A Site for Photographers by Photographers

Community > Forums > photo.net > Lenses > MTF tests VS Real world...

MTF tests VS Real world results

Lee L , Jan 09, 2000; 06:28 p.m.

The title kinda asks my question but the reason why I am asking is because I recently borrowed a Nikon N70) and 24-120 Nikko. The MTF results are no that great, I think around 2.6 out of 5 with the lens but My pics look great. (When the Sigma 28-70 f/2.8 has a 3.5 for half the price)The only noticable thing about the lens I do not like is the vigetting at 24mm. How can a lens with a relativly low MTF score on the lens result in great looking pics? What other factors are there to look for?


    1   |   2     Next    Last

Brian Walsh , Jan 09, 2000; 07:06 p.m.

If the "MTF results" you cite ("about 2.6 out of 5" vs. "3.5 for half the price") are from photodo.com, you might go back there and check how the folks behind that site think you might best use the numbers that they generate from the weighted averages of their MTF tests. I don't think that such numbers are at all useful when comparing very different focal lengths or focal length ranges.

I've been car shopping recently. I give the Audi TT a 4.9 out of 5, while a stolid all-wheel-drive Volvo wagon rates a solid 4. Based on the higher score and the lower price, I'd say the tiny sports car is the clear winner, but my wife seems to think the family wagon might be more useful in a wider range of applications. She also thinks I shouldn't compare apples with oranges.

Bob Atkins , Jan 09, 2000; 07:17 p.m.

"What other factors are there to look for? "

There are no other factors. If a lens gives great images, it's a great lens, no matter whether it gets a 5.0 or 0.0 on some published "test".

You have just learned an important lesson, namely that you can't use published MTF numbers to decide which lens to buy. Now you can laugh at people who pick their lenses only on the basis of the photodo numbers. Think how much money you'll save...

Todd Frederick , Jan 09, 2000; 08:00 p.m.

A quick side question; what is an MTF test?

Brian Walsh , Jan 09, 2000; 09:51 p.m.

Several brief articles at the photodo site explain MTF measurements and interpretation. You can find them at:


Michael Lopez , Jan 10, 2000; 12:19 a.m.

MTF is a good summary measure of the sharpness and contrast of lens optics. It does not address other issues, such as the sturdiness of a lens, whether it suffers from pincushion or barrel linear distortion, and *whether the lens is suited for the kind of photographs you want to take*. A wide angle lens with terrific MTF figures won't be of much use to you if you need a powerful telephoto lens to take photographs of wildlife. The problem I have with a 28-70mm zoom lens is that I like to take portraits at about 100mm focal length, and the world's best 70mm lens won't give me the perspective I prefer. BTW, if you now decide that you want a 100mm lens for portraits, you can get a fixed focal-length (non-zoom) 100mm lens of more than adequate quality for far less than the cost of a 24-120 Nikkor.

MTF is most important to people who either are very fussy about image sharpness, or who need to reproduce high-quality big enlargements, which magnify every small flaw on the negative. For example, pro photographers taking pictures that they hope will become tack-sharp full page illustrations in glossy magazines worry about the things which MTF measures.

If your photographs rarely are enlarged beyond 4x6 inches, bear in mind that most people have a hard time noticing differences in MTF at that size. On the other hand, if you often want to enlarge your photographs to 8x10 or 8x12 inches, getting a high MTF is important. If the lens you have can produce the results you want, that's really all that matters.

Neal Vaughan , Jan 10, 2000; 01:14 a.m.

I hate photodo. The results posted dont even come CLOSE to real world experiences in my opinion.

lee, you ask what other things to loook for. Photodo tests all of its lenses at infinity. So a lens that is optimized to work at a closer distance, say a portrait or macro lens, may get a bad grade on photodo but may be a spectacular performer in its intended purposes.

Ivan Verschoote , Jan 10, 2000; 09:21 a.m.

My own experience learns:

1) buy only brand names

2) most of the time; the more expensive, the better

3) read photo.net carefully - the best lenses are getting the most feedback

4) compare different magazines (or websites) and you'll be surprised of the results you will get

5) ask a friends experience

6) be aware: some brands are often (discreetly) promoted by magazines by the means of a better figure

7) besides optics, think about mechanics and its practical use

8) .... my inspiration is finished....

Michael Lopez , Jan 10, 2000; 10:43 a.m.

I shouldn't post so late at night ...

SQF (a log-weighted integral of MQF over a certain range of frequencies) is the good summary measure of which I was thinking. