why can i only get to f5.6 but other times 5,4 ? confused
- Thread starter topcat07
- Start date
ecoleman
Suspended / Banned
- Messages
- 6,992
- Name
- Elliott
- Edit My Images
- No
It sound like you are using a cheaper zoom lens. These lenses have a variable aperture which means you may be able to get f3.5 at its widest zoom, but only f5.6 at the extreme.
Take a look at the f numbers on your lens casing is will probably say something like f3.5-f5.6 or similar.
More expensive lenses will only have one f number ie f4 or f2.8. This means the lens is capable of the same aperture throughout the zoom range.
Take a look at the f numbers on your lens casing is will probably say something like f3.5-f5.6 or similar.
More expensive lenses will only have one f number ie f4 or f2.8. This means the lens is capable of the same aperture throughout the zoom range.
OP
topcat07
Suspended / Banned
- Messages
- 431
- Edit My Images
- Yes
Hi Danny,
it is a 18-55mm kit lens which come with the nikon d40 i don't change it
not sure what you mean by the variable bit im afraid
OP
topcat07
Suspended / Banned
- Messages
- 431
- Edit My Images
- Yes
bingo!!!
at 18mm i can get to f3.5 but at 55mm only 5.6
thank you very much was driving me crackers
Teflon-Mike
Suspended / Banned
- Messages
- 1,076
- Name
- Mike
- Edit My Images
- Yes
The aperture is a hole, basically, that restricts how much light might pass through the lens.
f-number is the dimensionless ratio of the focal length of the lens divided by the aperture diameter.
Confusion arises when we say we set the aperture to a certain f-stop, rather than a certain diameter; a convenience as the f-number combines the variables of lens length and aperture diameter to a single ratio, that passes the same amount of light regardless of the lens length, or hole diameter.
If you use a hand held light-meter; it will provide an Exposure-Value or EV-Number for the reflected light level. Dial that number onto the calculator scale, and it will give you a range of shutter speeds and corresponding f-stops you can set on the camera. Doesn't matter if you are using a 10mm wide angle or a 1000mm telephoto.. anything wider or longer or in between. same f-number, same exposure value.
OK....
F-number = lens length / aperture diameter; So; aperture diameter = lens length / f-number
Lets look at a fast-fifty; 50mm fixed focal length 'Prime' lens; with a fastest aperture of f-1.4. The aperture diameter then will be 50mm/1.4 = 35.7mm
OK, lets try this with a 500mm telephoto. aperture diameter 500mm/1.4 = 357mm.... a little over a foot! and that is the size of the hole behind the lens... you'd need a pretty BIG diameter lens body to fit a hole that big in! And that would need a lot of very carefuly ground glass to fill it.
Which is why longer lenses tend to have slower maximum apertures.
So.... 55-300mm zoom, and we have a fastest aperture of f3.5 at the short 55mm end. 55/3.5 = 15.7mm.... if the aperture diameter stays the same, then at the long end, 300mm, f-no = lens length / aperture diameter: 300/15.5 = f19... same diameter aperture, f-number changes merely because the ratio has been changed, zooming to a different focal length.
OK.... You have an 18-55mm lens, with fastest aperture, f3.5 at 18mm, f5.6 at 55mm
18/3.5 = 5.142mm
55/5.6 = 9.821mm
This implies that the aperture does change diameter.... but odds are it doesn't.
I have Nikon Kit 18-55 next to me. at 18mm zoom setting, the front element is aprox 90mm from the lens mount. At 55mm zoom setting, its aprox 100mm, and its at its shortest at 85mm or so at about the 30mm zoom setting.
Err.. doesn't make sense does it? The 'lens length' we speak of seems utterly unrelated to the focal length...
This is because of another convenience of terminology.. when we say focal length and lens length, we have omitted the word 'equated' or 'equivilent'... in the old days, lenses were just that; a single bit of glass set in a tube that far from the film... of course more bendy it was more magnification it offered, but then the further from the film it needed to be to focus.... so you didn't get more magnification just from the longer tube, but the bow of the glass.
However... if you put a second lens behind the first... you could get double magnification.. which takes us to what are known as 'compound lenses' which is what we are holding.
I think the Nikon 18-55 has seven 'elements'... so the original focal length to magnification relationship is completely lost; and the numbers on it, are the 'equivilent' to single element lenses of old.
