Beyond 400mm

[Ray]

Hi Ray,

Thanks for the reply and the info. I have a great copy of the 100-400 which is as sharp as my 400 5.6L at 400 mm f/5.6. My 100-400 even compares well vs. my 500 f4 IS at the center of the frame. I believe my copy of the 100-400 can harness the resolution of the 50D's sensor very well, at least in the central portion of the frame where it matters for my bird photography.  This is because I see a good improvement of captured detail on a per frame basis (as opposed to per pixel basis) when using this lens with a 1.4x TC on a 40D, which use effectively doubles the pixels/bird from the same distance.


Here are links to my user report:

1. 100-400 vs 400 5.6L - http://birdphotoph.proboards.com/index.cgi...&thread=352

2. 100-400 compared to my 500 f4 IS - http://birdphotoph.proboards.com/index.cgi...&thread=353

Romy

Romy,
You are very lucky to get a 100-400 which appears to be so good. For a long time I've been a strong advocate of  thorough MTF lens testing according to international (ISO) standards. I believe that every lens shipped, for serious photography, should come with extensive MTF charts so the customer knows exactly what he/she is buying.

The alternative is 'cherry picking'. This can be very time-wasting, and also requires a co-operative retailer. In general, it's a very unsatisfactory approach, because those who have the relationships which facilitate such cherry picking, benefit, and those who don't, and who pay the same price for their lens, get the rejects.

However, I'm impressed with the thoroughness of your tests. Can I persuade you to buy a 300/F2.8 IS and do comparisons with the 1.4x and 2x extenders?  

Fike makes a very valid point, that the 300/1.4x combination gives you an 420/F4 lens which has faster and more accurate autofocussing than the 100-400 at 400mm. I can only assume, because I can find absolutely no comparisons on the internet between the 300/1.4x and the 100-400 at 400mm, that the differences in absolute resolution are too small to be an issue.

It's a fact of photography that sharp results are often a combination of accurate focussing and high shutter speed, rather than some slight instrinsic resolution advatage of one lens over another.

I know my 100-400 is not as sharp at F5.6 as it is at F8, but the difference is very marginal. The difference in DoF between F5.6 and F8 is far greater. This is where practical considerations come into play and often reveal the irrelevance of pixel-peeping advantages. If focussing is not 'spot on' then a bird's eye at F5.6 might look disturbingly less sharp than at F8. However, it's not always possible to get focussing spot on. Using F5.6, it's more critical to get accurate focussing. Using F2.8 must be a nightmare.

The 50D with its very high resolution LiveView LCD screen, facilitates tremendously accurate manual focussing. At 10x magnification, a 400mm lens becomes a 4,000 mm lens (compared with an actual 4,000mm lens with no magnifiaction on the same LCD screen.)  Of course we don't have any real 4,000mm lenses for DSLRs. The Canon 1200/F5.6 costs as much as a house.

There's simply no excuse for misfocussing with the 50D when manual foussing is practicable.

Out of interest, I went back to my 5D/50D comparisons, and selected images comparing the 50D at F5.6 with the 50D at F8, at 400mm. Whilst the the 100-400 is very marginal sharper, in the plane of focus, at F8, the most significant increase in sharpness results from the increased DoF that F8 offers.

Below, I've taken the same scene at F5.6 and at F8, and cropped a vertical strip from the centre to show the variation in DoF. In my opinion, the resolution differences that occur due to DoF are far more significant than resolution differences at the plane of focus. This scene was taken in the late evening and is across a river, from one bank to the other. Atmospheric distortion is minimal. (It's fortunate that my neighbours are not given to prancing around in the nude, otherwise they might be quite concerned about a guy across the river with a telephoto lens   ).

The plane of focus is shown in the lower half of the middle image. Above that and below, you can see the loss of reolution in the F5.6 shot due to DoF. At the plane of focus, we're nitpicking.

[attachment=12738:0013_50D...ll_scene.jpg]  [attachment=12739:top_of_c...5_6_v_F8.jpg]  [attachment=12740:middle_o...5_6_v_F8.jpg]  [attachment=12741:bottom_o...5_6_v_F8.jpg]

[fike]

I have also found that my keeper rate with the 100-400 is much better after calibrating it with the 50D's auto-focus micro adjustment.  I adjusted about 5 units from center.  Makes a big difference for bird work.

[Ray]

I have also found that my keeper rate with the 100-400 is much better after calibrating it with the 50D's auto-focus micro adjustment.  I adjusted about 5 units from center.  Makes a big difference for bird work.

