Teleconverter with 100-400 on 50D

[fike]

I have been experimenting with my 1.4x teleconverter on my 50D using my 100-400 f/4.5-5.6 L IS lens.  I am unable to detect in prints that there is any additional detail captured with the teleconverter over the 100-400 alone.  I am trying to figure out if perhaps the teleconverter has become a limiting factor with the additional resolution of the 50D sensor.  Is it possible that with the teleconverter there is so much additional distortion or diffraction that the 50D sensor is out-resolving the lens/teleconverter, even at f/8 or f/11?  Should I give up on using the teleconverter? (I mostly have because focusing is so hard with this combination.)

[pegelli]

I have been experimenting with my 1.4x teleconverter on my 50D using my 100-400 f/4.5-5.6 L IS lens.  I am unable to detect in prints that there is any additional detail captured with the teleconverter over the 100-400 alone.  I am trying to figure out if perhaps the teleconverter has become a limiting factor with the additional resolution of the 50D sensor.  Is it possible that with the teleconverter there is so much additional distortion or diffraction that the 50D sensor is out-resolving the lens/teleconverter, even at f/8 or f/11?  Should I give up on using the teleconverter? (I mostly have because focusing is so hard with this combination.)

Hi fike.

I use a different mount/brand but with a good 1.4x converter (the Kenko Pro 300 DG 1.4x) I'm able to extract more detail on a 12 MP sensor vs. just cropping the image (mainly on a 80-200/2.8 or 200/2.8 since the Kenko pro 300 1.4x physically doesn't fit my 100-400 lens)

By the time I use a generic 2x converter (Tamron MC7) there is little difference between using the converter and just cropping.

I have one lens (a 200/2.8 APO) with a dedicated APO 2x converter that can again extract more detail vs. cropping.

All above observations are at the same effective f-stop

Do you use a branded canon converter or a generic one? From my limited dataset above (I don't have any other converters I can test) one of the conclusions is that not all converters are equally well performing.

Hope this helps

[Ray]

Is it possible that with the teleconverter there is so much additional distortion or diffraction that the 50D sensor is out-resolving the lens/teleconverter, even at f/8 or f/11?  Should I give up on using the teleconverter? (I mostly have because focusing is so hard with this combination.)

Yes. Waste of time using an extender with the 100-400 on the 50D. Extenders unavoidably cause a lens to be a poorer quality lens of longer focal length. You'd probably find a marginal improvement shooting a line chart from a relatively close distance because black lines on a white background are very contrasty. But real world scenes rarely have such high contrast so there's rarely any benefit. However, I would not say no benefit whatsoever under any circumstances.

Consider also, that the 100-400 is noticeably sharper at F8 than at F5.6 (at 400mm). With 1.4x extender you would expect to get sharpest results at F11. You lose autofocussing and you need to use a higher ISO to get the same shutter speed. Because the focal length has been increased, the same shutter speed may not be fast enough, so you might need to use 2 stops higher ISO, hand-held. It's not worth the hassle.

[fike]

Yes. Waste of time using an extender with the 100-400 on the 50D. Extenders unavoidably cause a lens to be a poorer quality lens of longer focal length. You'd probably find a marginal improvement shooting a line chart from a relatively close distance because black lines on a white background are very contrasty. But real world scenes rarely have such high contrast so there's rarely any benefit. However, I would not say no benefit whatsoever under any circumstances.

Consider also, that the 100-400 is noticeably sharper at F8 than at F5.6 (at 400mm). With 1.4x extender you would expect to get sharpest results at F11. You lose autofocussing and you need to use a higher ISO to get the same shutter speed. Because the focal length has been increased, the same shutter speed may not be fast enough, so you might need to use 2 stops higher ISO, hand-held. It's not worth the hassle.

My subjective tests with the camera tripod mounted using f/11 showed no noticeable difference with the canon 1.4x and without.  When I used my old 70-200 f/4 L (not IS), I was able to get a bit more detail on the 30D.  My hypothesis is that with the 50D, the pixel density is so high, that the added resolution (at lower ISOs) renders a higher quality image than a lower pixel density and the teleconverter.

