Since the 1DX hasn't delivered yet there aren't many in the wild for testing. Canon disables AF at f8 on the 1DX so the camera body has to be tricked. There was a thread regarding 5DIII AF (same system or nearly so) at f8 recently, using two 1.4X TCs, one of them non-reporting. Performance was said to be OK but not outstanding.

http://www.naturescapes.net/phpBB3/viewtopic.php?f=57&t=214533

Another way to accomplish the same goal with a 2X TC (1.4X on an 800mm) is to place a 12mm non-reporting extension tube between the lens and TC. (The lens will no longer focus to infinity.) Since a 2X has fewer optical elements than two 1.4X TCs I'd be interested to see how well that works.

Everyone here brings up great points. For what it's worth I don't care too much about F8 autofocus. I find that shooting my 600mm prime and cropping produces much better image quality then using a 1.4x or a 2x converter. I shoot with a 1DIV and was really glad I could autofocus at F8, but after taking hundreds of shots with my 2x TC and trying another copy I gave it up. Not only were images softer with the TC but autofocus is so painfully slow at F8 that it missed most of the time. I was trying to shoot BIF to stationary wildlife and was never happy with the images.

Now I shoot without TC's on this lens and get tack sharp images consistently. Cropping in post works far better for me. And I get faster shutter speeds because I can shoot wide open at F4 and use lower ISO for cleaner shots.

I confirmed this by testing with Reikan FoCal software. IQ comparisons between using my lens without any TCs and using a 1.4 or a 2.0x concluded that the image quality loss by using the TC's was far greater then simply cropping images shot without a TC. I then tested by shooting static subjects with and without TC's and cropped in photoshop so that the field of view was the same between images. The cleanest prints by far were shot without any TC. The only negative is that the resulting cropped file is a lower Megapixel count.

I do regularly use my 1.4x on my 300 F2.8 and get excellent image quality, but even on that lens the 2x degrades the image so much I am better off cropping in post.

From what I have heard Nikon TC's do a much better job at preserving image quality and until canon improves their TC glass I will go without and not be worried about a camera that can't focus at F8....

It seems that no two of us work exactly alike and the important thing is to find what works best for you. That said, both the 1.4 and 2x have been important tools for me with 300, 500 and 600mm lenses for many years and my experience has been that they have given me better images than I have achieved by cropping--which I am sometimes forced to do when things happen quickly.

On the other hand, I don't use a 1DIV and I am definitely not a tester.

I never was a tester until I got into camera bodies with Microadjust Autofocus (MA). After calibrating my lenses I found that all my lenses on both my camera bodies required a certain degree of adjustment to ensure that the lens actually focused on what my focus point in the body was hitting on.

Before I calibrated my lenses I noticed that I often got better photos with my 2x then without it and actually found that when I added the 2x to my lens it "corrected" for the MA issues I was having on that lens. So without the 2x my lens was focusing in front of my subject rendering softness in areas like the eye of an animal where I was aiming. Once I put the 2x it optically corrected this issue so that my lens was actually focusing on my target.

However once I learned about MA (And had a camera that could adjust it) I found that I was able to dial in my lens to my body and.... voila!.... images became tack sharp in a way I had never experienced previously.

I completely agree with you on finding what works best for some people doesn't always work best for others. I think there is so much variability between camera bodies and lens copies that we often only find what works best for ourselves with that specific equipment. For me the testing only takes a few minutes, and when my equipment works better I have more fun in the field and spend less time complaining about my gear. But I am not a Pro and only get to shoot when I get out of work... or have the sniffles.....

The most recent issue of Outdoor Photographer has an image in it where George Lepp shot an Eagle with an 800mm f5.6 with stacked 1.4x and 2.0x TC and it looks tack sharp to me!!!

Happy shooting.

Chris I think we're pretty much in agreement. I view calibrating lenses as something a little different than testing but maybe I don't quite understand about further adjustments after calibration.

FWIW, on a 50D withthe 300 f/4L IS USM I noticed an MFA differenceof +4 between no TC and a TCIII 1.4x. Reikan software shooting target at 50xeffective focal length distance to target wide open.

+4 is considered quite acceptable usually. Either way the camera recognized the lens and lens +TC combo as separate entries in the MFA lens table.

As for the peak IQ of each by Reikan's scale:

746 @ MFA 16 combo
922 @ MFA 12 no tcIII

Many of us do care very much about autofocus at f8 with teleconverters/extenders, and validly so. But the point of view that rezzing up a bare lens image (I choose that term deliberately) produces better results than TCs/extenders comes up often enough to merit further discussion -- and perhaps a bit of clarification. I’ve occasionally had the same experience, but only when the TC/extender is a really poor match to the telephoto lens. (There may be other factors as well.) For instance, the Canon EF 2X II on my 800mm is truly awful. Of course, one loses autofocus so one has to focus manually, so that may be a contributing factor. By comparison, rezzing up a 800mm + 1.4X image does give me a sharper result with my 1Ds MkIII. The EF 2X III is a little better but still not great. I haven’t tried the same comparison with the 2X III, but I wouldn’t be surprised if it were worse than a rezzed up 1.4X image also.