This is a feature I've yet to try. Been busy doing other things. But the micro-adjustment was definitely a major consideration for me when deciding to upgrade from a 40D to 50D.

However, assuming perfect focus, it still appears to be the case that the 100-400 is not as sharp as it could be at all 3 apertures of F5.6, F8 and F11. Whilst Romy's tests suggest that his copy of the 100-400 really is as sharp as can be expected and on a par with the 400/5.6 prime, we can't be sure in absolute terms just how sharp it is. The 400/f5.6 and 500/f4 are lenses that also vary in quality.

When I bought a 400/5.6 some years ago, I was surprised to find that it wasn't any sharper than, and sometimes not quite as sharp as, my 100-400 at all apertures. I returned the lens, of course. In such circumstances, it wasn't clear to me whether my copy of the 100-400 was above average and the copy of the 400/5.6 I tested just average, or whether my copy of the 100-400 was just typical and the 400/4.6 a lemon. Perhaps they were both average. I notice that the copies of these lenses that Photozone tested with a 350D appear to be about equally matched at 400mm from F5.6 to F11. The 400/5.6 prime appears to be slightly sharper at the edges of the cropped format frame at F5.6. One might assume that on a FF 35mm DSLR, the better edge performance at F5.6 would be more obvious. However, the 100-400 at both F5.6 and F8 appears to be very marginally sharper in the centre, in the Photozone tests.

Such small differences are irrelevant. For all practical purpose the 400/5.6 prime and the 100-400 zoom at 400mm are about equal in the PZ tests, but that doesn't mean that the 100-400 is a sharp lens. Of course, it goes without saying that the IS feature of the 100-400 can have great practical benefit. I would bother to use a 400/5.6 in place of the 100-400 zoom unless the prime was substantially sharper.

[attachment=12772:PZ_400_5..._100_400.jpg]

[fike]

...

The 50D with its very high resolution LiveView LCD screen, facilitates tremendously accurate manual focussing. At 10x magnification, a 400mm lens becomes a 4,000 mm lens (compared with an actual 4,000mm lens with no magnifiaction on the same LCD screen.)  Of course we don't have any real 4,000mm lenses for DSLRs. The Canon 1200/F5.6 costs as much as a house.

There's simply no excuse for misfocussing with the 50D when manual foussing is practicable.

...

I haven't found this to be true.  At 400mm and 10x magnification, even with image stabilization on, the smallest vibration of the camera makes focusing very hard because it moves the entire viewable area.  I kind of gave up using live preview with manual focus at those long focal lengths.  It might be better if I used a heavier tripod, but as it is I use a carbon traveller tripod--fairly steady, but not that steady.

[BJL]

I haven't found this to be true.  At 400mm and 10x magnification, even with image stabilization on, the smallest vibration of the camera makes focusing very hard because it moves the entire viewable area.

What if you back off to 5x, for an image still larger and better resolved than the corresponding part of the OVF image? 5x displays a roughly 950x640 crop, and the 50D LCD is really about VGA resolution of 640x480 (with dodgy "x3" pixel counting), so that is close to full sensor resolution anyway, and way ahead of the resolution of the image that the OVF gets off the ground glass. The step up from 5x to 10x goes well beyond "100%" and adds little detail to the LCD preview image.

[Ray]

I haven't found this to be true.  At 400mm and 10x magnification, even with image stabilization on, the smallest vibration of the camera makes focusing very hard because it moves the entire viewable area.  I kind of gave up using live preview with manual focus at those long focal lengths.  It might be better if I used a heavier tripod, but as it is I use a carbon traveller tripod--fairly steady, but not that steady.

Fike,
My travel tripod is the Manfrotto Carbon Fibre 190CXPRO4 with 460MG head. With the 100-400 it's amazing to see just how much that 10x image on the camera's LCD screen quivers and wobbles on that tripod with the slightest breeze, so slight in fact one can hardly notice that there is a breeze.

But it's equally amazing how steady that image becomes when IS is enabled. With the older versions of IS, one is advised not to use IS when the lens is mounted on a tripod. However, movement is movement. That LiveView is an excellent device for letting one know when IS should be enabled, if the camera is tripod mounted.

Manual focussing with LiveView when IS is enabled is no problem, when camera is on a tripod. However, it's still a bit awkward when the lens is hand-held.