Couple the minimal (if any) improvement with the teleconverter with the loss of autofocus using my setup and I think that the teleconverter will need to come out of my camera bag for now.  I'll swap-in my 50mm f/1.4 in its spot.  That has some pretty practical every-day uses.

[stever]

a bought a 50D in hopes of getting more detail from the 100-400, but got no improvement over the 40D (after micro-adjusting the 50D to +8 which is about what it required on all of my lenses) and returned it.  the 50D was significantly better than the 40D only with the 100 macro (and i assume it would have been with the 50 1.4).  the 50D was also not as sharp as my 5D

i have used the Canon 1.4x with the 100-400 on the 40D and think it gives some benefit, but not sure it's worth the trouble

i did find a 1.4x II to be sharper than my old 1.4x (which is noticeably sharper than the Kenko) on my very sharp 200 f2.8, but doubt that it would make a noticeable difference on the 100-400

[fike]

a bought a 50D in hopes of getting more detail from the 100-400, but got no improvement over the 40D (after micro-adjusting the 50D to +8 which is about what it required on all of my lenses) and returned it.  the 50D was significantly better than the 40D only with the 100 macro (and i assume it would have been with the 50 1.4).  the 50D was also not as sharp as my 5D

i have used the Canon 1.4x with the 100-400 on the 40D and think it gives some benefit, but not sure it's worth the trouble

i did find a 1.4x II to be sharper than my old 1.4x (which is noticeably sharper than the Kenko) on my very sharp 200 f2.8, but doubt that it would make a noticeable difference on the 100-400

From the 30D to the 50D, the detail difference is quite noticeable.  I also micro-adjusted my 100-400 at around +8, though I forget the exact number.  My other lenses didn't need the same adjustment.

[Ray]

a bought a 50D in hopes of getting more detail from the 100-400, but got no improvement over the 40D (after micro-adjusting the 50D to +8 which is about what it required on all of my lenses) and returned it.  the 50D was significantly better than the 40D only with the 100 macro (and i assume it would have been with the 50 1.4).  the 50D was also not as sharp as my 5D

It's always been the case that good lenses are appreciated whatever the film type or sensor pixel density.

In olden days of film, if you wanted sharpest results in circumstances where a slow shutter speed was not a handicap, you needed to use a fine-grain film (equivalent to a high-pixel-density sensor).

Whether or not the film or sensor outresolves the lens is merely academic. Image resolution is always a product of lens resolution and sensor resolution. Increase either one and image resolution is inevitably increased (assuming perfect technique, of course).

The question that needs to be answered is, in practical terms just how significant is any increase in sensor pixel count, or lens resolution?

When using a teleconverter, you are basically taking one step forward and one step backwards. One step forward in terms of extended focal length, and one step backwards in terms of lens quality.

Even with the Canon D60, I rarely saw any improvement in detail using the 1.4x extender (version II) with the 100-400 zoom at 400mm. And that's with extreme pixel peeping at equivalent print sizes on the monitor that are far beyond the maximum print size of my Epson 7600.

Since we have few details of sensor and lens performance (separately) at MTF 50, it's difficult to demonstrate any results by way of calculation without making inaccurate assumptions, but one can see from the Photozone charts below that the 100-400 is a pretty poor performer compared with a top class zoom such as the Canon 70-200 F4 IS.

The reason the 100-400 is so popular is that it is still as good or better than an expensive prime with 2x extender.

However, looking at those Photozone charts, I wonder if the 70-200 F4 IS with 2x extender might get as close as matters to the 100-400. I still think it would be marginally worse (at F4 with 2x extender) than the 100-400 at 400 at F8, but perhaps only at a pixel-peeping level which may not matter even for large prints.

If anyone reading this happens to own a Canon 100-400, 70-200/4 IS, 2x extender and a 50D, don't be shy. Post some comparisons.