And its quite handy; you can change the 'equivilent' focal length by moving one or two elements inside the lens body, without moving the front element so far, as its all down to the ratios of distances between the elements inside.
So, back to the aperture... we have a leaf-blade iris, and we are talking 'equivilents'.
The actual hole will be at least in front of the rear-most element, so there will be some magnification of its true size in relation to the film sensor, and depending where it is in the groupings and how they move; whatever the actual diameter of the fastest aperture is... when you work out the 'equivilents' of that size to the effective focal length, you get the different f-numbers for different focal lengths.
So it's not an effect of a "Variable Aperture"... strictly speaking a variable aperture is the iris that lets you change the aperture diameter.... you can change the f-number on a fixed focal length lens, by changing the diameter of the aperture.
This change in f-number is the result of a FIXED diameter aperture, and a variable FOCAL LENGTH changing the F-Number ratio as the focal length is changed.
f-number is the dimensionless ratio of the focal length of the lens divided by the aperture diameter.
Confusion arises when we say we set the aperture to a certain f-stop, rather than a certain diameter; a convenience as the f-number combines the variables of lens length and aperture diameter to a single ratio, that passes the same amount of light regardless of the lens length, or hole diameter.
If you use a hand held light-meter; it will provide an Exposure-Value or EV-Number for the reflected light level. Dial that number onto the calculator scale, and it will give you a range of shutter speeds and corresponding f-stops you can set on the camera. Doesn't matter if you are using a 10mm wide angle or a 1000mm telephoto.. anything wider or longer or in between. same f-number, same exposure value.
OK....
F-number = lens length / aperture diameter; So; aperture diameter = lens length / f-number
Lets look at a fast-fifty; 50mm fixed focal length 'Prime' lens; with a fastest aperture of f-1.4. The aperture diameter then will be 50mm/1.4 = 35.7mm
OK, lets try this with a 500mm telephoto. aperture diameter 500mm/1.4 = 357mm.... a little over a foot! and that is the size of the hole behind the lens... you'd need a pretty BIG diameter lens body to fit a hole that big in! And that would need a lot of very carefuly ground glass to fill it.
Which is why longer lenses tend to have slower maximum apertures.
So.... 55-300mm zoom, and we have a fastest aperture of f3.5 at the short 55mm end. 55/3.5 = 15.7mm.... if the aperture diameter stays the same, then at the long end, 300mm, f-no = lens length / aperture diameter: 300/15.5 = f19... same diameter aperture, f-number changes merely because the ratio has been changed, zooming to a different focal length.
OK.... You have an 18-55mm lens, with fastest aperture, f3.5 at 18mm, f5.6 at 55mm
18/3.5 = 5.142mm
55/5.6 = 9.821mm
This implies that the aperture does change diameter.... but odds are it doesn't.
I have Nikon Kit 18-55 next to me. at 18mm zoom setting, the front element is aprox 90mm from the lens mount. At 55mm zoom setting, its aprox 100mm, and its at its shortest at 85mm or so at about the 30mm zoom setting.
Err.. doesn't make sense does it? The 'lens length' we speak of seems utterly unrelated to the focal length...
This is because of another convenience of terminology.. when we say focal length and lens length, we have omitted the word 'equated' or 'equivilent'... in the old days, lenses were just that; a single bit of glass set in a tube that far from the film... of course more bendy it was more magnification it offered, but then the further from the film it needed to be to focus.... so you didn't get more magnification just from the longer tube, but the bow of the glass.
However... if you put a second lens behind the first... you could get double magnification.. which takes us to what are known as 'compound lenses' which is what we are holding.
I think the Nikon 18-55 has seven 'elements'... so the original focal length to magnification relationship is completely lost; and the numbers on it, are the 'equivilent' to single element lenses of old.
And its quite handy; you can change the 'equivilent' focal length by moving one or two elements inside the lens body, without moving the front element so far, as its all down to the ratios of distances between the elements inside.
So, back to the aperture... we have a leaf-blade iris, and we are talking 'equivilents'.
The actual hole will be at least in front of the rear-most element, so there will be some magnification of its true size in relation to the film sensor, and depending where it is in the groupings and how they move; whatever the actual diameter of the fastest aperture is... when you work out the 'equivilents' of that size to the effective focal length, you get the different f-numbers for different focal lengths.