[Ray]

What if you back off to 5x, for an image still larger and better resolved than the corresponding part of the OVF image? 5x displays a roughly 950x640 crop, and the 50D LCD is really about VGA resolution of 640x480 (with dodgy "x3" pixel counting), so that is close to full sensor resolution anyway, and way ahead of the resolution of the image that the OVF gets off the ground glass. The step up from 5x to 10x goes well beyond "100%" and adds little detail to the LCD preview image.

BJL,
For really critical, hairsplittingly accurate focussing, it's surprising how much easier it is to focus with the higher resolution screen of the 50D at 10X, compared with the 230,000 dot screen of the 40D at 10X.

Recently, when comparing resolution differences between the 50D and 40D, photographing a banknote using the Canon 50/1.4 at various apertures, I was concerned that at wide apertures, the results from the 50D would show more of an advantage than was warranted because more accurate focussing was possible. With the 40D, I was never quite sure if focussing was really spot on, even at 10x magnification. After some trial and error, I decided the presence of faint color aliasing artifacts on the LiveView screen, due to a particular spacing of lines on the banknote, was a more certain indication of dead-accurate focussing.

[cmox]

Fike,
My travel tripod is the Manfrotto Carbon Fibre 190CXPRO4 with 460MG head. With the 100-400 it's amazing to see just how much that 10x image on the camera's LCD screen quivers and wobbles on that tripod with the slightest breeze, so slight in fact one can hardly notice that there is a breeze.

But it's equally amazing how steady that image becomes when IS is enabled. With the older versions of IS, one is advised not to use IS when the lens is mounted on a tripod. However, movement is movement. That LiveView is an excellent device for letting one know when IS should be enabled, if the camera is tripod mounted.

Manual focussing with LiveView when IS is enabled is no problem, when camera is on a tripod. However, it's still a bit awkward when the lens is hand-held.

After some unpleasant and very costly experiences with aluminium, carbon fibre and basalt tripods I followed the advice of some nature photographers. Many of them here in Germany swear by Berlebach tripods. They are made of ashwood and dampen vibrations. Price and weight are in between aluminium and carbon fibre. Stability outclasses even Gitzo carbon tripods that cost a lot more.

The one I use most is this one here:
http://www.berlebach.de/?bereich=details&a...sprache=english

It is more than stable enough for the 560mm in bright daylight without the converter at 1/125.

For the 560mm plus 2x converter I prefer to set the camera and lens in concrete... I have an older version of this one plus this one borrowed from a friend:
http://www.berlebach.de/?bereich=details&a...sprache=english

One of them is attached directly to the lens, the other directly to the camera, without a quick release system. If my Telyt were an "IS" lens I could probably leave one of those at home :-)

They also have a "Conan the Barbarian" edition for the tough guys:

http://www.berlebach.de/?bereich=details&a...sprache=english  

[Ray]

After some unpleasant and very costly experiences with aluminium, carbon fibre and basalt tripods I followed the advice of some nature photographers. Many of them here in Germany swear by Berlebach tripods. They are made of ashwood and dampen vibrations. Price and weight are in between aluminium and carbon fibre. Stability outclasses even Gitzo carbon tripods that cost a lot more.

The one I use most is this one here:
http://www.berlebach.de/?bereich=details&a...sprache=english

It is more than stable enough for the 560mm in bright daylight without the converter at 1/125.

For the 560mm plus 2x converter I prefer to set the camera and lens in concrete... I have an older version of this one plus this one borrowed from a friend:
http://www.berlebach.de/?bereich=details&a...sprache=english

One of them is attached directly to the lens, the other directly to the camera, without a quick release system. If my Telyt were an "IS" lens I could probably leave one of those at home :-)

They also have a "Conan the Barbarian" edition for the tough guys:

http://www.berlebach.de/?bereich=details&a...sprache=english  

I'm afraid my enthusiasm for photography does not extend to humping around an 11Kg tripod, or even a 5Kg tripod. I want a 150-400/F5.6 IS lens which is as sharp at F5.6, F8 and F11, as the Canon 70-200/F4 IS is at F5.6, F8 and F11. That's not too much to ask, is it?

I'm prepared to pay a reasonable premium.

[BJL]

For really critical, hairsplittingly accurate focussing, it's surprising how much easier it is to focus with the higher resolution screen of the 50D at 10X, compared with the 230,000 dot screen of the 40D at 10X.

No surprise: at 10x, the 50D is giving about a 480x360 crop displayed at 640x480 vs the 40D's roughly 390x290 crop displayed at 320x240; the lower res. of the 40D LCD itself would surely give the 50D a visible advantage. But that is a different comparison that my 5x vs 10x.

