Tamron SP 70-300mm f/4-5.6 Di VC USD : Review

Introduction

This is a review of the Tamron SP 70-300mm f/4-5.6 Di VC USD telephoto lens for the Nikon mount. The test was once again made using a D700 full-frame DSLR. This is a well-built lens, very solid without any wobbling parts, and competes directly with the Nikkor AF-S 70-300mm f/4.5-5.6 VR lens which has been very successful among amateurs and enthusiasts. Here’s how the D700 combines with the Tamron lens:

This has been a very welcome addition to the market, because the Nikkor had been alone price-wise, and third-party options were not up to the standards. Fortunately, the Tamron is a very viable alternative and it even has some characteristics that surpass its direct rival, as we will see. The Tamron belongs to the SP (Super Performance) line, the same line that has, among others, the highly praised 17-50mm f/2.8 (DX) and 28-75mm f/2.8 (FX) lenses. This 70-300mm is the first Tamron lens having the new USD (Ultrasonic Silent Drive) auto-focus motor for fast and silent focusing, and also the first including their innovative optical stabilizer mechanism, VC (Vibration Compensation). Indeed, the lens is very silent and rather quick focusing, but may hunt at times in areas with low contrast, the same happening in low light situations, otherwise the lens focuses quickly. The lens also has special glass elements, LD (Low Dispersion) and XLD (Extra Low Dispersion), which are employed to take care of chromatic aberrations. One thing that’s always good to have is IF (Internal Focusing) and this lens got it, and because of that the front element never rotates when focusing, therefore using a polarizer is no problem.
At a retail price of €389 right now, at a first sight and looking at the specifications, the lens seems to have a great price/perfomance ratio, but that’s the thing we’ll find out later in this review. On the D300 I had before, this lens was stellar for the price, having very good sharpness corner to corner at every aperture and only with a slight drop at 300mm, being a terrific combination for quick operation speed and optical performance in all focal lengths for users who don’t normally shoot in low light situations.

Technical Specifications

Focal length 70 – 300mm
Maximum aperture f/4 – f/5.6
Minimum aperture f/32 – f/45
Field of vision 34 – 8 degrees (on FX)
Weight 765 g
Dimensions 144 x 80mm (194 x 80mm extended)
Optical construction 17 elements in 12 groups (1 LD element, 1 XLD element)
Aperture blades 9
Filter diameter 62mm
Minimal focus distance 144cm (125cm from the front element)
Hood HA005, petal-shaped
Mount Nikon F

Mechanical Characteristics

Zoom ring Plastic with rubber finish
Focus ring Plastic with rubber finish, no infinity stop
Focus throw 160 degrees
Focus motor Ultrasonic Silent Drive, allows full-time manual focus override
Optical stabilizer Vibration Compensation, up to 4 stops capability
Front element rotation while zooming No
Front element rotation while focusing No
Internal focusing Yes
Lens extension while focusing No
Lens extension while zooming Yes, up to 51mm

Handling

The Tamron is a big fat lens and even with the D700 is a little front-heavy, but not overly so. Playing with the zoom ring feels like plastic against plastic, but nevertheless it’s not only smooth enough for quick operation, but also acceptable enough for the times when you need precise framing. Some people complain that the zoom ring is a little stuck on their copy, but that might be true when the lens is brand new; I don’t remember mine having a stuck zoom ring. The focus ring feel is similar, but since it’s lighter it’s easier to turn around.
The lens allows full-time manual focusing without the need to turn the lens or camera to manual focus. Common with other built-in auto-focus motor lenses, the focus ring never stops rotating and is capable to focus past infinity. The Tamron focuses as near as 144cm from the focus plane (125cm from the front element) at 300mm, resulting in a maximum magnification ratio of 1:4, which is less magnification than the previous Tamron 70-300mm was capable of (1:2).
The lens has a long petal-shaped hood and the caps are of good quality. The front cap is similar to Nikon’s, allowing you to take it off and put it on without the need to take the hood off. The lens has a metal mount, which is a must-have for such weight, but doesn’t have a tripod collar, so be sure to have the camera well attached to the tripod socket when using the combo for long exposures.

Resolution

For the resolution test I shot the 5 Euro bill in the studio. Focus whas achieved using Live View to avoid auto-focus imprecisions and to compensate for an hypothetical field curvature.
The first column shows a crop of the image center, the corner crop is on the second column and the third column shows a crop of the extreme corner. Each row represents an aperture setting, from maximum to f/22 in full stops. Here are the results:

70mm

At 70mm the center resolution is already excellent right from f/4 and only deteriorates at f/22 due to diffraction. The corner resolution is on a very good level until f/11 and drops noticeably at f/16. The borders are on a much lower level; by f/4 the sharpness is only acceptable and improves slightly at f/5.6, but reaches a good level from f/8 to f/11 before dropping again at smaller apertures. All in all, a very good performance at this focal length (not that, from now on, if you’re considering using this lens on DX, use the first and second columns only as reference, since the second column shows crops of the same area as the extreme corners on DX).
Moving on to other focus lengths now…

100mm

At 100mm the resolution at the center and corners is excellent already wide-open and only drops by f/22. The extreme corners are good wide-open at reach very good figures by f/5.6 and the resolution drops a bit at f/16. From 100mm to 135mm seems to be the lens’ sweet spot.

200mm

At 200mm the center resolution is always on an excellent level but the corners start to show some limitations. Wide-open and until f/8 the resolution figures are only on an acceptable level and the extreme corners follow closely. At f/11 there’s a sudden increase in quality in the entire frame and that continues at f/16. By f/22 the resolution drops a little but not by much as in previous focal lengths.

300mm

At 300mm, resolution drops noticeably in the entire frame but remains good to very good at the center. On the downside, the corners are never on the same level, and for good edge-to-edge sharpness one has to stop down to f/16 or even f/22. This is not noticeable in the real world, trust me, because at 300mm you tend to put the subject at the center anyway.

Overall, the Tamron shows a strong performance, producing sharp and contrasty images at all focal lengths, and its strongest selling point is the resolution and contrast at 300mm. This is much better than the Sigma 70-300mm f/4-5.6 APO Macro I had before and better than its direct rival, the Nikkor AF-S 70-300mm f/4.5-5.6 G ED VR I had played with.

The only thing that changed my opinion about the Tamron now, compared to the performance delivered with my previous D300, is the noticeable variation in exposure. It is very visible, at all focal lengths, that the lens underexposes a lot at the maximum aperture and less so stopped down, exposes well at f/8 and overexposes at f/11, before underexposing again by f/16. The weird thing is that it wasn’t noticeable at all with the D300, when only I had some minor overexposures in bright sunny days which were corrected simply dialing to -0.3 or -0.7 EV, independently of the aperture I was using. Now I can’t say that my lens doesn’t underexpose or overexposes, but instead a combination of the two that is dependent on the aperture. I can’t really say if this is because of heavy vignetting, but if so it affects the entire pictures and not only the borders, but also there’s a bit of overexposure at f/11 that is visible in the camera histogram. This will be more visibly explained below in the vignetting test.

Distortion

For the distortion test I shot a brick wall, again:

70mm

135mm

300mm

The lens distortion is practically absent at 70mm but starts producing a certain amount of pincushion at 135mm which stays constant until 300mm. On the field this is practically invisible and can be easily corrected in post-processing if needed.

Vignetting

In this test I shot a white wall at home using tungsten white balance and setting exposure manually. The exposures were judged by the camera histogram and are considered to be spot on when it’s centered.

70mm

At 70mm, vignetting is strong wide-open and less so at f/5.6, before improving considerably at f/8. At smaller apertures it disappears completely.

135mm

At 135mm, the same behaviour happens exactly as before.