On the other hand, if the optics are excellent and the optical match between the prime lens and the TC/extender is excellent then from a theoretical point of view the lens + extender must give a better result, all else being equal. That is, same camera/sensor, same distance to subject, same everything. I ran this question by Roger Clark last night in an email (http://www.ClarkVision.com), and he put the answer in terms of sampling. By magnifying the image, one is, in essence, sampling the image at a higher frequency than with a bare lens. The sampling of more data combined with excellent optics and an excellent match between lens and TC/extender is critical. If the quality assumption breaks down one will find that rezzing up the bare lens may give a better result, as many claim.

To further add to the discussion, as it happens I carefully test almost every lens I buy with an Edmund Scientifics test chart based on the USAF 1951 resolution test pattern. A few months ago I bought the new Canon 300mm II lens. I tested it extensively with my 1Ds MkIII, not only bare but also with the 1.4X III and the 2X III extenders. Because of space limitations in my basement, all three combinations were tested at the same distance to the test target, making these test images an ideal candidate for providing data relative to this discussion. This morning I pulled three frames into Photoshop, at 300mm, 420mm, and 600mm, all at f8. I rezzed up copies of the 300mm image by 1.4X and 2X for comparison to the 420mm and 600mm images.

On this outstanding lens these latest generation high quality Canon extenders easily bested rezzed up bare lens images. For those of you familiar with the USAF 1951 pattern, the 420mm 1.4X extender image resolved three steps better than the rezzed up 300mm image. The 600mm 2X image resolved an even better four steps vs the 300mm image rezzed up. Plus, the rezzed up images were, well, rather ugly – blotchy and pixelated. Not surprising, really, if you think about what was done to them.

Note once again that the answer is not all one way or the other. It does depend of the quality of the optics involved. Finally, as others have noted, it is critical that each combination be AF microadjusted for quality results.

Michael is correct on Canon, however on Nikon, uprezzing beats the 1.7x and 2x on all but the f/2 and f/2.8 primes (200/2, 300/2.8, 400/2.. On the f/4 primes and the 70-200/2.8, uprezzing soundly bests images taken with the 1.7x and 2x. On the 1.4x, the TC beats uprezzing.

As I noted, it all comes down to the optical quality of the combinations involved, and where uprezzing wins it implies that there's room for improvement in optical quality. RE Nikon TCs, since my only experience is with Canon, that's all I can comment on. In any event, experience and testing tells the tale.

I would add that if your long lens technique is not great every increase in magnification/tele-extender will add potential of blurring from vibrations and lens flex unless taken with strobe/flash or very high shutter speed.

A good rule for point and shoots has always been 'optical zoom is better than digital zoom'. Same must apply to SLRs, unless, as Michael notes, the system is lacking somewhat optically. That's been my general take on these things anyway.

Agree so far, but there's a few more variables to take into consideration.
The tc will bring a loss of light, which may result in the iso being raised to keep shutter speed up, or there may be sufficient ambient light to leave the iso where it is.
The loss of light may hamper the autofocus sufficiently that focus isn't bang on, or it may not.
The extra magnification may cause a vibration in the field to be more of an issue, or long lens technique may be good enough anyway.
There's an extra mechanical coupling that might or might not just introduce a slight misalignment.
I find myself moving away from the 2x on my 500/4, and occasionally wishing for a new 600/4...

Bear in mind that most point and shoots have variable apertures, so I think that analogy holds. I have pretty much given up on the canon 2xII because I think that optically limits my system in most cases. The 2xIII might be a different story.

The advantage conferred by a 1.4X TC is considerable compared to any additional benefit provided by a 2X. Still, in favorable circumstances, and for certain combinations, there may be some. The only way to know for sure is testing/experience. (Personally, for info prior to purchase, I like the lens tests at The-Digital-Picture.com.) Whether the improvement is sufficient and whether appropriate situations occur often enough to justify the investment is the key question. Community experience with older 2X TCs implies that it is not so easy to design a 2X TC that is good enough and well enough matched to a maker’s prime lenses to yield a positive benefit compared to a rezzed up 1.4X TC image. Canon’s new 2X III does seem to be a welcome exception. I’m waiting hopefully to see how the 500mm II and 600mm II perform with the 1.4X III and 2X III extenders.