P. S. Nice to see you acknowledge elsewhere in this thread that there is such a thing as equipment that is too heavy to justify the IQ improvements that it can bring! In your case, 11Kg tripods; in my case, lenses with effective aperture diameters bigger than about the 70mm of a 200/2.8 or 300/4 [edit: or a 400/5.6]

[cmox]

I'm afraid my enthusiasm for photography does not extend to humping around an 11Kg tripod, or even a 5Kg tripod. I want a 150-400/F5.6 IS lens which is as sharp at F5.6, F8 and F11, as the Canon 70-200/F4 IS is at F5.6, F8 and F11. That's not too much to ask, is it?

I'm prepared to pay a reasonable premium.

My enthusiasm for photography does not extend to buying an IS lens with a focal length beyond 400mm... that's the sonic barrier where prices become really nasty. Sure, I would love to own a 4/600 lens. But as I do not use such a beast every day, I prefer to carry some more weight. In most cases it is that 3 Kilo tripod which is more than sufficient for my purposes in 80% of the cases. If I had a 150-400/F5.6 IS lens I would defintely buy a lighter tripod.

[Ray]

No surprise: at 10x, the 50D is giving about a 480x360 crop displayed at 640x480 vs the 40D's roughly 390x290 crop displayed at 320x240; the lower res. of the 40D LCD itself would surely give the 50D a visible advantage. But that is a different comparison that my 5x vs 10x.

BJL,
Perhaps you can clear up a point I'm not sure about. Is the anti-shake sensor in the Olympus 4/3rds cameras continuously in operation during focussing, or does it spring into action only at the time the shutter is fully pressed?

If it's not continuously active during focussing, then I can understand that you would often prefer to use the 5x magnification instead of 10x. That would also explain why Olympus have not provided a 960,000 dot LCD screen on their latest models.

[BJL]

Is the anti-shake sensor in the Olympus 4/3rds cameras continuously in operation during focussing, or does it spring into action only at the time the shutter is fully pressed?

I do not have an IS body yet, but I am rather sure that stabilization is continuous in Live View mode. My comment about 5x vs 10x was mostly a question, about whether zooming in on 4% if the image area might work better than zooming in on 1% of it, when working without stabilization.

That would also explain why Olympus have not provided a 960,000 dot LCD screen on their latest models.

No, the reason for that is simple: the new VGA resolution (640x280x3) LCD is a 3" model, whereas the FourThirds bodies so far use smaller 2.7" screens. An articulated 3" LCD probably fits poorly with the 4/3 DSLR body size targets, at least for models like the E-620. Likewise, the Nikon D60X, hotly rumored to be announced on April 14, supposedly stays with the 2.5" LCD of the D60 and D40 for its articulated LCD, hence staying at about 240,000 dots.


P. S. AFAIK, the 3" LCD's are all this one from Samsung:
http://www.digitalcamerainfo.com/content/S...ch-VGA-LCD-.htm
http://www.gizmag.com/go/5999/

[Ray]

I do not have an IS body yet, but I am rather sure that stabilization is continuous in Live View mode. My comment about 5x vs 10x was mostly a question, about whether zooming in on 4% if the image area might work better than zooming in on 1% of it, when working without stabilization.

BJL,
I would have thought so too. However, the mysterious lack of Live View on the Sony A900 got me wondering about this issue. The fact is, when looking through the optical viewfinder, sensor shift can have no effect on the viewed image during focussing. It therefore does not make sense to have the anti-shake sensor active for no good reason, which is why I thought that perhaps the intital anti-shake designs allowed a very brief anti-shake movement of sensor just prior to, and during, the flipping of the mirror. There's no point in making something more durable than it needs to be, as Henry Ford would agree, if he were still alive.

Creating a full frame anti-shake sensor appears to have been something of a technological challenge for Sony. Perhaps the technological challenge of creating a FF sensor which could shake continuously during manual focus in Live View mode, was too much, or too expensive. Hence, no Live View for the A900.

If image wobble is exaggerated in Live View with magnification, then 10X magnification will exaggerate it to a greater degree than 5x magnification or no magnification. The choice is really about how critical the focussing needs to be. With the 100-400 and 1.4x extender, I see no advantage in trying to focus in Live View mode when the camera is hand-held. The movements are too great for IS to stabilise the image. I prefer to manually focus through the optical viewfinder with such a lens, with IS enabled. Shorter focal lengths are more manageable in Live View mode.

