Archive | Lens Reviews RSS for this section

Tamron SP 24~70 Di VC USD Review


This is going to be a quick review of this lens. There’s not a lot to say since the lens pretty much works as advertised. I’ve used this lens for a couple of months and have tested it and compared it to my other lenses.

The lens is intended as a ‘walk-around’ workhorse for full frame cameras and is in direct competition with Canon and Nikon offerings. The first point to make about the lens is the cost. Canon’s 24~70 f2.8 has a list price in NZ of $3,390, Nikon’s 24~70 f/2.8 is $2,400. The Tamron is only $1,398. Of course all of these lenses would be available cheaper if you shopped around, so this is for comparison only.

The focal length range is very useful. At the wide end 24mm gives you a decent wide angle. I find 28mm a little conservative, and I much prefer 24mm. At the long end it is just long enough for looser portrait shots. With a wide maximum aperture you have plenty of scope for shallow depth of field shots. But if you have an APSC camera (Canon 700D, Nikon D5200 and similar) the focal length range is not so useful, taking you only from standard to moderate telephoto range, with no wide-angle)

At first glance the lens looks like it would be the ideal companion to a Canon 6D or Nikon D610.

In terms of sharpness the lens performs very well. The lens is as sharp as my prime lenses. In the centre the lens retains its sharpness very nicely except for the most extreme apertures. In the corners there is some loss of sharpness and vignetting at wide apertures, but compared to other lenses this is nothing more than what one could expect. Stopping down from f2.8 to f/3.2 produces a noticeable improvement with very little loss of light. The lens performs brilliantly at f5.6 and f/8, with only a slight fall off in quality as you move away form the ideal apertures.

The lens does exhibit some rather obvious barrel distortion at the wide end and pincushion distortion at the 70mm end. Seeing the rather obvious results on my shots of a test chart I went into the dojo and photographed a rack of naginata – this subject being full of horizontal and vertical lines which should be about the worst case scenario to show this up. Fortunately the results are not nearly as bad as I expected. You can see the distortion if you look for it, but it’s not too much in the way, especially considering the subject.

Tamron-1980 Tamron-1981

Can you tell which is which? I’m sure you can if you know what you are looking for, but if you can’t, then the distortion will never worry you. The distortion can also be corrected by applying the Lens Profile correction in Lightroom.

In terms of construction the lens is well put together, and feels more solid than most Tamron lenses I’ve used in the past. The switches for manual/auto focus and the on/off switch for the VC (image stabilisation) are a little small and fiddly. For the VC this is not a problem since it is not something that I would want to accidentally knock out of  place, but I would prefer a more accessible autofocus switch since this is one control that I will access during shooting.

The focus ring can be adjusted during autofocus, but the manual warns against doing so, saying that this could cause damage to the autofocus system. This does seem a little confusing – if indeed the lens can be damaged this way it would be a very easy mistake to make, especially if you did not bother to read the manual. The auto focus is fast and silent.

The closest focusing distance is not a true macro lens, but better than I expected for a general purpose lens. The shot below will give you a practical idea of how close you can get. The battery is an AA size.


The lens appears to be weather sealed, but to what extent it is I do not know. I don’t plan to leave it out in the rain, but it’s nice to know there is some attempt to prevent weather damage in the design, should an emergency arise.

The lens is not small. when you put it on a full size DSLR you have quite a big camera – here the lens is mounted on a 6D, beside a more compact Canon 350D.


But overall how do I think it stacks up?

I think this is an excellent lens. The image quality is excellent. The fiddly auto focus switch is the only real niggle I have with it, but I can live with that especially at the price.

And the price really is an issue. What we have here is a well built 24~70 with great optical performance and good build quality that is half the price of the Canon or Nikon lenses, and it is stabilised. Neither the Nikon or Canon equivalents are stabilised and that really sets the seal on the lens for me. Stabilisation is a really important factor in low light photography. I always seem to be shooting in situations where there is not enough light, and I know form the student work I see that this is true for most people.

So we have to ask… I’m sure the Canon and Nikon lenses are very nice – but why would you pay twice the price for a lens that is probably no better optically than the Tamron, yet is not stabilised?

In short this lens would seem to be the obvious lens to own if you have a Nikon 610 or Canon 6D. And the more I think about it I can’t see why you’d want to spend more money even if you had a 5DIII or D800E

The shot below is taken at 70mm, f/5, 1/30th of a second at 1600 ISO. I could have probably dropped down to f/8 or maybe even f/11 if I had needed more depth of field, but if the lens had not been stabilised I’d probably be thinking about opening the aperture until I had a higher shutter speed.



Going full frame on a tight budget

I was going to write this up as a proper comparison review, but I think it works better as a story. The moral of this story is to know exactly what it is that youcan purchase more wisely if you know exactly what you want your camera to do.

I decided to go to a full frame camera. The use of full frame sensors offers great low light performance and really good control over depth of field. If you are someone from my generation it kind of makes you feel like the lenses are behaving the way they are supposed to. The key to making an economically sensible camera purchase is to know exactly what you want from the camera. These days all the major manufacturers make great cameras, but some are better fits for individual approaches than others. Look for the camera that seems to have been designed with you in mind.