300mm

At 300mm, only the widest aperture is affected by vignetting but in a very strong way. At other apertures there isn’t any noticeable light losses towards the edges.
As said above, the lens is affected by exposure differences that affect the entire image, as opposed by vignetting which only affects the borders of an image by any degree. I also could not find an explanation for the overexposure seen at f/11 from 70mm to 135mm. This is the first time I had a problem like this one and I’m not sure if the problem is related to this particular copy or any other sample. There are many people who complained about their copy producing under- or overexposed pictures, but mine gives exposures that were unpredictable at first, but can be corrected dialing exposure compensation accordingly depending on the aperture used. This phenomenon is more visible in Matrix metering mode and less so in Center-Weighted mode. I found no differences between aperture and shutter priority, or manual exposure modes. I will try to repeat the test if I have the chance of getting another copy.

Chromatic aberrations

For this test I shot a car roof from above, on a very sunny day early in the afternoon.

70mm

300mm

The Tamron employs low dispersion glass elements to reduce chromatic aberrations and it’s clearly visible that the lens performs as advertised. Aberrations are negligible at almost every aperture, except at its maximum where minor vestiges can be found on extremely high contrast situations, but only at the widest focal length. That’s a very good performance for a lens in this price point.

Coma

Coma is an important requirement in astrophotography and usually affects the corners of most lenses. Lenses that are affected by coma produce comas (hence the name) instead of bright light points in dark backgrounds. One way to test coma is using a LED source of light at home in a dark room.
I put the light source at the center (first column), corner and extreme corner of the frame (second and third columns, respectively), at maximum aperture and stopped down.

70mm

300mm

The lens isn’t much affected by coma and the only distortions visible are in the halos around the center, where they stretch the more the points are closed to the borders, but the points themselves remain perfectly circular. This is not uncommon for lenses with such moderate maximum apertures.

Flare

I shot a building in construction in front of me, with the sun sneaking from a window, to see if I could see any flare vestiges. I started to shoot directly against the sun, then placed the sun at the corner and finally made some shots with the sun just outside the frame.

Shot directly against the sun.
Shot with the sun placed at one corner of the frame.
Shot with the sun just outside the frame.

The images are not much affected in all situations, which shows that the lens has good resistance to flare keeping contrast on a high level. The worst case is when the sun is placed at one corner, where the multiple internal reflections of light may be noticeable in the opposite corner.

Bokeh

Thanks to the 9 aperture blades, we can expect circular out of focus highlights from this lens. I took a defocused picture at the widest aperture of the city lights and got crops of the center, corner and extreme corners. The test was repeated for the subsequent two stops.

70mm

300mm

The bokeh from this lens is perfectly circular at the center and suffers from the cats eye distortion due to vignetting in the corners and more so in the extremes. The out of focus highlights are rather nervous on the inside at 70mm but get perfectly smooth at 300mm, and the edges aren’t much accentuated, resulting in smooth transitions and backgrounds that are soft and not distracting at the longest focal length. All in all, the lens has good bokeh characteristics which are amongst the best in its class, especially when we take the moderate apertures involved into account.

Macro/Close-up

The Tamron was capable to focus as close as 144 centimeters from the sensor plane, which means 125 centimeters from the front element, resulting in a maximum magnification raio of only 1:4 at 300mm. I shot an 1 Euro coin and this is what to expect at the minimum focus distance:

70mm

300mm

Image stabilization

This lens introduced Tamron’s new image stabilization technology, VC (Vibration Compensation), to compensate for unwanted small movements, and it’s advertised as giving an advantage of 4 stops. This means that at 300mm, one can shoot with speeds as slow as 1/20 seconds.
To test it, I shot the back of a street lamp from my window. The first column shows crops of the subject shot with VC off, and the second column shows them with VC on for comparison. Here are the results at 300mm:

It shows to me clearly that the VC works as advertised, resulting in sharp images at 1/20 seconds. Of course, one has to build some technique to hold the camera and lens steady, so that the benefits of this technology can be applied. The image taken at 1/20 is a little softer but due to diffraction (an aperture of f/32 was used).
Contrary to Nikon’s VR mechanism where 2 axis (up and down) are used, VC used 3 axis: up, down and yaw (up-left, down-right) to compensate for diagonal shakes. One more difference to VR is that VC appears to work with full power right from the moment you press the shutter button halfway, and stays there moments after the picture is taken. VR works with less power and only commutes to full power when you press the shutter button down to take the picture. This is the reason why the images in the viewfinder are almost static with VC, when compared to what is visible with VR. This doesn’t mean that VR is much worse, because it’s not, but I found VC to give me more keepers. And it works as advertised, which is amazing.