Out of curiosity, and because there was a clear moon last night and little breeze, I mounted my 50D with 100-400 and 1.4x extender on my main tripod, the rather basic but too-heavy-to-travel (2.55kg) Manfrotto 141RC, to check out image stability in Live View mode.

One curious effect of such high magnification on the LCD screen, was the visible real time movement of the moon across the LCD screen. I had to keep re-adjusting my tripod to get the moon in the centre of the screen, during the few minutes of the experiment. I could actually see the movement of the moon from right to left of the screen.

With tripod firmly on the concrete floor of the verandah, the image was rock-steady (apart from this gradual drift), even at 10x magnification. At this degree of magnification, one sees only a portion of the moon on the LCD screen. Quite amazing! If only my lens were better! Unfortunately, as soon as one touches the focussing ring, the wobble begins.

[samirkharusi]

Just general comments on an interesting thread:

Nobody seems to worry about "seeing", the shimmering unsteadiness of the atmosphere that made astronomers launch the Hubble Space Telescope. An irrelevance? Definitely not, if you are using a 400mm lens on a current Canon DSLR with the 5micron or finer pixel pitches. Any astrophotographer will notice it on the very first pixel peep! At normal daytime temperatures seeing is a much worse threat than misfocusing if you are shooting anything at more than a 100m away with 400mm or longer teles. Just shoot outside your window at dawn and at noon and pixel peep. Even in the dead of night, seeing at zenith is between 2 and 3 arc-seconds FWHM (Full Width at Half Max). During the daytime it probably shoots up a few to several times larger. With a 400mm focal length and a 50D you are sampling that shimmering stew at 2.5arc-sec per pixel and no amount of IS nor tripod rigidity, nor super accurate focusing will alleviate it. And for Heaven's sake, you are shooting horizontally through that stew! Welcome to the new world of pixel peeping.

The theoretical best that Canon lenses can do may be seen from their published MTF curves. I put them up on my website for easy reference, including for when they are used with 1.4x and 2x extenders:
http://www.pbase.com/samirkharusi/canon_mtf_curves
It looks to me that it is wishful thinking to expect the average 100-400mm to equal the 400mm/5.6 when used wide open at 400mm. One would have to be super-lucky with his sample of the zoom and super-unlucky with his sample of the prime, or else be examining only the on-axis performance. Owning and having owned more than 10 Canon primes and zooms (all the way to the 600mm), and having compared a couple of them to a couple of premium astro-imaging scopes, I have come to the conclusion that these MTF data are actually excellent indicators of the lens' imaging abilities, despite all my purchases having been random, not cherry picks.

Ah, I also have a small gallery of astrophotos (where every pixel is stuffed carefully with photons over many hours, drip by drip, and then massaged lovingly over many days and nights in postprocessing, to squeeze any peformance the camera or the lens might be capable of!) displaying a whole range of focal lengths from 14mm to 4000mm here:
http://www.pbase.com/samirkharusi/focal_lengths
Sorry, this stuff is not really for zooms, but the occasional extender was deployed.

[Ray]

Nobody seems to worry about "seeing", the shimmering unsteadiness of the atmosphere that made astronomers launch the Hubble Space Telescope. An irrelevance? Definitely not, if you are using a 400mm lens on a current Canon DSLR with the 5micron or finer pixel pitches. Any astrophotographer will notice it on the very first pixel peep! At normal daytime temperatures seeing is a much worse threat than misfocusing if you are shooting anything at more than a 100m away with 400mm or longer teles. Just shoot outside your window at dawn and at noon and pixel peep.

We're aware of this problem, Samir   . The second post in this thread makes reference to the problem. It was very apparent when taking my test shots across the river, just a couple of hundred metres away. Any time other than early morning or late afternoon was pointless. Heat shimmering, dust and atmospheric pollution are a great barrier to getting sharp images with a long telephoto lens.

I presume that an indication of the clarity of the night sky would be the number of stars one can see. Often, in the countryside where I am, the night sky is ablaze with stars which one simply doesn't see in the city because of the lights and the pollution.

It looks to me that it is wishful thinking to expect the average 100-400mm to equal the 400mm/5.6 when used wide open at 400mm.

For me the wishful thinking that the 400/5.6 prime would be sharper than my 100-400 at 400mm at F5.6 and at any aperture were dashed when I compared them. Photozone also found these two lenses to be about equal at all apertures.

[BJL]

Even in the dead of night, seeing at zenith is between 2 and 3 arc-seconds FWHM (Full Width at Half Max). During the daytime it probably shoots up a few to several times larger. With a 400mm focal length and a 50D you are sampling that shimmering stew at 2.5arc-sec per pixel ...