So it's not an effect of a "Variable Aperture"... strictly speaking a variable aperture is the iris that lets you change the aperture diameter.... you can change the f-number on a fixed focal length lens, by changing the diameter of the aperture.
This change in f-number is the result of a FIXED diameter aperture, and a variable FOCAL LENGTH changing the F-Number ratio as the focal length is changed.
HoppyUK
Suspended / Banned
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
Not quite sure what your point is here Mike, but a lens' physical length that you seem to be referring to, and the focal length, are not the same thing at all. You can't mean that, or maybe I've misunderstood. Focal length is the distance from the rear nodal point to the sensor, when the lens is focused at infinity. And that nodal point can often be outside the physical lens.
As far as physical aperture size is concerned, very few modern lenses obey the focal length/aperture diameter f/ratio literally, apart from a few fairly basic primes. The f/number is decided by focal length and the size of the aperture that the lens 'sees' in an optical sense (also at infinity).
Last edited:
- Messages
- 5,607
- Name
- Stephen
- Edit My Images
- Yes
Which is why we have telephoto designs (which are physically shorter than their focal length) and retrofocus designs (which go the other way).
The earliest extreme wide angle lenses on 35mm SLR cameras were a standard design, and hence came closer to the film plane than the mirror allowed, and hence had to be used with mirror lock up and a separate viewfinder. In those days "extreme" meant 21mm.
Unless there are other factors in play, the simple way of designing a zoom lens results in a constant aperture anyway. The "other factors" that mean that many aren't is that you need a physically larger lens (and hence larger elements) to achieve this - so the normal cost saving results in a variable aperture zoom.
Since it is vaguely relevant to this topic, autofocus has also had a side effect that is unexpected to many. The older lenses usually focused by physically moving the whole lens. Autofocus and zooms changed that; because autofocus drains batteries, and zooms are heavy. Hence, it became better to move elements within the lens rather than the whole lens. The side effect is that as you focus closer (which implies a greater lens to film separation) the focal length decreases. Try it at home and see; the image size you get at (say) 3 feet using a zoom lens set at 50mm will be noticably smaller than the image size using a 50mm prime manual lens.
HoppyUK
Suspended / Banned
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
Physical length and focal length are totally unrelated, nothing to do with different types of lens design.
It's much harder to design a zoom with a constant maximum aperture than a variable one.
It's not so much that internal focusing inherently changes focal length at closer distances, more a common side effect. Lenses with this 'focus-breathing' or 'focal-breathing' characteristic, as it's euphemistically called, vary a lot - anything from zero effect to a heck of a lot. At one extreme, most primes are generally not affected, whereas consumer grade super-zooms suffer badly.
Example: Nikon 70-200 2.8 is more like 140mm instead of 200mm at close distance; Canon 70-200 2.8 Mk2 a lot better, around 190mm; Tamron 70-200 2.8 non-VR stays very close to 200mm.
It's much harder to design a zoom with a constant maximum aperture than a variable one.
It's not so much that internal focusing inherently changes focal length at closer distances, more a common side effect. Lenses with this 'focus-breathing' or 'focal-breathing' characteristic, as it's euphemistically called, vary a lot - anything from zero effect to a heck of a lot. At one extreme, most primes are generally not affected, whereas consumer grade super-zooms suffer badly.
Example: Nikon 70-200 2.8 is more like 140mm instead of 200mm at close distance; Canon 70-200 2.8 Mk2 a lot better, around 190mm; Tamron 70-200 2.8 non-VR stays very close to 200mm.
- Messages
- 5,607
- Name
- Stephen
- Edit My Images
- Yes
Sorry, but I disagree and stand by my points. Can you point me to a reputable source that will prove me wrong, because I'm basing my statements on my memory (which may be at fault) and my understanding (which may be at fault) of the relevant passages in Sydney Ray's Applied Photographic Optics.
I fail completely to understand the distinction between something "inherently changes the focal length" and "a common side effect". To my simple mind, a side effect is caused by whatever it's a side effect of; and therefore in this case is caused by the internal focusing.
Edit to add: I've said my piece, and won't discuss it further as I found on another forum that that was unacceptable practice. I don't understand why, but people made things very unpleasent for me for wanting to debate things.
I fail completely to understand the distinction between something "inherently changes the focal length" and "a common side effect". To my simple mind, a side effect is caused by whatever it's a side effect of; and therefore in this case is caused by the internal focusing.
Edit to add: I've said my piece, and won't discuss it further as I found on another forum that that was unacceptable practice. I don't understand why, but people made things very unpleasent for me for wanting to debate things.
Last edited:
HoppyUK
Suspended / Banned
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
Sorry if I've offended you, it certainly wasn't intended. I don't want to debate anything either - these are facts, so there should be no debate involved.
Simply this: if internal focusing inherently changes focal length, then all internal focusing lenses would be affected. And many are, but not all, and not all to the same extent. It's basically one of the many compromises that a lens designer has to juggle with, and they can choose to fix it or not.
Again, apologies
- Messages
- 5,607
- Name
- Stephen
- Edit My Images
- Yes
I know I said I wouldn't post again, but I do want to make it clear that I wasn't offended. If my post came across in that way, I unreservedly apologise for creating that impression.
Apologies also for this Parthian shot. It may not be a fact, but I still stand by my interpretation of Sydney Ray's opening sentence in section 20.4.3 of Applied Photographic Optics, 3rd edition, p 195 section. The section is headed "Internal focusing":
"A most succesful and versatile method of focusing a lens by optical means is to use a floating element or group which moves axially internally to alter the separation of elements and hence reduce EFL to satisfy the conjugate equation."
Italics in the source, bolding is mine. The remainder of the discussion (and mathematics) makes perfect sense to me. If you can point out my error, or give me a reference that I can follow up, I'd be grateful. But not an internet source, please - there's so much error there that I don't trust it
Apologies also for this Parthian shot. It may not be a fact, but I still stand by my interpretation of Sydney Ray's opening sentence in section 20.4.3 of Applied Photographic Optics, 3rd edition, p 195 section. The section is headed "Internal focusing":
"A most succesful and versatile method of focusing a lens by optical means is to use a floating element or group which moves axially internally to alter the separation of elements and hence reduce EFL to satisfy the conjugate equation."
Italics in the source, bolding is mine. The remainder of the discussion (and mathematics) makes perfect sense to me. If you can point out my error, or give me a reference that I can follow up, I'd be grateful. But not an internet source, please - there's so much error there that I don't trust it
HoppyUK
Suspended / Banned
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
Stephen, the way I read your post, it implied that all internal focusing lenses (ie pretty much everything these days) have a reduced focal length at closer focusing distances.
That is certainly true in a lot of cases, more than the manufacturers would like us to believe, but it is not true of all IF lenses.
Let me put it another way: modern lens design is very complex and a lot of the classic optical conventions are not followed. There are several examples in this thread, such as the way telephoto and inverted-telephoto/retrofocus designs turn basic focal length calculations their head. Or that the f/ratio cannot simply be calculated by dividing focal length by the physical diameter of the aperture. Or that internal focusing lenses inherently have reduced focal length at close distances.
Sometimes all these things apply; sometimes none of them. And there are plenty of examples either way.
That is certainly true in a lot of cases, more than the manufacturers would like us to believe, but it is not true of all IF lenses.
Let me put it another way: modern lens design is very complex and a lot of the classic optical conventions are not followed. There are several examples in this thread, such as the way telephoto and inverted-telephoto/retrofocus designs turn basic focal length calculations their head. Or that the f/ratio cannot simply be calculated by dividing focal length by the physical diameter of the aperture. Or that internal focusing lenses inherently have reduced focal length at close distances.
Sometimes all these things apply; sometimes none of them. And there are plenty of examples either way.
Teflon-Mike
Suspended / Banned
- Messages
- 1,076
- Name
- Mike
- Edit My Images
- Yes
Point was, that people were explaining that kit 18-55 f3.5/5.6 lens gives different max aperture settings at either end of the zoom range by virtue of having a 'variable' aperture.
Kit lens does indeed have a 'variable' iris aperture, in so much as you can change the diameter of the hole that lets light through the lens, but the different f-numbers at either end of the zoom range are almost certainly NOT due to the iris diameter changing as you zoom.
Aperture is the hole in the iris; maximum aperture SIZE (probably) doesn't vary... the aperture SETTING, the f-number varies....... because it's a ratio of focal length to aperture diameter, so changes as you change the focal length.
But, do the maths, and the sums don't add up. f3.5 @18mm means you ought to have 5.1mm aperture diameter. f5.6@55mm means that the hole ought to be 9.8mm in diameter, suggesting it does in fact change....
