Canon EF 100-400 f/4-5.6L IS "Soonish" [CR2]

[richy]

With the current 100-400L model, what is the minimum shutter speed you need at 400mm with IS to get a decent keep rate for your picture? I assume it would be less then 1/400. Maybe 1/200 with the current IS? Any thoughts as to how much of an improvement a new IS version would make using these parameters?

I was always afraid that at 400mm and f5.6, even with IS it must not be always obvious to get great shot on a cloundy day?

I believe it has a 2 stop IS system, therefore 1/100 @ 400mm assuming you don't need faster to stop subject motion and that the IS is a full 2 stops at 400mm.

Should the new one have a 4 stop system that would drop to 1/25 roughly. Its not an exact science with IS. As you say its more a 'good percentage of keepers' thing rather than guaranteed crystal shots.

 

[Meh]

With the current 100-400L model, what is the minimum shutter speed you need at 400mm with IS to get a decent keep rate for your picture? I assume it would be less then 1/400. Maybe 1/200 with the current IS? Any thoughts as to how much of an improvement a new IS version would make using these parameters?

I was always afraid that at 400mm and f5.6, even with IS it must not be always obvious to get great shot on a cloundy day?

IS improves blur due to camera shake which depends a lot on the person holding the camera. If you've got shaky hands to start with, been up all night drinking, and had 7 cups of coffee... then you'll need a much faster shutter even with IS.

The rule-of-thumb of 1/focal length is the minimum hand-holding shutter speed. I've also heard of 1/2L as a better guideline for the average person with average hand steadiness.

IS in the 100-400L claims a 2 stop improvement. So if for you minimum hand-holding shutter speed is 1/400 then with the IS turned on it should be 1/100.

 

[lol]

The 1 / focal length guide applies to 35mm equivalent focal length, so for crop sensor it does get that bit worse. In practice I'm ok down to say around 1/80 at 400mm on crop sensor, hand held, unsupported. Another stop or so with some support e.g. leaning on something.

Have been using the 70-300L more recently, and you can definitely feel the extra IS on that in challenging situations.

 

[Meh]

The 1 / focal length guide applies to 35mm equivalent focal length, so for crop sensor it does get that bit worse.

Why would that be true? Not questioning your personal experience but what's the technical reason for that. Possibly because with a crop-body you're standing further away for the same framing, so the angle of view is less, and vibrations due to hand shake would then cross a greater angle. Ok, true assuming same framing which is the case 99% of the time.

 

[richy]

The 1 / focal length guide applies to 35mm equivalent focal length, so for crop sensor it does get that bit worse.

Why would that be true? Not questioning your personal experience but what's the technical reason for that. Possibly because with a crop-body you're standing further away for the same framing, so the angle of view is less, and vibrations due to hand shake would then cross a greater angle. Ok, true assuming same framing which is the case 99% of the time.

It's roughly true in my experience but not specifically due to it being a cropped sensor (as far as I can tell), more that cropped sensors see higher pixel densities (a 5d2 would be about 8mp as a crop with the same density but a 7d would be around 45-47mp as a full frame). In digital medium format, as sensors mp counts have increased drastically that requirement for better technique to avoid blur has increased along with it. I find with the same 50mm lense I need 1/60 on the 5d2 but 1/90 on a 7d for a high keeper rate, However 1/60 is about right for my 50mm on my rz67. If 1/fl holds true on a 6x7 like it does on 35mm it should also hold true for apsc. I guess if canon made a 45mp full frame sensor (sony probably have a 45mp apsc in the works ;-) ) it would probably need better technique or a faster shutter speed. By 'better technique' I refer to methods such as a tripod, IS, holding breath etc, I'm being clumsy with the wording, been awake too long, I mean just taking steps to avoid operator induced blur.
I found (as many did) the 7d to be more demanding than the 10d, 20d or 40d in that respect.

 

[AJ]

The size of the front element is not what determines the f-stop.

It pretty much does for telephoto lenses. e.g. the 200/2.8 has a 72 mm front element.

For normal and wide lenses the front element is not what decides f-stop. This critical lens element is towards the rear.

If this lens is indeed a 100-400/4-5.6 with a 82 mm front element then the extra front element size may ease vignetting at the short end. I doubt it would do much for the long end.

