Nude Photo Pro

A few months ago I purchased a 24-105L after years of using mainly prime lenses and I have to say I have been generally impressed by the performance of the zoom. When it first arrived I did a quick test and felt that in most cases it was a more than reasonable match for my Canon primes in professional use, so it would seldom be necessary in practice to take off the zoom in favour of the primes. This was good news if true at least on from the standpoint of convenience.

Let me say immediately that pixel peeping is not my hobby and that certain small differences seen on screen will never be visible in a printed catalogue or large inkjet print for example, especially after sharpening, however, there is no harm in knowing the limitations of your lenses so a test like this is always worthwhile doing if you have any doubts.

Recently during a forum discussion I was asked about the tests I ran by somebody who believed that a zoom could never come close to a prime, but of course I had already deleted those images, so unfortunately I had nothing to show. A few days later I decided to run the test again in a fairly similar set up to try and establish how valid my initial comparisons were. When I ran the original test it was a nice day during the summer, but now we are in November and I’ve had to wait a week just to find an hour or two during the day when it wasn’t pouring with rain.

This review isn’t perfect and few are, but I think it serves as a reasonable indication of the various merits of each lens. To test this out, all images were shot using single shot auto focus on a 1DsII that was mounted on a Gitzo tripod and head as shown below. Each picture was taken with a cable release and IS was turned off on the zoom to make it a level playing field. I relied on the built in matrix metering for the exposures and every lens was used with with no filters, but with their standard Canon lenshood.

Canon 1DsII on Gitzo tripod

The subject to be photographed for testing was a fence in my garden. That is about as dull as you can get and a million miles away from my usual model photography, but this particular test is simply aimed at establishing whether or not a good quality L series zoom is a reasonable substitute in terms of sharpness for some of my Canon prime lenses. The fence has good texture, so it is a useful way to determine how well the lenses perform.

Test image fence in the garden

I ran the test by shooting a couple of frames at each focal length from F4 through to F22 and all at 100 ISO. F4 is the widest aperture on the zoom, so it seemed like the only practical place to start and theoretically hands a slight advantage to the primes, since most lenses start to improve a couple stops down from maximum aperture. The primes also benefitted slightly from a rare brightening of the weather that would increase the contrast slightly and possibly lead to a perception of greater sharpness.

None of the images had any sharpening applied and the closeups viewed at 100% are taken from Raw files processed in Photoshop CS4 and saved as Tifs. The screenshots were all saved for web at 70% quality which is visibly very close to the originals but saves on the download size for this web page.

This was rather a lot of work to write and prepare, so I have decided to split this into various parts and I will create a separate post each time when comparing the 24-105L against the 35, 50mm, 85mm and the 100m macro.

That is probably enough explanation, so let’s get on with the results…

I don’t own a 24mm prime so I started at 28mm to compare side by side with 28mm F2.8. As with all the lenses, I initially ran through the full aperture range, however, when preparing this blog post it soon became evident that it would become the longest post in history if I included every single image, so for the purpose of reasonable compromise I have decided to limit it to the images shot at F4, F8 and F16.

There are some focal lengths that I just don’t like for some reason and 28mm is one of them, so this was probably the comparison of least interest to me, since just about everything I do for work purposes shooting nudes and beauty images etc. can be handled within a range of about 35-100mm. Nevertheless it’s good to know how they stack up and whether I need to pack that prime when leaving on a job.

First of all 28mm @ F4.

Canon 28mm edge sharpness at F4

Canon 24-105L edge sharpness at 28mm on F4

The above two images are relatively close at the extreme edge, but it it looks like the zoom is already starting to show some advantage even just a little way in and that difference increased notably in the centre area as shown below.

28mm prime @F4 sharpness at centre

24-105L @ F4 on 28mm central sharpness

To me the difference here is like day and night. The prime is clearly disappointing but I feel the zoom has acquitted itself admirably well and while I expect good results from an L series zoom I never would have believed that such a difference was possible a couple years ago. Yes the prime will certainly sharpen up to an acceptable level in print, but clearly the zoom has the advantage.

I have used the 28mm prime on a couple of small jobs in the past and it was always OK but I was never overly impressed and this appears to demonstrate that it’s not such a fantastic performer. On this basis I’d choose the zoom every time, but let’s look at the other apertures to see if they make any difference.

Now at 28mm on F8

Canon 28mm @F8 edge sharpness

Zoom on 28mm at F8 edge sharpness

These two above are fairly close and I think it depends in part where you look which is best.

28mm prime @ F8 centre sharpness

Zoom at 28mm on F8 centre sharpness

Here again the zoom is clearly better and I can see little justification at this point for favouring the prime over the 24-105L zoom.

Now at 28mm on F16

28mm prime @F16 edge sharpness

Zoom at 28mm on F16 edge sharpness

Here above we see once again that the lenses are relatively close at the extreme edges of the image and here at last as we move towards the centre we can see the prime starting to close the gap on the zoom though it is never quite as good.

28mm prime on F16 centre sharpness

24-105L zoom on 28mm at F16 centre sharpness

Viewing this test very much confirms what I saw in my earlier tests that the 24-105L is more than capable of holding its own against certain Canon prime lenses and in this case there is no doubt in my mind that it out performs the 28mm F2.8 prime by a healthy margin right across the aperture range.

Wide angle lenses are severely tested by the high resolution sensor on the 1DsII so you really need good glass to do the camera justice and as you can see the 28mm F2.8 does seem to be struggling here. I have no doubt this will be the worst of my prime lenses, so with the other tests I expect the differences to be less clear cut and in certain cases the primes may well pull ahead as common theory would normally expect.

In a day or two I shall make a second blog post comparing the 24-105L against my 35mm F2.0. In the meanwhile I would be interested to hear any of your thoughts of comments.

Edit:

The second part test of 35mm F2.0 v 24-105L is now available here.