Is your zoom lens what it claimed to be?

[mpenza]

Minolta Dimage 7's actual focal length appears to be smaller than claimed:
http://webpages.charter.net/bbiggers/DCExperiments/html/d7_focal_length.html

Is the lens on your camera the same?

 

[Darren]

I would stick my neck out here and state that ALL lenses do not strictly conform to their stated specifications.

For example, that shiny new 80-200f/2.8 will most likely be a 82-197f/2.9 IF you really wanted to be exact to the extreme.

If this right? Is this wrong? Who knows ... most likely this was a marketing approach as its easier to group lenses into categories rather than having weird-numbered focal lengths.

Photodo.com has tested the effective focal length in most of their tests - Photodo MTF Lens Charts - CLICK ME!!

Example - Nikkor AFS 28-70f/2.8 Effective focal-length is 29-67mm; Nikkor AFS 80-200f/2.8 Effective focal-length is 81-193mm.

 

[mpenza]

A bit is acceptable but 10% is quite a lot (28mm to 200mm is 7.14x, but tests show it to be only 6.5x)!

 

[Darren]

Could there be some factor in the interpolation of the image data from the CCD? Looking at the way the test was done (from the link in your first post), I suspect the interpolation is reducing the number of pixels, in effect, giving you a shorter focal length :dunno:

I could be wrong, just hypothesizing a theory.....

 

[Ian]

Originally posted by mpenza
A bit is acceptable but 10% is quite a lot (28mm to 200mm is 7.14x, but tests show it to be only 6.5x)!

I tend to agree with you, however the fact is that most camera lens elements aren't for the most part individually tested and that manufacturing tolerances are actually a lot less tight than most people would imagine.

I recall reading an article a few years back that compared elements from several samples of a len of the same brand and model, that is the same lens and on accurate measurement of the elements they found a 5-6% variation from memory on the individual elements. These figures sounded about right to me as they are on par with commercial telescope manufacturers who generally cite a 5% tolerance in their optics.

Even high quality optics such as Edmund Scientific's Tech spec lenses have typical focal length tolerances of +/- 1-2% depending on design and they are more accurate than the lenses employed in most camera lenses. Now when you tote up the total tolerance of say 16 elements at 1% per element the system tolerance becomes quite sloppy!

Food for thought perhaps?

 

[mpenza]

hmm.... then we might need to test more when we buy any camera/lens ;p to try enough units to find one that exceeded the specifications

 

[Ian]

Originally posted by mpenza
hmm.... then we might need to test more when we buy any camera/lens ;p to try enough units to find one that exceeded the specifications

In reality it doesn't make a whole lot of difference as camera systems when looked at as a whole are a mixture of very tight tolerances and exceedingly sloppy tolerances.

It should be remembered that lens focal lengths are 'nominal' as are rated aperture values.

Camera shutters are a great example of how inaccurate many camera parts are with most shutters having a 1/3rd stop speed tolerance when new. Now before everyone starts jumping up and down I should point out that the human eye is not generally capable of differentiating a light level change of less than 1/3rd stop anyway so it makes no difference in the real world.

The sad part is that very few of us ever get the opportunity to 'cherry pick' lenses from camera shops let alone from the warehouse

However as I've been advising for years, if you don't rest a lens before purchasing it then you can be saddled with a lens that isn't much good.