Summary

Build quality 7 Mostly high quality plastic but all parts are tight together
Handling 7 Nice feel overall with silent and fast AF in most situations, the zoom ring could be damped for smoother operation
Resolution 7 Very good center sharpness, good corners and acceptable extreme corners in most apertures, amongst the best in class
Distortion 9 A little pincushion distortion but not noticeable in everyday shots
Vignetting 8 Strong at biggest apertures, negligible afterwards (not considering the variations in exposure)
Chromatic aberrations 9 Rarely visible, if any only at 70mm at the widest apertures
Coma 9 Not perfect, but the lens almost doesn’t have issues here
Flare 8 May produce some flare in harsh conditions, but contrast stays always on a high level
Bokeh 7 Circular but very nervous with accentuated edges at 70mm, but gets pleasantly smooth at 300mm
Overall 79% Very good lens on DX and FX, a valuable lens if you need stabilization and can’t afford any of the 70-200mm offerings

Samples

Here are some samples of pictures I made with this lens. Settings: native JPEG, picture control set to Landscape mode, no post-processing applied except reducing to 600 pixel width.

300mm, f/5.6, 1/125s, ISO 200
300mm, f/11, 1/160s, ISO 200
70mm, f/4.0, 1/1600s, ISO 200
70mm, f/4.0, 1/2000s, ISO 200, -0.7 EV
270mm, f/5.6, 1/1250s, ISO 200
70mm, f/11, 1/320s, ISO 200
100mm, f/4.2, 1/1250s, ISO 200
Advertisements

Voigtländer Color-Skopar 20mm f/3.5 SL II Aspherical : Review

Introduction

I present you my first review, of the Voigtländer Color-Skopar 20mm f/3.5 SL II Aspherical lens for the Nikon mount. The test was made on a D700 full-frame DSLR. The lens has a very compact design, which is commonly referred as “pancake”, and if you’re like me you’ll appreciate the size and weight of this lens, making a relatively lightweight kit with the D700 for walking around the streets. Watch the funny combo below:

Its retail price of €449 seems fair for a pancake lens. Having a maximum aperture of only f/3.5 it’s not the fastest prime lens out there, but this aperture limitation is common for such design. In fact, there are not many pancake lenses as cheap as this one, and particularly on Nikon land these pancake lenses are a rarity. The lens includes an aspherical element to reduce aberrations and distortions. As a manual focus lens, it can be a bit limiting for you depending on your shooting style, but that’s also true for every manual focus lens such as Nikkor AI or Zeiss ZF lenses. The fact that the lens has a wide focal length eases this limitation a bit, being very easy to focus anywhere you want.
I didn’t write a review of this lens when I had my previous D300, but from the tests I did and all the real world shootings I made, I found this lens to be a very good performer on DX stopped down. I even went on a trip to Holland only with this lens and didn’t need anything else on the streets. On DX, the focal length is just perfect and it was very handy for taking pictures of churches and other tall buildings, street scenes and urban landscapes. This time I’m showing you the results of a much more demanding test (on FX), so let’s find out if the Voigtländer is worth your consideration if you’re searching for a wide angle lens.

Technical Specifications

Focal length 20 mm
Biggest aperture f/3.5
Smallest aperture f/22
Field of vision 94 degrees (on FX)
Weight 205 g
Dimensions 29 x 63 mm
Optical construction 9 elements in 6 groups, 1 aspherical element
Aperture blades 9
Filter diameter 52 mm
Minimal focus distance 20 cm
Hood LH-20N (optional)
Mount AI-S, CPU integrated

Mechanical Characteristics

Zoom ring n/a
Focus ring All-metal with rubber finish, no infinity stop
Focus throw 170 degrees
Focus motor n/a
Optical stabilizer n/a
Front element rotation while zooming n/a
Front element rotation while focusing No
Lens extension while zooming n/a
Lens extension while focusing Yes, 3 mm
Internal focusing No

Handling

The Voigtländer is a very compact lens as you can see above in the pictures, and being all-metal it’s very well built and surprisingly heavy for such size. Being a manual focus lens, probably the most important aspect is how the focus ring feels on your fingers, and it does indeed feel great. The focus ring is perfectly damped and focusing is very smooth. I had been shooting with some manual focus lenses (AI-S and Samyang) but focusing with this lens feels even better than those. It feels smooth and has a long throw of about 170 degrees for precise focusing. This is mostly useful in near distance, of course, because from about 10 meters to infinity there’s almost no need to move the focus ring again. Beware, though, that the lens is capable to focus past infinity which isn’t good news for astrophotographers. Fortunately most cameras today have Live View, which is of precious help if you want precise infinity focus. The focus ring has a rubber finish for a perfect grip.
The less positive things about this lens is that it doesn’t have a lock on the aperture ring, therefore watch out and keep the aperture ring on f/22, otherwise you’ll get a fEE error on your LCD display. Because the lens is so tiny and there’s almost no space between the focus and aperture rings, attaching and dettaching the lens from the camera can be a bit difficult, but with a firm grip on both focus and aperture ring, along with the fixed 3mm spacer in between, it can be done but be warned that the lens gets almost glued to the camera.
Anyway, my only thumbs down goes to the front cap which gets off so easily with just a small touch. There were many times that I had my camera and lens on my shoulder pack and the cap was off. Just throw it out and get a Nikon one.

Resolution

For the resolution test I shot a near distance object and a far away building to find out if there were any visible differences in image quality. I focused using Live View and focused again using this method when moving the target to the corners. Turns out that the lens performs very similarly and all distances, regarding resolution and vignetting, therefore I’m not showing the far distance results this time.
The first column shows a crop of the image center, the corner crop is on the second column and the third column shows a crop of the extreme corner. Each row represents an aperture setting, from maximum to minimum in full stops. For the near distance test I used a 5 Euro bill as target. Here are the results:

You can see that the center resolution is already very high and the images have very good color and contrast. This is always true until f/16, from which diffraction starts making effect. The corner resolution is very weak at f/3.5 and gets acceptable only by f/8, where I mainly use the lens on landscape photography (or f/11 for that matter if there’s enough light). The extreme corners are dismal – this is the effect of putting a pancake lens on a full-frame camera – resolution practically doesn’t exist wide open, and coupled with heavy vignetting this makes an awful combination for available light photography. The extreme corners get much better only by f/16.
If you want corner-to-corner sharpness this lens has to be stopped down to f/16 at least, or f/11 if you can tolerate some mushy corners. Depending on your subject even f/8 can be used, as long as the corners don’t have fine details.
In conclusion, this is a lens to be used on sunny days or with the help of your tripod. This is a huge disappointment to me because the lens is great on DX!

Distortion

For the distortion test I shot a brick wall, of course:

The lens distortion is well controlled. Vertical lines are always straight and there’s no signs of barrel distortion, which is great for taking pictures of buildings. The problem resides on horizontal lines, and this mustache distortion can be very noticeable if the target has many horizontal lines. This type of distortion is very difficult to correct in post-processing, so it is best to avoid placing horizontal lines on the top or the bottom of the frame, where distortion is mostly visible.
As an example, here’s a “real world” photo of an interior. You can see that vertical lines remain straight as long as you hold the camera parallel to the ground. Because there weren’t any horizontal lines due to the framing angle I used, this shot seems perfectly balanced:

Vignetting

In this test I shot a white wall at home using manual white balance.

As detected previously on the resolution test, vignetting is very heavy wide open and things start to get even only by f/8. Notice that this is mostly visible on shots of white walls like this. As a matter of fact I can make a nice use of the vignetting wide open on shots of flowers and other small stuff like that. Fortunately most or all vignetting can be removed in post-processing.

Chromatic aberrations

It was a clear bright sunny day with the sunlight reflecting on my brother’s car and I searched for a spot where to make the shots. I shot the FIAT logo at all apertures because it was so shiny:

The use of an aspherical element is an important feature on any lens, and it’s pretty clear that Cosina made an excellent job to maintain the CA levels negligible. From all the shots I made with this lens there wasn’t a single day I noticed any fringing issues. The lens earns the maximum score here.