Does this suggest that once the focal length is about 100,000 times the pixel spacing, seeing will limit resolution even with cool night air?  And so probably "50,000 times pixel spacing" or less during the day? (And less for Ray in the Australian tropics!?)

If so, it seems that focal lengths beyond "100,000x" will rarely improve resolution of the subject, and one might as well just stay at that maximum focal length (really a maximum angular resolution) and crop when a narrower field of view is wanted?

I would like to think so, since my future cameras will almost certainly all have a pixel size under 5 microns, which would make 400mm the most that I need worry about being able to afford or carry, with a 400/5.6 sitting at my maximum lens size target.

[samirkharusi]

Does this suggest that once the focal length is about 100,000 times the pixel spacing, seeing will limit resolution even with cool night air?  And so probably "50,000 times pixel spacing" or less during the day? (And less for Ray in the Australian tropics!?)

If so, it seems that focal lengths beyond "100,000x" will rarely improve resolution of the subject, and one might as well just stay at that maximum focal length (really a maximum angular resolution) and crop when a narrower field of view is wanted?

I would like to think so, since my future cameras will almost certainly all have a pixel size under 5 microns, which would make 400mm the most that I need worry about being able to afford or carry, with a 400/5.6 sitting at my maximum lens size target.

Not too long ago, when pixels were still very expensive, astrophotographers would very carefully match their focal lengths to their seeing, using Nyquist Critical Sampling, i.e. 2 pixels for FWHM. For long-exposures, seeing is indeed the limiter. For short exposures (say, under a tenth of a second) seeing does not blur the image so much as it distorts the image. Nobody wants his straight walls to come out wavy. This is, in my opinion, a bigger headache for daytime exposures than for astro planetary imaging. Planetary imaging is normally done using video, with the distorted frames later stacked. I have not yet seen this technique used for daytime photography of static subjects, but in theory it could enable much longer focal lengths. It works very well on the Moon (which is a sunlit landscape). The relevant FWHM in this usage becomes the diffraction limit of the lens aperture, not the seeing FWHM. I.e. we will then be chasing large aperture optics, focal length is easily adjustable using tele-extenders (Barlows in astro usage).

The highest resolution planetary images, still limited by our atmosphere, can be obtained with a C14, a fairly common 14" aperture amateur scope. Larger apertures have so far not proven capable of ever higher resolution from ground level, beneath the atmosphere, with current amateur image processing software. For a 5 micron pixel pitch we can achieve diffraction-limit critical sampling with a 7,000mm focal length on a C14. So I would expect that the ultimate daytime, long tele, would be a C14 at 4000mm to 7000mm focal length, shooting video (not stills) and later stacking the frames. For planetary imaging I typically shoot 2000 frames, grade and parse them down to the best 15%, and stack those. I get planetary images that look far better and more detailed than if I look visually through the same scope. Only one catch, a C14 is large and heavy, not something to carry to the beach, but could still be useful to somebody doing technology spying on static subjects. Here is how small a Canon 1Ds looks like on a C14:
http://www.pbase.com/samirkharusi/image/37431993
OK, the front shiny half is only the lenshood (dew shield)

[BJL]

For short exposures (say, under a tenth of a second) seeing does not blur the image so much as it distorts the image. Nobody wants his straight walls to come out wavy.

That is better news for me then. My super-telephoto subjects are usually animals and other nature scenes, not industrial espionage, so if the dominant effect of seeing is a little bit of waviness, I can live with that.

[samirkharusi]

That is better news for me then. My super-telephoto subjects are usually animals and other nature scenes, not industrial espionage, so if the dominant effect of seeing is a little bit of waviness, I can live with that.

Yes. And for static subjects you can still use stacking to get rid of it. I have an image of Saturn, taken with a standard Canon 600mm/4.0L IS using the video stack technique, and a 1.4x extender coupled to a 5x extender to yield a focal length of 4200mm, through 27 lens elements! I used a $100 webcam with its lens (and UV/IR blocker) removed. It is possible that one could do better using an IR blocker, but I simply forgot to use one; force of habit since I normally use the C14 mirror scope for planetary videos and that yields better images with IR included (allowing faster frame rates and possibly more steady seeing in IR):
http://www.samirkharusi.net/televue_canon.html
These days with the video-capable new DSLRs, such exercises would be simpler to try out. And the DSLR sensors are way better than in a 5-year old webcam...