 

[Loswr]

The 1 / focal length guide applies to 35mm equivalent focal length, so for crop sensor it does get that bit worse.

Why would that be true? Not questioning your personal experience but what's the technical reason for that. Possibly because with a crop-body you're standing further away for the same framing, so the angle of view is less, and vibrations due to hand shake would then cross a greater angle. Ok, true assuming same framing which is the case 99% of the time.

Because the 1/focal length approximation derives from the relationship between the angular motion and the angle of view. Thus, at wide angles, where a given amount of angular motion represents only a small fraction of the angle of view, the effects of camera shake are less likely to have an impact on the image. Since the crop factor reduces the angle of view, you need to account for that in the 1/focal length approximation.

 

[Cropper]

Just read on another website that both the 500 II and the 600 II delivery dates have been postponed indefinitly. The formal anouncemment of the 200-400 x1.4, scheduled to occur on February 2011 with a realease date during this year, has also been postponed indefinitly.

In light of these information I wonder what might the definition of "Soonish" be for Canon. :-(

 

[Loswr]

Just read on another website that both the 500 II and the 600 II delivery dates have been postponed indefinitly.

Link? No one has told B&H, who still list December for estimated availability of the 500L II.

 

[WildBill]

here's a link: http://www.the-digital-picture.com/News/News-Post.aspx?News=1439&utm_source=twitterfeed&utm_medium=facebook

 

[Loswr]

Thanks! Bummer...especially since the 500/4 IS MkI is now officially discontinued, and not available from retail outlets anymore. IIRC, the exact same thing happened with the 300/2.8 IS II and 400/2.8 IS II - d/c'd the MkI then the MkII was delayed several months.

Makes one really question 'soonish' - even if we see a prototype 'soonish' the announcement will be delayed, then the announcement will occur but the release will be delayed, then delayed again. C'mon, Canon, get it together! If they don't have the 500L and 600L II lenses in peoples' hands in time for the Olympics, they're going to take a big hit.

 

[Meh]

The 1 / focal length guide applies to 35mm equivalent focal length, so for crop sensor it does get that bit worse.

Why would that be true? Not questioning your personal experience but what's the technical reason for that. Possibly because with a crop-body you're standing further away for the same framing, so the angle of view is less, and vibrations due to hand shake would then cross a greater angle. Ok, true assuming same framing which is the case 99% of the time.

Because the 1/focal length approximation derives from the relationship between the angular motion and the angle of view. Thus, at wide angles, where a given amount of angular motion represents only a small fraction of the angle of view, the effects of camera shake are less likely to have an impact on the image. Since the crop factor reduces the angle of view, you need to account for that in the 1/focal length approximation.

Essentially that's the logic I walked myself through but it is relative to the "same framing" I think. If the camera is in the same position the APS-C image is nothing more than a crop from the FF image circle that the lens renders on the focal plane and the effect of angular motion imparted by shake can not be greater in the crop area than it is in the rest of the FF image.

One thing is different however; circle of confusion is smaller on an crop sensor because we enlarge more for any given print size. Is that the actual physical phenomenon at play... when the angular motion that occurs during the exposure is larger than the CoC we begin to see the blur?

 

[Mt Spokane Photography]

With the current 100-400L model, what is the minimum shutter speed you need at 400mm with IS to get a decent keep rate for your picture? I assume it would be less then 1/400. Maybe 1/200 with the current IS? Any thoughts as to how much of an improvement a new IS version would make using these parameters?

I was always afraid that at 400mm and f5.6, even with IS it must not be always obvious to get great shot on a cloundy day?

I've used mine hand held at 1/25 sec out my back window. contrast was poor, but the image was usable, even if not in the excellent range. It amazed me that it did so well at such a slow shutter speed.

I saw our baby goats pllaying king of the mountain on our horse, and the camera had my 100-400 on it, so I just took the image less than a second before the horse rolled. No time to properly set up for the shot.

1/25sec, f/11, ISO 800. Obviously, if I had time, I'd have set a wide aperture, faster ISO, and had a faster shutter.

The other day, our young baby goats ambushed the horse as he walked by their 3 ft high house where they were playing, and one jumped on his back while he was standing by them. He immediately started walking so no chance to get the camera. The horse loves his goats, and they love him.