I decided to go for the Canon 6D. This camera has the same core functionality as its more expensive sibling, the 5D mark III, but less a few bells and whistles. Perhaps if I was a sport photographer I’d miss the more sophisticated autofocus system in the more expensive camera or the competing Nikon D600, but I’m not, and I don’t. What is important to me is low light performance. It always seems to work out that I’m trying to take a hand-held photo right at the edge of the cameras low light performance (and, from looking at student work, I think a lot of other people are in the same boat!). The Canon 6D is a relatively no-frills camera (at least compared to other models at the same price point) with really good image quality and low light performance. Sounds like a camera designed for me!

The trouble is that in moving to a full frame camera you need to update all your lenses. If money was no object I’d just buy the 16~35 f:2.8 L series lens and the Tamron f:2.8 24~70 stabilised lens, but those two nice lenses would cost more than the camera. In fact, the budget that I needed to work on gave me enough for a body, but very little left over for lenses.

Again the solution is to know exactly what you need. Most of my shots involve wide angles, and as I mentioned, low light performance is important. I’ve been noting the focal lengths that I’ve been using on my Pentax K-5, and I’m consistently using the wide end of by 10~20mm zoom. In fact I think I’m using the wide end because that’s as far as the ring goes, and I’d be better off using a setting of around 12mm or 13mm. At that length, the lens itself doesn’t dominate the image, and my choice of angle is what makes or breaks the image. At 10mm it tends to be the drama of the ultra-wide angle that sells the image, and I want my photography to be more about my vision than a technical gimmick.

Converting 12mm or 13mm to full frame focal lengths that means I want a 20mm lens, and because I tend to work in poor light, it should ideally be a fast lens.

The other lengths I use are a short telephoto for shots of people, and a slightly-wider-than-standard length for buildings.

My first purchase, because it was cheap, was a used Canon 20~35mm zoom. I was influenced by Ken Rockwell’s review of this lens ( Unfortunately, while Ken often has some wise things to say, he also says things that are quite wrong, and says them with such complete conviction that they seem very believable. In this case I think he got the lens wrong. To handle, the lens is certainly light, but feels flimsy and cheaply constructed. The lens performs adequately, but never brilliantly, and has very significant distortion and chromatic aberration in the corners. Ken mentions the in-camera correction, but fails to mention that this degrades an already less-than-stellar performance further. It’s inexpensive, and you get what you pay for.


Then I came across a used 20mm f:2.8 Canon prime lens for around $NZ450. This lens looks very similar to the 20~35, but feels a lot more solid and better constructed. Also it’s faster, as prime lenses usually are. Comparing the two lenses, the prime is significantly sharper when both lenses are at their best (in the centre at f:8), and the prime lens is far superior in the corners. The prime does vignette when it is wide open, but I can live with that since I often add a little vignetting to images in Lightroom anyway. The prime lens completely out performs the zoom in every respect (except the ability to zoom). So the zoom will soon be for sale on Trade Me – it will be great for someone on a tight budget, but the prime suits me better..

For a short telephoto I already had the perfect lens in my cupboard – an old Tamron SP 90mm f:2.5 prime lens. This is a great focal length for people shots, and this lens is one of Tamron’s best. It’s razor sharp, fast, has great colour, and even has a reasonable macro performance. All I needed was an appropriate adaptor. The lens does not autofocus, and does not record the aperture in the EXIF data, but I can live with that. Manual focussing seems to help me concentrate, and since I’m not shooting sports it is not a problem.

To replace the standard-ish lens I picked up an old 28mm Tamron lens for $NZ25, again manual focus, but I tend to use this on buildings, and they don’t move around very much. I also had, in the cupboard, an ancient Pentax M42 50mm with an adaptor and a Pentax 135mm K mount lens* that also have wonderful optics. I think for anyone on a tight budget who doesn’t want to sacrifice picture quality, you need to be looking at vintage manual focus lenses. While there are plenty of lemons out there, there are also lenses that are the equal (or better) than the best lenses available today, manufactured to higher standards and with beautifully damped focus rings. They are less convenient, and really not useful for action shots, but I think that’s a reasonable compromise.

The result is that my main lens has all the convenience of a modern lens: autofocus and aperture set by the camera, and recorded in the EXIF data. The lenses I use less frequently have superb optical quality at the cost of manual focus. The thing I’ve noticed about using prime lenses is that you tend to see differently, especially if you go out shooting with only one lens. With everything else stripped away your choices are what to point the camera at, and what angle to use – core photographic choices. With fewer choices, I seem to see more.

The main point is that, if you know exactly what you want a camera to do, you can tailor your purchases to exactly that, and minimise the outlay.

