Don’t forget to consider the total image pipeline, not only single elements of it.
Noise Map D5100 vs. D7000
There has been rumors after the Nikon D5100 was announced, that it will have better low iso capabilities due to advances in noise reduction. A good opportunity to check the NR settings of the D5100 vs. the D7000. Nikon’s engineers have significant leeway in tuning the software pipeline to optimize the resulting photo. I did run my little D5100 through all ISO settings with the 4 different NR modes (off, low, normal, high). It is completely identical to the settings found in the D7000. Another indication that the sensors and software pipeline are very similar and Nikon kept the changes to a minimum.
Here are the NR values as seen in Capture NX2:
This part grew longer than expected. I did check also the AF low light capabilities of the D5100, D7000 and D3s with moving targets (cars) in the night. In case you are interested, this will be covered in the next part. I did also run most appropriate lenses against an ISO 12233 resolution chart to check the resolution of the most common lenses (with some high end references).
With that, good night for now,
Next page: Part 4 – Battery life, low light performance of the autofocus, dynamic continous night shots