 

[Loswr]

Essentially that's the logic I walked myself through but it is relative to the "same framing" I think. If the camera is in the same position the APS-C image is nothing more than a crop from the FF image circle that the lens renders on the focal plane and the effect of angular motion imparted by shake can not be greater in the crop area than it is in the rest of the FF image.

The effect of the angular motion on the image circle is no different, true. But the image is being sampled from a smaller area of the image circle with a crop sensor.

The AoV will affect the magnitude of the impact of camera shake. Let me illustrate: A 400mm lens on FF has a diagonal AoV of 6°12', so if your angular motion (camera shake) occurs within arcs of 30", then that motion represents 0.135% of the diagonal angle of view. But, that same 400mm lens on APS-C has a diagonal AoV of 3°52'. So, the same amount of camera shake (sensor size won't affect how shaky your hands are!) represents 0.216% of the diagonal AoV.

 

[Meh]

Essentially that's the logic I walked myself through but it is relative to the "same framing" I think. If the camera is in the same position the APS-C image is nothing more than a crop from the FF image circle that the lens renders on the focal plane and the effect of angular motion imparted by shake can not be greater in the crop area than it is in the rest of the FF image.

The effect of the angular motion on the image circle is no different, true. But the image is being sampled from a smaller area of the image circle with a crop sensor.

The AoV will affect the magnitude of the impact of camera shake. Let me illustrate: A 400mm lens on FF has a diagonal AoV of 6°12', so if your angular motion (camera shake) occurs within arcs of 30", then that motion represents 0.135% of the diagonal angle of view. But, that same 400mm lens on APS-C has a diagonal AoV of 3°52'. So, the same amount of camera shake (sensor size won't affect how shaky your hands are!) represents 0.216% of the diagonal AoV.

Math is correct (I presume, I didn't check it) but without the additional criterion of enlarging the crop area to the same viewing size, I could apply your math to a software crop of a FF image (same lens, same camera) and come to the same conclusion which would be incorrect because the blur must be the same throughout the entire FF image.

 

[Loswr]

I could apply your math to a software crop of a FF image (same lens, same camera) and come to the same conclusion which would be incorrect because the blur must be the same throughout the entire FF image.

Correct. If you cropped in software, the reduced angle of view would still mean a relatively greater impact of the existing blur, which might take it from below threshold to above threshold for detectability. In the same way, cropping an image in post affects sharpness and DoF.

 

[Meh]

I could apply your math to a software crop of a FF image (same lens, same camera) and come to the same conclusion which would be incorrect because the blur must be the same throughout the entire FF image.

Correct. If you cropped in software, the reduced angle of view would still mean a relatively greater impact of the existing blur, which might take it from below threshold to above threshold for detectability. In the same way, cropping an image in post affects sharpness and DoF.

I may just be quibbling, or I perhaps I'm just wrong, but I'm not comfortable with your phrasing. It's not the cropping per se that causes these effects. Rather, it is the greater enlargement of the smaller image (usually to achieve the same viewing size) and that is what the CoC is a measure of.

 

[Loswr]

Sorry, yes - for the same final output/reproduction size (i.e. back to the CoC, in a way).

 

[LetTheRightLensIn]

The 1 / focal length guide applies to 35mm equivalent focal length, so for crop sensor it does get that bit worse.

Why would that be true? Not questioning your personal experience but what's the technical reason for that. Possibly because with a crop-body you're standing further away for the same framing, so the angle of view is less, and vibrations due to hand shake would then cross a greater angle. Ok, true assuming same framing which is the case 99% of the time.

Because the 1/focal length approximation derives from the relationship between the angular motion and the angle of view. Thus, at wide angles, where a given amount of angular motion represents only a small fraction of the angle of view, the effects of camera shake are less likely to have an impact on the image. Since the crop factor reduces the angle of view, you need to account for that in the 1/focal length approximation.

If you want to even up blur as percent of total image then apply the crop factor.

If you want to get what you can out of the sensor then you need to use factors relating only to pixel density and crop could end up needing less shutter speed than FF in rare cases (10D vs 5D2).

 

[LetTheRightLensIn]

here's a link: http://www.the-digital-picture.com/News/News-Post.aspx?News=1439&utm_source=twitterfeed&utm_medium=facebook

Wow, they look to be aiming to miss the London games.
I think the rumors that the disaster hurt things way more than anyone lets on were true.