Another question about the L 100-400 zoom....

According to the manual:

Mode 1: corrects vibrations in all directions. It is mainly effective for shooting still subjests.

Mode 2: Corrects vertical camera shake during following shots in a horizontal direction, and corrects horizontal camera shake during following shots in a vertical direction.

:)
 
jerry12953 said:
Thanks Ian....

Does anyone have a general "modus operandi" for using this lens in various situations?
Click to expand...

I don't do any panning shots so mine pretty much stays on Mode 1 all the time - I think it's only gone on mode 2 when I first had it for a quick test.

I think that mode 2 is just for panning shots and mode 1 for everything else, I could be wrong but that's how I use the lens anyway!
 
blacksheep said:
I don't do any panning shots so mine pretty much stays on Mode 1 all the time - I think it's only gone on mode 2 when I first had it for a quick test.

I think that mode 2 is just for panning shots and mode 1 for everything else, I could be wrong but that's how I use the lens anyway!
Click to expand...


Snap, don't think I have ever used mine in mode 2:D
 
jerry12953 said:
Thanks Ian....

Does anyone have a general "modus operandi" for using this lens in various situations?
Click to expand...

Typically I will disable IS completely if shooting at shutter speeds at or over 1/800, and in particular when following a subject of unpredictable motion.

When I do enable IS I do pay close attention to the mode I use. Using the wrong mode can be as bad, or worse, than turning it off completely.

If I am making slow, controlled movements, rather than fast tracking pans I may well use Mode 1, but the thing to watch out for is when the IS tries to fight against the movements you are making, so slow, steady movements which need a little shake control, such as when aiming at a subject approaching head on, will normally benefit from IS Mode 1.

If you are doing pans then clearly Mode 2 is the right tool for the job, but the lens must decide whether you are panning horizontally OR vertically and then stabilise shake in the other plane. If you are making a diagonal pan or random panning movements, rather than exactly horizontal or vertical pans, then the IS will fight you so it would be better to turn it off.

It would be worth reading what Thom Hogan has to say about using Nikon's VR system. I'm sure that similar arguments apply well to Canon's IS system....

http://www.bythom.com/nikon-vr.htm
 
tdodd said:
Typically I will disable IS completely if shooting at shutter speeds at or over 1/800, and in particular when following a subject of unpredictable motion.

When I do enable IS I do pay close attention to the mode I use. Using the wrong mode can be as bad, or worse, than turning it off completely.

If I am making slow, controlled movements, rather than fast tracking pans I may well use Mode 1, but the thing to watch out for is when the IS tries to fight against the movements you are making, so slow, steady movements which need a little shake control, such as when aiming at a subject approaching head on, will normally benefit from IS Mode 1.

If you are doing pans then clearly Mode 2 is the right tool for the job, but the lens must decide whether you are panning horizontally OR vertically and then stabilise shake in the other plane. If you are making a diagonal pan or random panning movements, rather than exactly horizontal or vertical pans, then the IS will fight you so it would be better to turn it off.

It would be worth reading what Thom Hogan has to say about using Nikon's VR system. I'm sure that similar arguments apply well to Canon's IS system....

http://www.bythom.com/nikon-vr.htm
Click to expand...

Interesting link Tim, thanks :) I'd not seen that before, and from a respected source, but I've heard about this alleged VR problem at higher shutter speeds. I say alleged, because I've yet to see any conclusive evidence of it. And that is the single most significant ommission from Thom's article; I understand the theory, but what does it actually look like in practise? If you can indeed see it, then why not post a picture or two and if it is anything meaningful, then the scene is set for a beneficial for-and-against debate. The flip side is that we know very well of the benefits of VR.

I get the impression he's on a bit of a mission when he talks about the shift in framing you can get with VR enabled. Well yes, again in theory, but is that in any way remotely significant?

Then he's talking about Nikon's VR, and the critical sampling frequency of 1000Hz. What does Canon's IS work at? What's the difference between the IS system in Jerry's 100-400L (one of the very first incarnations) and later versions like the 4-stops IS system in my 70-200L 4 IS? Then again, cheaper kit lenses like the 18-55 and 55-250 appear to have another system again.

He says turn VR off above 1/500sec, which, in the absence of anything other than theoretical hypothesis and frankly anecdotal evidence, seems rather more irresponsible than Nikon saying leave it on. He makes no qualification for focal length, merely referencing the sampling frequency re nyquist. Despite quoting some seemingly real life situations, his recommendations don't seem very well grounded in reality.