*Pentax K mount lenses can mount, with an adaptor on Canon EF-S bodies (650D, 60D, 7D, etc) but don’t work on full frame bodies. The EF-S mount has the mirror further back, but when you put a K-mount lens on a full frame body the aperture lever will strike the camera’s mirror. The solution sounds rather harsh – take the back off the lens and cut the aperture lever off. This ruins the lens in terms of using it as a K mount lens, but the fact is that an M Series lens is actually a pain to use on a Pentax body anyway – you need to use manual exposure and meter by using the depth-of-field preview button. With the aperture lever snapped off you could use Aperture priority automatic. As you stop the lens down the viewfinder would get darker, but that’s probably better than the other way.

Samyang 8mm Fisheye compared with other lenses

There are a number of reviews on the Samyang fisheye on the net, and I’m not going to spend a lot of time repeating what has already been said. To sum up the obvious points, the lens is available under a wide variety of names – Opteka, Pro Optic, and Bower to name a few, and the lens is identical except for the badge. It’s a solidly constructed lens that is inexpensive and has great optics. On the down side it is completely manual focus. On a Pentax body the lens will integrate with the cameras electronics, but with other manufacturers’ bodies the lens control is by the aperture ring, meaning that the camera has no control over the aperture, and the image in the viewfinder gets darker as you stop down.


I’ve been using this lens for a couple of years now, and I’m quite used to it. It’s one of the most useful lenses I own, but then using a fisheye is an acquired taste (and one that I have acquired).

The need to focus manually is not usually a problem since all fisheye lenses have massive depth of field anyway. However it is possible to get everything slightly out of focus if you are not careful. Sometimes what looks OK in the viewfinder is actually just slightly out of focus, enough to ruin a shot. Because the depth of field is so huge it’s less obvious that the main subject of the shot might be slightly out of focus. To avoid this problem I’ve taken the unusual step of focusing using the distance scale on the lens. Since the huge depth of field makes focusing very forgiving this actually works surprisingly well.

What is special about this lens, though, is its ‘projection’. Unlike other fisheye lenses, this lens is remarkably free of distortion. What I’ve done below is to set up a scene that reveals this and I’ve shot it using three lenses – the Sigma 10~20mm f:3.5 (a good quality wide-angle zoom), the Pentax 10~17mm fisheye zoom (a more conventional fisheye lens), and the Samyang. I’ve used the two zoom lenses at both extremes of their ranges, and moved the camera back and forth so that it was roughly the same shot size, although, as you will see, the lenses all work quite differently. The camera was a Pentax K-5.

To start with let’s look at the Sigma 10~20mm at it’s 20mm setting. Here we have the most normal of all the shots, with minimal distortion. The two pillars are straight. Notice the two circular objects in the bottom left – a yellow ball and a reflector. Both of these are very slightly distorted. You need to look closely to see it (but it will be obvious in the next image). [Click to see the images at a larger size]


Here I’ve moved the camera forward and zoomed wide. This kind of scenario is very difficult for any ultra-wide angle and we would expect to see some problems. The two pillars are showing barrel distortion, bending outwards. In a perfect world these would be as straight as the 20mm shot, but some barrel distortion is inevitable, and it would seem that Sigma has done a good job at controlling it. But look at the two circular objects. Both are now clearly distorted. All rectilinear wide-angle lenses distort in this way, and often objects near the corners can be quite misshapen.


Lets move on to the Pentax fisheye at 17mm. Here we can see the pronounced barrel distortion that gives Fisheye lenses their name. What is considered undesirable in a rectilinear wide-angle like the Sigma is treasured in the fisheye, but at 17mm the effect is not very strong, and in fact can be corrected in software without too much degradation of the image. Look at the circular objects. These are now almost round again – but in fact are slight distorted in the opposite direction to the Sigma shot.


Moving on to the Pentax Fisheye at 10mm. Now we have the classic fisheye look with massive barrel distortion. Look at the circular objects – here they are distorted just as much as they were in the Sigma 10mm shot, but the distortion is in a different direction. While the Sigma distorted the circles by squeezing them towards the centre, here they are being squashed on a line around the centre. We might describe the sigma’s distortion as radial or centripetal, and the fisheye’s distortion as circular.


Lastly we come to the Samyang lens. This time the camera is very close to the wall – perhaps only a metre away. As expected the fisheye effect is even more pronounced – after all the Samyang lens is an 8mm compared to the Pentax’s 10mm (but focal lengths at this range are not accurate anyway). Look at the yellow ball. It is nearly round, perhaps not perfectly round, but remarkably round compared to the amount of fishy-ness in the image. The reflector is actually slightly distorted, but that is because it is a disc shaped object and we are now so close to it that we are looking at it side on. The ball, being a sphere, reveals how free of distortion this lens is.


It’s also worth noting that, comparing the Pentax at 10mm and the Samyang, while the Samyang has just a little more in the shot at the top and bottom, it has quite a bit more of the room at the left and right edges. What this means in effect is that the Samyang lens seems more ‘open’ and the Pentax seems a little ‘cramped’. Using the Pentax after using the Samyang feels a little disappointing because it is that little bit less wide (I mean wide here in the sense of wide from one side to the other, not in the sense of wide-angle lenses).

The big plus that the Samyang lens has is that it renders shapes near the edges with remarkable accuracy considering the amount of fishy-ness it produces.