Wednesday, 11 May 2016

Are Tele-converters worthwhile

Most people who know me know I don't use a teleconverter and indeed I don't even own one. I know a lot of people who do own and use them but most of the time I'm either unimpressed by the results or even when they look ok I tend to think they would probably have looked better without one and doing a slightly tighter crop.
I certainly know several people who have purchased them but abandoned using them after a few days because they are so disappointed in the results - though it must be said that I also know a couple of people who swear by them.
Anyway this last week a friend of mine lent me his 1.4x converter (one of those who abandoned) so I set about trying to get hard evidence that my perceptions ( i.e.that they are a waste of time) were correct.

I have previously tried to get an answer to this question by searching the internet but have failed to find a report where the results with and without converter are compared directly.


Since my lens ( Canon 400mm f5.6) is not image stabilised all test shots were taken with the camera mounted on a tripod. In every test multiple images (normally 10) were taken so as to avoid the possibility of camera shake or a slightly missed focus influencing the results.

The targets used for these tests were either A4 photographs of birds where lots of plumage detail could be seen or a test chart published by Bob Atkins (ref    )

The targets were set up approx 12- 15 yards from the camera (a Canon 7D mk 2).

Downstream processing was zero other than cropping and resizing so that both set ups gave an image of exactly the same size. Obviously the greater magnification achieved with the converter in place meant the reduction in pixels was significantly greater for this set up. For example using the Bob Atkins target the cropped image without the converter was 940 pixels wide whereas with converter it was 1325; both were reduced to 800 pixels for use in this blog. No attempt was made to optimise image quality in Photoshop.

The first set of shots were taken using a picture of a little green bee-eater as the target. On the back of the camera I was convinced the results were better without the converter but when the shots were processed as described above there was very little between the best shots at either setting though more pictures were poor/unacceptable when the converter was used. I now think this may have been due to the low shutter speeds -  it wasn't a very bright day and I had 1/500th without converter and 1/320th with (more on this later):

Without converter
With converter
Whilst the differences are small if I really had to choose between the two I'd go with the shot taken without the converter.

A day or two later I repeated this test using a picture of an Indian Pitta.

without converter
With converter

In this test I was focusing on the breast plumage (NOT the eye) and the results were pretty clear - the image after re-sizing was better WITH the converter than without. This time I didn't see the failure rate I saw with the bee-eater shots.  I attribute this to the greater shutter speed  achieved on this day - it was sunny and even with converter I was getting 1/5000th sec.

The final test was with the Bob Atkins chart. Here the results were again unequivocal - the more detailed images were achieved WITH the converter.

No converter
with converter

As before these are straight out of the camera.
The place to see the difference in sharpness and resolution is with the  4.5 and 5.0 sets of horizontal and vertical lines - without the converter complete resolution is achieved with the 4.5 set but lost with the 5.0 set. With the converter complete resolution is achieved at 5.0 and is almost achieved at 5.6; it's gone by 6.3. (this is easier to see on the raw images than on the blog shots but just look at the 4.5 block and the difference in quality is clear)
For these shots the camera was around 14-15 yards from the target and the "4.5" target width (horizontal and vertical lines) is 5.0 mm wide and approx 3.5mm high.

I did repeat the test without a converter but using f8 (the aperture you get when using a 1.4x converter) but the results were identical to those achieved at f5.6.

So there you have it. I was wrong. In my hands and under my test conditions the 400mm f5.6 with converter outperforms the lens alone.

Using the converter does have significant drawbacks in that the best you can get is f8. This has a significant impact on the shutter speed, decreasing your ability to freeze any movement of the subject and makes the set up more sensitive to camera shake.  This latter point was very noticeable during the test where, even though I was on a decent tripod and sitting in a chair, the amount of camera shake with the converter was a lot more apparent than without. Also when the light was poorest (the bee-eater shots) several shots were lost to motion blur when using the converter. I didn't try using remote or auto shutter release or self timer. Finally it should be remembered that the 400mm f5.6 does not have image stabilisation so this problem may not be so apparent for an IS lens.

I also expect the converter will increase the time the lens takes to focus but I didn't try to assess this not least because I have no idea how to attempt it in a quantifiable manner.

So where does this leave me.

I think I have to see how much my friend wants for his converter; if the price is right I'll buy it. For most of my work I will continue to use the lens without a converter especially for motion shots where shutter speed and speed of focus is paramount (I have enough trouble getting decent hand held shots as it is; I doubt if I'd get any at all with a converter in place). I would certainly try to use it if the subject was standing still and/or the light was decent - though it would have to stand there long enough to put the thing on!

How do I reconcile my results with the perceptions I mentioned at the start. Certainly those who have given up on a converter only hand hold so perhaps their problem is motion blur. Certainly one of the guys who swears by them uses a tripod a lot of the time minimising motion blur. So perhaps my previous conclusion were due to the greater difficulty in holding the set up steady with a converter rather than the intrinsic quality of the image.