I've tried to produce evidence of problems with IS at high shutter speeds. Unfortunately the only long lens I have is the 70-200L as I sold the 100-400L a little while ago. On more than one occasion I've tried hard to find evidence of the IS clipping sharpness but I can't see anything reliably, despite going through a range of shutter speeds and different movements with IS on and off. Sometimes I think the non-IS image is better (very, very fractionally, almost invisibly) but then I can't replicate it and for sure the evidence of shake in other images when it is turned off is all too clear to see.

Be interested in your thoughts :)
 
From my own experience shooting BIF for the last three years with my 100-400 I can say that I have witnessed some very odd effects in the viewfinder when accidentally using the wrong IS mode. I can also say that there have been times when I have experienced very poor IQ with absolutely no logical reason to explain it. I'm talking in particular about softness throughout the frame or a sort of halo around bright, high contrast subjects. This has happened with different bodies - 50D, 1D3, and in circumstances where I have absolutely not saturated the sensor and leaked charge from pixel to pixel. Whenever I see such problems and bench test the lens, using a tripod, there is no fault to be found with the AF performance or the IQ. Looking back, I think the most likely explanation is that I had been using IS when it really would have been better not to. Unfortunately the EXIF data does not record whether or not IS was used, so I can never be sure.

Although I have never used one, one of the finest birding lenses for those not wishing to spend over £1,000 is reputed to be the 400/5.6L. It is sharp wide open and focuses fast. It does not have IS and that does not appear to stand in the way of achieving excellent results. With that in mind, and wondering whether adding the lens to my collection would be the solution to my spurious IQ problems, this year I have been disabling IS for pretty much all my BIF shooting. I would say on reflection that my keeper rate has increased.

So, all very anecdotal and unscientific, but I really believe that IS is best left off for tracking random motion, motion which is not complementary to the abilities of the IS system (diagonal pans, for example), or when shooting at high shutter speeds as one might when using the unstabilised 400L. Now, is there any sort of rational explanation for my own findings? I think it might well be contained in Thom Hogan's article. Maybe newer IS systems are better than the 100-400, but given that this thread is about the 100-400 and its first generation IS it does not seem unreasonable to me that it is not state of the art and may have limitations at high shutter speeds.


Here is a recent example of a hand held panning shot at 400mm and 1/800, with IS off and shown at 100% with no sharpening adjustments. Since it is a simple horizontal pan I guess IS mode 2 might have been OK, but what would happen if the plane started to climb? I'd be screwed. By the look of things I don't think I needed IS anyway. Others may disagree.

20100707_173058_.JPG



How about this one - which panning mode should I have used? Well none were appropriate. Maybe viewed at 100% my panning is not quite perfect, but I don't think IS would have done me any favours at all..

20100707_174242_.JPG



Sure, both shots need some edits, but they don't need IS, IMHO. :)


Now, at the other end of the spectrum, when shooting at low shutter speeds and long focal lengths I'll happily take the best IS on offer. Here's a hand held shot at 400mm and 1/60. Of course, viewed at 100% this is testing the IS massively beyond the 2 stop performance expected of it. Of course, it didn't quite rise to the challenge, but it hasn't done a bad job, especially considering I was squeezed into an awkward shooting position, hemmed in on both sides by people while seated in a crowded grandstand.

20100707_175102_.JPG



Now for a change of subject, this was shot at only 105mm and 1/160, but IS was turned off since the bikes were moving in all sorts of directions and IS would have screwed things up. Even if it had worked, I'm not sure it would have helped.

20100707_180056_.JPG


So, like Thom Hogan says, turn it on when you need it. Turn it off when you don't.
 
Okaaaayyyyy, we seem to be getting into technical realms here which are well beyond my ability to understand......:lol:

But thanks for the general advice Tim, which once again is invaluable.

There's just one more question - to IS or not on a tripod?

Otherwise I'll let you two technical heavyweights carry on without me. Don't want to be a gooseberry....;)
 
With the 100-400 the advice when using a tripod is simple. TURN IS OFF.

Well, that's what Canon says. However, even when using a tripod you can get undesirable vibration, such as that caused by wind, so I think there may be situations where IS may offer an advantage when using a tripod. There is probably a narrow range of shutter speeds for which using IS may make sense - probably between around 1/10 and 1/400, but only if you are seeing visible shake in the system. Slower than 1/10 and you may get drifting of the IS system, which will smear the image across the sensor. Faster than 1/400 and, with the aid of a tripod, I think we're getting into that territory where the value of IS starts to become moot.

For an easy life, follow Canon's advice. If you want to maximise your results you may need to experiment. If you're shooting indoors then you really shouldn't need IS on a tripod, but don't forget mirror lockup and remote or timed release and all that good stuff. :)
 
tdodd said:
From my own experience...

<snip>
Click to expand...

Thanks Tim. I appreciate your time.

I think it's fair to conclude that the venerable 100-400L may be an exception and certainly there are situations that will fox any IS/VR system and if you get it fighting against you there will be problems. I certainly got some unexpected drifting effects at very long speeds with that lens, but they provide a two position switch for these things afterall, plus a third position - off.

That isn't the main thrust of Thom's article though. He's talking about VR's sampling frequency introducing very slight but permanent vibrations that at higher shutter speeds are more intrusive than the camera movements they are supposed to be cancelling out. But he stops short of showing us what it actually looks like, and I can't replicate it :thinking:

A bit off topic I guess ;)
 
tdodd said:
With the 100-400 the advice when using a tripod is simple. TURN IS OFF.

Well, that's what Canon says. However, even when using a tripod you can get undesirable vibration, such as that caused by wind, so I think there may be situations where IS may offer an advantage when using a tripod. There is probably a narrow range of shutter speeds for which using IS may make sense - probably between around 1/10 and 1/400, but only if you are seeing visible shake in the system. Slower than 1/10 and you may get drifting of the IS system, which will smear the image across the sensor. Faster than 1/400 and, with the aid of a tripod, I think we're getting into that territory where the value of IS starts to become moot.

For an easy life, follow Canon's advice. If you want to maximise your results you may need to experiment. If you're shooting indoors then you really shouldn't need IS on a tripod, but don't forget mirror lockup and remote or timed release and all that good stuff. :)
Click to expand...

:thumbs:

So, if possible, try with and without IS and see what happens......
 
For Hoppy.....

I just mounted my 100-400 onto my 7D and went to fire off 20 sequential shots each with IS and without IS. Nothing fancy, just one static target with plenty of contrast and hard edges by which to judge sharpness and shake. The light is a bit poo but I managed to get 1/1600 at 400 ISO. I focused just once, using One Shot, and then rattled off 20 shots in a slow burst, changed the IS setting and repeated the drill. Focus was targetted at the letter "I", which is why I have lined them up side by side for comparison. I recommend ignoring the other letters as the surface was curved and slightly angled.

I would say that unless pixel peeping, and even then, there is very little to choose between the two sets of results. Here are a couple of corresponding pairs of shots for comparison. I won't say whether the shots with the IS are on the left or the right, but if you can spot a difference in sharpness then feel free to comment.

20100707_183848_.JPG


20100707_184030_.JPG



Of course, this is an easy test for the IS, with a static subject and nothing to trip the system up. It might be a different story when trying to track a moving subject and shooting at high shutter speeds. I'm not sure how easy it would be to execute a controlled test to compare results with and without IS for such things - I'm thinking BIF, for example.


I guess one thing to consider is that with IS enabled you do have the free floating elements wibbling about trying to do their thing. With IS disabled those elements should be locked down and simply perform like other elements of the lens. I guess in a way this is like the argument about using AI Servo to shoot static subjects. Do you want your AF system trying to determine and adjust for movement when there is none, and would you rather have lens elements jiggling about when at high shutter speeds there really is no need? Perhaps it really comes down to personal choice, and perhaps the abilities and needs of the individual.

For me, since I can see no visible benefit from using IS at high shutter speeds, but do shoot in conditions where IS could do positive harm to my images, for me it is better left off when shooting at 1/800 and above. For slower speeds, like I said, it is a Godsend.
 
tdodd said:
For Hoppy.....

<snip>
Click to expand...

Thaks very much for that Tim. I could post a similar set if pictures, where if the IS is in fact making a negative contribution at high shutter speeds, it is effectively invisible.

As you say, it is very difficult to do any controlled tests, because when there is any evidence of blurring, it's always more likely that this is simply from camera shake and not from the IS at all. Or from diffraction etc (see below). Certainly if it all starts to happen at 1/500sec as Thom Hogan suggests, this is not a fast shutter speed at all, in fact below the hand-held rule's threshold for say a 400mm lens on a crop body.

I did a control shot with the camera firmly bolted to a tripod, IS off, mirror up, etc etc. At f/8 with my 70-200L 4 IS on a 5D2. It was pretty sharp!

I then did exacly the same shot, but hand held, with and without IS. In all cases, when the shutter speed rose sufficiently to rule out camera shake, the images matched the control for sharpness. The only difference was that I could go about three stops lower with the IS on quite reliably, and four or even five stops if I was prepared to accept a low keeper rate (very low at 5 stops ;) ).

Well actually, that's not strictly accurate as there was some variation in the images, but these differences were consistent with having to change the f/number and ISO to balance the shutter speed. The most noticeable variance was due to diffraction at f/22, but this of course was with the longer shutter speeds where there is no dispute about the benfits of IS.

On the panning thing, particularly at unusual angles, I think you have a good point :)
 
You must log in or register to reply here.
Back
Top