Coma

Coma is an important requirement in astrophotography and usually affects the corners of most lenses. This can be tested on the field, on a starry night, to check for weird distortions on the stars. Basically, lenses that are affected by coma produce stars that are no longer light points but comas instead (hence the name). One other way to test it is using a LED source of light at home in a dark room.
I put the light source at the center (first column), corner and extreme corner of the frame (second and third columns, respectively), at f/3.5 and f/5 (one stop down).

You can see that at f/3.5 the lens produces high amounts of coma at the extreme corners, although it gets better by stopping down. This is no big deal since the moderate maximum aperture is not very attractive for astrophotographers anyway, but if you’re taking this lens to the field be sure to use at least f/5 to avoid coma. Other than shooting stars and LEDs, there aren’t any more real world examples I can remember that may produce this coma aberration, so this shouldn’t scare you away.

Flare

I walked around my backyard doing many shots trying to get any flare vestiges. I started to shoot directly against the sun, then placed the sun at the corner and finally made some shots with the sun just outside the frame.

Shot directly against the sun.
Shot with the sun placed at one corner of the frame.
Shot with the sun just outside the frame.

I took many more pictures and tried hard to produce flare but without success. That demonstrates that the Voigtländer has top quality coatings inside and the results put it along with the best lenses I used, with terrific flare resistance, and all of this without the need of the LH-20N hood available separately.

Bokeh

The Voigtländer has a moderate maximum aperture and a very wide angle, so the defocused quality of an image doesn’t get the same importance as with the fastest professional grade lenses. The lens has 9 aperture blades so we can expect pretty circular out of focus highlights.
For this test I took a defocused picture at f/3.5 of the city lights and got crops of the center, corner and extreme corners. The test was repeated for f/4 and f/5.6.

The bokeh from this lens is pretty circular at f/3.5 thanks to the 9 aperture blades, as expected, except at the extreme corners where weird distortions are visible. The bokeh is very nervous internally and the edges are a bit pronounced which of course isn’t attractive at all. Stopping down to f/5.6 starts producing polygonal highlights. Generally, with this lens it’s better to shoot close-up subjects with the maximum aperture, where the center sharpness is already very good, to get the best out of focus renderings.

Macro/Close-up

The Voigtländer was capable to focus as close as 20 centimeters from the sensor plane, that is 12 centimeters from the front element. I shot an 1 Euro coin and this is what to expect at the minimum focus distance:

Summary

Build quality 9 The front cap gets off too easily, otherwise the lens is a jewel
Handling 10 Superb feel of a very solid lens, very easy to focus manually, with exposure being controlled automatically by the camera
Resolution 6 Very good center sharpness, poor corners and horrible extreme corners even at moderate apertures
Distortion 7 Some mustache distortion to be aware of, especially when placing horizontal lines at the top and bottom of the frame
Vignetting 6 May work on your favor or against you, depending on the subject, but it’s easy to correct
Chromatic aberrations 10 None, this is probably the best lens I have with such resistance
Coma 7 Wide open there’s coma at the extreme corners, but stopping down minimizes the problem
Flare 10 Top notch, no vestiges found whatever I have tried
Bokeh 5 Circular only at f/3.5 but very nervous with accentuated edges, and gets polygonal at smaller apertures
Overall 75% Excellent lens on DX, but look elsewhere if you need good edge to edge resolution and low vignetting on FX

Samples

Here are some samples of pictures I made in my backyard. Settings: native JPEG, picture control set to Landscape mode, no post-processing applied except reducing to 600 pixel width.

f/3.5, 1/200s, ISO 200
f/3.5, 1/160s, ISO 200
f/3.5, 1/320s, ISO 200
f/3.5, 1/200s, ISO 200
f/11, 1/15s, ISO 200, -0.7 EV
f/8, 1/400s, ISO 200
f/3.5, 1/320s, ISO 200
f/4, 1/50s, ISO 800