[attachment=12142:PZ_100_4...00_F4_IS.jpg]

[stever]

i did some tests using Imatest and found the 70-200 f4 not to be very happy with converters, particualrly the 2x.  the 200 f2.8, on the other hand is fairly happy with converters but with the 2x still couldn't match the center sharpness of the 100-400 (although it was as good in the corners).  the 100-400 and 200 f2.8+2x are capable of making good 13x19 wildlife prints , i don't think the 70-200 +2x will be acceptable even for an 11x14 (although i haven' tried)

i went thru this exercise to find a travel lens substitute for the 70-300 DO - mine is very disappointing from 200 to 300 (if someone has better experience i'd like to hear about it).  so far the 200 has met expectations.  at 200 it's as sharp corner to corner as the 100 macro even at 2.8, still very sharp with 1.4x (better than my 300 f4 which needs to be sold), and useable with 2x.

[dseelig]

I had 2 70-300 do the first I did not like the second after a trip to canon was a very nice lens 1 stop from wide open . Wide open it was amatuer hour one stop down very nice. I sold it abd went from a 70-200 f4 lis 100-400 now back to a 70-200 f 4 lis . I found the 100 -400 very dissapointing at 400 wode open so I sold it. Would lie a light 400 but want is oh well a 300 2.8 with an extender will do for now. David PS those that ahve put a 2 x on a 200 2.8 was it version one of the 2x version 11 thanks David

[Ray]

i did some tests using Imatest and found the 70-200 f4 not to be very happy with converters, particualrly the 2x.

Was this the new 70-200 F4 IS? The new version seems remarkably sharp at 200mm. However, even if it were to come close to the 100-400, one would still sacrifice autofocus with a 2x extender.

This is why the 100-400 is so popular. There's nothing that can match it in terms of weight and convenience. The 70-200 F2.8 doesn't match it (Michael did a comparison some time ago). The 300/2.8 IS no doubt would match it and even exceed it, but that's in a different ball park regarding price and weight and suffers the inconvenience of a lack of zoom.

You say the 200/2.8 with 2x extender has not quite the centre sharpness of the 100-400, so that lens doesn't match it. In addition to the inconvenience of a lack of zoom, it also has no IS. This is not an option for me.

I'm surprised Canon have not upgraded the 100-400 IS by now. If they could give it the same performance from F5.6 to F11 as the 70-200 F4 IS, fully extended, such a lens would sell like hot cakes.

[stever]

yes, new 70-200 IS.  is suspect the same problem as Michael found with the 70-200 f2.8 -- zooms don't like extenders.

from reports and personal experience, there has been significant variation in the 100-400. i have a copy that is nearly as sharp at f5.6 as f8.  i believe that Canon has improved consistency over time, but you'd have to test a lot of lenses to prove this.  It may also be that the lens is somewhat delicate.

i don't consider the 200 f2.8 a substitute for the 100-400, rather a black substitute for the 70-200 f4 (depending on what other lenses you're carrying).  with the high ISO capability of the 40D and 5D, the lack of IS is not a deal breaker for me

I agree that Canon should have a higher priority for upgrading this lens, an 82mm filter and another 1/2 pound would be a welcome tradeoff for better IQ.

[David Sutton]

I was looking for a longer lens than the 70-200 f4, and read that many people found the extender a waste of time on either the 100-400 or the 70-200 compared to just cropping. Furthermore, a lot of copies of the 100-400 were reported as good but not really razor sharp. I opted for a 400mm f5.6 on the grounds that I would probably use a zoom at the longest length anyway, it takes the same size filters as my 24-105, lack of IS won't matter too much for birding as I'll need a fast shutter speed for that, and the 400mm had very fast autofocus (which it does when I can locate the bird fast enough looking through the thing   ). And it is dead sharp when properly supported. I think it should also take a 1.4 extender okay, though probably it would be a waste of time trying for wildlife.  The downside is the long minimum focus distance.
David

[Ray]

I opted for a 400mm f5.6 on the grounds that I would probably use a zoom at the longest length anyway, it takes the same size filters as my 24-105, lack of IS won't matter too much for birding as I'll need a fast shutter speed for that, and the 400mm had very fast autofocus (which it does when I can locate the bird fast enough looking through the thing   ). And it is dead sharp when properly supported. I think it should also take a 1.4 extender okay, though probably it would be a waste of time trying for wildlife.  The downside is the long minimum focus distance.
David

There seems to be quite a bit of QC variability with the 400/F5.6 prime. Michael gave it a good review, comparing it with his copy of the 100-400. The copy I bought a few years ago was actually slightly worse than my 100-400 at 400mm regarding resolution. I didn't compare autofoussing speed. I returned the lens for a refund.

Checking Photozone's assessment of this lens, I see that the copy they tested is also no better than the copy of the 100-400 they tested. At F5.6, the 400 F5.6 prime is slightly sharper at the edges, but not sharper in the centre at any aperture, and sometimes not quite as sharp even, although the differences are propably too small to be of any significance.

I'm not sure that the lack of IS is of little consequence for birding. Birds are usually fairly stationary for brief moments, in between twitching and sudden movements. If the exposure takes place during a twitch, then neither a fast shutter speed nor IS will ensure a sharp result.

The 1/FL guideline for shutter speed to overcome camera shake applies to a 35mm equivalent focal length and an 8"x12" print. If you want razor sharpness and intend printing larger than 8x12", a minimum shutter speed without IS would be 1/2FL(35mm). With the 400/F5.6 prime on a 50D, hand-held, I'd want to use a minimum shutter speed of 1/1200th sec. for all shots.

[attachment=12177:PZ_400_5..._100_400.jpg]

[David Sutton]

Are we talking MTF(resolution)? To quote Photozone "Overall the MTF values are marginally better than the classic EF 300mm f/4 USM L with 1.4x II but quite a bit better than both the EF 300mm f/4 USM L IS with 1.4x II and the Canon EF 100-400mm f/4.5-5.6 USM L IS (@ 400mm).
My copy is approaching the 70-200, though I haven't sat down and pixel peeped.
Though I got it mainly for BIF, I quite agree about the minimum shutter speed hand held. However I find it really too heavy to use this way and get consistent results, and I use it on a monopod. Works fine this way even a little bit slower than 1/400 for stationary objects. The small DOF is more likely to affect sharpness, and I've got this wrong more than once. In low light a flash also works fine if it doesn't upset the subject.
David

[Ray]

Are we talking MTF(resolution)? To quote Photozone "Overall the MTF values are marginally better than the classic EF 300mm f/4 USM L with 1.4x II but quite a bit better than both the EF 300mm f/4 USM L IS with 1.4x II and the Canon EF 100-400mm f/4.5-5.6 USM L IS (@ 400mm).

There's obviously a conflict here. The LW/PH at 50% MTF clearly show that, on balance, the MTF values of for the 400/F5.6 prime are no better than the 100-400 at 400mm. In fact, there's only one figure that is marginally higher for the 400mm prime, and that's border MTF at F5.6. All the other LW/PH figures, down to F11 and including centre and border, are marginally higher for the 100-400 at 400mm.

Photozone mention that they reserve the right to withdraw or replace the MTF results for any lens if it comes to their attention that the results are atypical. There's been some talk that later issues of the 100-400 are better than the first batches that hit the market. I don't know if I've got one of those better copies. I bought my copy in 2002. Michael would have bought his earlier.

I can only suggest that the reason for this conflict of statement and measurement is due to the fact that Photozone have retested the 100-400 and neglected to correct that reference in the 400/5.6 review that the lens is sharper than the 100-400 at 400mm.

[David Sutton]

There's obviously a conflict here. The LW/PH at 50% MTF clearly show that, on balance, the MTF values of for the 400/F5.6 prime are no better than the 100-400 at 400mm. In fact, there's only one figure that is marginally higher for the 400mm prime, and that's border MTF at F5.6. All the other LW/PH figures, down to F11 and including centre and border, are marginally higher for the 100-400 at 400mm.

Photozone mention that they reserve the right to withdraw or replace the MTF results for any lens if it comes to their attention that the results are atypical. There's been some talk that later issues of the 100-400 are better than the first batches that hit the market. I don't know if I've got one of those better copies. I bought my copy in 2002. Michael would have bought his earlier.

I can only suggest that the reason for this conflict of statement and measurement is due to the fact that Photozone have retested the 100-400 and neglected to correct that reference in the 400/5.6 review that the lens is sharper than the 100-400 at 400mm.

Yes, re-read their review and there is a conflict indeed. Bizarre.