Teles with Teleconverters (on x2 crop factor cameras)

Too much is never enough - or is it?

why

Many photographers starting out into 'birding' or other telephoto work always lament that their lens just isn't long enough. I grew up in the time of 35mm film (well and other films too) and a 300mm lens was something exotic that I just dreamed of.

For a long time I had to make do with a 80-200mm zoom (which actually did some good work for me) and eventually (as time passed) bought a few 300mm zooms to see. I liked them but they were often not as good at f5.6 (looking better at f8). The step up to a high quality fast (like f4) prime lens was out of my league.

Much later in my life (like a few years ago) I discovered the benefits of rapid obsolescence: NB the wealth of older Manual Focus lenses the market.

By this time I've moved over to Panasonic micro4/3 and could take advantage of insanely low prices in FD lenses (which pretty much could not be used on anything else previously {except NEX}).

The 300mm class is a good one (especially in micro4/3)  for a number of reasons:

• lighter weight (have you picked up a 600mm)
• lower prices (I paid sweet-F-all for my 300mm)
• good handling characteristics
• bright

I have owned both the Canon FD300mm f4 and the Olympus OM 300mm f4.5 (see my views on that here), and found them both neck and neck. Back in that article I made the observation that using the 300mm on a G1 body was like using it on a 35mm Full Frame camera with a teleconverter (and perhaps perfect film scans for those of us using film still).

Actually I've personally found that 300mm is enough to get you good shots of BSB's (that's Bloody Small Birds) and that Manual Focus actually helps you to get images of the thing you want in focus. In this image an AF system would drive you nuts focusing on the branches either side of the bird and after struggling with the camera the bird will be gone.

I often don't think there's more point in getting more than effectively 600mm as being closer actually gets you a better shot.

Well, since teleconverters exist (and can be quite good back in the Film Daze) I wondered about how it would actually work (in reality) on the smaller format digital, especially considering we are already doing an effective optical x2 anyway.

I went out and got a Canon x2-A teleconverter (there is also an x2-B, but the A is designed specifically with 300mm in mind) and had a look at the images. The excellent MIR site shows more information on the Canon x2-A extender here. I picked that teleconverter because I was using a Canon lens which it was designed for and its well regarded. I would expect that any other teleconverter would likely give me only equal or inferior results.

summary

• Is it worth the effort? to me, not really ... you'll need to work it out for yourself and your needs
• you don't get something for nothing, there is an exact x2 tradeoff ... perhaps more
• can upscaling solve the problem for you? Upscaling does have advantages.

the data

So, I took two shots (well more really, I took f4 and f5.6 and the bloody wasp kept moving around so actually I took lots more) one with the Canon x2-A teleconverter.

FD300 overview

FD300 + x2-A

Well it sort of looks like a x2 enlargement doesn't it ;-) Considering that these are scaled back for the WWW if you aren't printing them large, then a crop of the 4000x3000 native image will look just as good. For your interest here is the FD300 image cropped to match the x2-A image (bringing it to 2000 x 1500) and then resized down again to 1600 for web.

Similar isn't it. If you're only shooting for WWW (and not street side signage or posters) then of course the native sizes allow a x2 'enlargement' by being so bloody good at 100% pixels that you can crop the guts out of the images and still have room for a good size.

But that's not the whole picture (of course), as the exposure data is important. Firstly I used 400ISO for this test because I personally feel that after 400ISO the image quality falls off.
FD300@f4 gave - 2500th of a sec (handy to seize movement)
FD300@f4 + x2-A gave - 800th of a sec which is starting to push shit up hill with a fork because as this is effectively a 1200mm that's below the 1 over focal length rule. Actually I had to use a cable release as well as my most stout tripod (and yes, that means I took more pictures, evaluated them found camera shake and took more again).

So yes, you do loose 2 stops of speed (recalling that a stop is either halving or doubling the speed) so two stops down from 2500th - 1250 - 625th (assuming stepless shutter speeds). Now if I'd used 2 stops more ISO to bring the speed up on the x2-A shots I'd need to be 1600, which is like popcorn if you ask me ... YMMV

Ok, so lets look at 50% pixel enlargements. Why do I choose 50%? Well my personal experience is that looking on the screen @50% is very similar to looking at a print closely.

Also its important that if you want to look at these images as I screen grabbed them, then right click the image and open it into a new tab: because blogger scales them down to fit onto the screen.

Ok, so again its still clear that its a x2 enlargement, but something else is starting to become clear ... the x2-A makes the image appear a little less sharp (or perhaps viewed another way, magnifies the "we've already reached" the limits of the lens point).

To be honest contrast isn't as good on the x2-A image either (if you ask me).

Ok, so lets do the mathematical version of the x2 enlargement using a bicubic upsample of the plain 300mm lens image (on the right still)

Actually its not too different if you ask me ... however ultimate pixel peeping (like why?) shows that its not actually exactly the same:

and we see that the bicubic upscale has added a little gritty noise to the image (in the greens even!) and so it may just be that in the ultimate call on this: if you need every last pixel of the image and you really need to get in closer then the x2-A works better than a bicubic upscale.

IF you can live with the much slower shutter speed (cos a blurry image from subject motion blur will suck worse) then there seems to be a slight advantage to the TC, but if you really need the speed (like say under forest cover) then crop and upscale will allow you 2 stops more speed (with perhaps only little difference in quality).

As it happens (not presented here) I prefer the look of the FD300 stopped down to f5.6, it was that bit sharper and more contrasty. However I didn't present that here because on the x2-A version the shutter speed was down to 320th of a sec and most images weren't clear.

Conclusion

Its a tough call to say one is better than the other but if shutter speed counts then upsampled images will work as well (and you won't need the adapter :-)

Lastly, I'm not disappointed that I have it as I've used it on the FD200f4 that I have and found that its not bad. So when I'm not wanting to drag along my FD300 (which is not only physically large, but weighs about 1Kg) I can put the FD200f4 (440g and something I'm likely to carry anyway actually ) and my adapter (200g) in instead.

Have Fun