When in all probability this isn't true; it wont be either 5.1mm or 9.8mm, but some other number; because in multi-element lenses, the 'lens length' and the 'aperture' setting, are usually calculated 'equivalents' not actual physical dimensions.
Hence putting ruler next to kit 18-55 at different zoom settings and showing how it length changes, getting longer then shorter again, ie: varying in length, with no notional relationship to focal length setting.. to explain how these numbers are, as far as user is concerned, 'imaginary' dimensions not tangible measurements.
So ... probably badly... to explain that a lot of misunderstanding comes down to terminology and confusion of expression.
We talk of 'Lens Length' when we mean 'Lens Focal length'....
We talk of 'Aperture' when we mean Aperture-Ratio or f-stop setting
And we 'omit' the qualifications of 'equivilent' or 'effective' from a lot of expressions, where things are calculated by comparison to reference standards, rather than actually measured physically.
But main point was, that the f-number is a ratio of aperture diameter to lens focal length; so the different f-numbers are due to changing the focal length, hence the aperture ratio or f-number.
NOT because the lens has a 'variable' aperture... the iris control that lets you change f-number by changing the hole diameter.
Asside: Set camera to A = Aperture priority and zoom to 18mm,
- Then select f8. Zoom through to 55mm, and the shutter speed only will change. The camera will close the iris down to whatever diameter is needed at that focal length to give f-8 ratio.
- Now set f3.5. Camera will stop the iris down to whatever diameter is needed to give f3.5 ratio at set focal length; but it can only do that at 18mm. As you increase focal length the ratio is going to fall, and at 55mm you'll get f5.6.
- Last go; set the camera to f4 and do the same. camera will set iris to whatever diameter is needed to give f-4 ratio.... but at 24mm it cant set a diameter wide enough, so gives it all its got, and will change the f-number until at 55mm it's giving f-5.6.
Lens makers COULD make 'constant' max aperture lenses simply by 'choking' the aperture, or closing the iris, at wider zoom settings; in which case your kit 18-55 would have an f5.6 max aperture throughout the zoom range.
I believe that this was actually quite common in the days of film and manual lenses.
- Messages
- 5,607
- Name
- Stephen
- Edit My Images
- Yes
Oh dear, I really, really didn't want to post again but...
If we're really going to be precise about things (and I desperately wanted not to make things more complex since this is the "Basics" section and some of what follows isn't) then the following points need to be understood.
1. The focal length of a lens isn't something that you can measure from any "obvious" point on a lens unless the lens has only one element.
2. A lens has two optically important points, the front and rear nodal points. Through each of these nodal points there is a plane (called the front/rear nodal plane) perpendicular to the axis.
3. Optical measurements of subject distance are made from the position of the front nodal point. Optical measurements of the image distance are made from the rear nodal point.
4. In telephoto designs the front nodal plane will usually be physically in front of (and hence outside) the lens; in retrofocus designs the reverse applies.
5. There is no reason why the front nodal plane should be in front of the rear one; they often are the other way round.
6. None of this calls into question any of the established optical conventions.. Focal length calculations are, and always have been, made from the rear nodal plane to the plane of focus when the lens is focused on infinity. And yes, there is an optical "definition" of infinity that works for all practical purposes. Modern lens design has not yet managed to circumvent the basic laws of optics.
7. None of this is in any way related to how the aperture is defined. It depends on a property that is related to, but not defined by, the physical size of the diaphragm. Surprising though it might seem, there is no reason to suppose that if you mount a lens in reverse the aperture (even fully open) will be the same. Some lenses have several stops difference between the engraved and actual aperture when inverted. I can supply figures and examples, but that would be excessive here. I said surprising; actually it should be obvious when you consider most modern lenses and all zoom ones.
So the statement that the f number is the ratio of the focal length and physical size of the diaphragm opening simply isn't true. There is nothing tangible in a lens that can be physically measured with a ruler that can be used together with the focal length to give the aperture directly.
8. Nothing has overturned the basic equation linking subject and object distances and focal length. It is still the case that to focus nearer you have to increase the distance between the rear nodal plane and the film or reduce the focal length. There is no alternative.
I will grant that it is possible in theory to move elements within the lens to move the nodal plane; I honestly don't know if this can be done and still maintain the focal length.
If anyone is disposed to question any of the above points, feel free to do so. As far as I'm aware they are all well known and well documented in evry book on photographic lenses (as well as general optics texts). I would be happy to have any misinterpretations pointed out, but I'm not writing an optics textbook on the forum so I won't reply again unless anyone can show me where I've slipped up.
If we're really going to be precise about things (and I desperately wanted not to make things more complex since this is the "Basics" section and some of what follows isn't) then the following points need to be understood.
1. The focal length of a lens isn't something that you can measure from any "obvious" point on a lens unless the lens has only one element.
2. A lens has two optically important points, the front and rear nodal points. Through each of these nodal points there is a plane (called the front/rear nodal plane) perpendicular to the axis.
3. Optical measurements of subject distance are made from the position of the front nodal point. Optical measurements of the image distance are made from the rear nodal point.
4. In telephoto designs the front nodal plane will usually be physically in front of (and hence outside) the lens; in retrofocus designs the reverse applies.
5. There is no reason why the front nodal plane should be in front of the rear one; they often are the other way round.
6. None of this calls into question any of the established optical conventions.. Focal length calculations are, and always have been, made from the rear nodal plane to the plane of focus when the lens is focused on infinity. And yes, there is an optical "definition" of infinity that works for all practical purposes. Modern lens design has not yet managed to circumvent the basic laws of optics.
7. None of this is in any way related to how the aperture is defined. It depends on a property that is related to, but not defined by, the physical size of the diaphragm. Surprising though it might seem, there is no reason to suppose that if you mount a lens in reverse the aperture (even fully open) will be the same. Some lenses have several stops difference between the engraved and actual aperture when inverted. I can supply figures and examples, but that would be excessive here. I said surprising; actually it should be obvious when you consider most modern lenses and all zoom ones.
So the statement that the f number is the ratio of the focal length and physical size of the diaphragm opening simply isn't true. There is nothing tangible in a lens that can be physically measured with a ruler that can be used together with the focal length to give the aperture directly.
8. Nothing has overturned the basic equation linking subject and object distances and focal length. It is still the case that to focus nearer you have to increase the distance between the rear nodal plane and the film or reduce the focal length. There is no alternative.
I will grant that it is possible in theory to move elements within the lens to move the nodal plane; I honestly don't know if this can be done and still maintain the focal length.
If anyone is disposed to question any of the above points, feel free to do so. As far as I'm aware they are all well known and well documented in evry book on photographic lenses (as well as general optics texts). I would be happy to have any misinterpretations pointed out, but I'm not writing an optics textbook on the forum so I won't reply again unless anyone can show me where I've slipped up.
HoppyUK
Suspended / Banned
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
None of that is in question.
Which is what I said.
It can be done, because it is done. It is not a given that every lens with internal focusing elements must have a reduced focal length at closer range, even though that is commonly the case. It's at the designer's discretion.
- Messages
- 5,607
- Name
- Stephen
- Edit My Images
- Yes
Then I have to bow to your superior knowledge and practical tests. I've certainly never seen a lens review that covered this particular point, and I'd certainly be very interested to know who publishes them. If, on the other hand, you've carried out the research yourself, then my hat is off to you.
Can you give me an example to I can research a little more? It's obviously difficult (even if possible) to prove a negative, but a positive is a doddle.
Can you give me an example to I can research a little more? It's obviously difficult (even if possible) to prove a negative, but a positive is a doddle.
HoppyUK
Suspended / Banned
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
I spend most of my life testing lenses these days, for various photo magazines (mostly Advanced Photographer and Digital SLR Photography). A lot of my findings are never published because there's precious little page space as it is and minor shifts in focal length are usually not relevant. When they are, I mention it, and I gave some examples earlier in the case of those 70-200s.
Measuring focal length is also not easy. Finding an accurate point to measure from is tricky, and then things like distortion throw it out anyway. Big shifts are easy enough to see and approximate by comparing field of view, but accurate measures take a lot more time and it's just not a priority. There are other things I'd like explore in far more detail way before a few mm shift in focal length. Flare control is one, that has a big impact on image quality but rarely gets mentioned in any measurable way. F/number vs transmission vs vignetting is another, also hard to do well bearing in mind any test regime I come up with has got to work with focal lengths from ultra-wide to super-tele and apertures from f/1.2 upwards.
But none of what I've said is at all contentious, and I've read similar references in other reviews often enough (eg DPReview, PhotoZone, TheDigitalPicture).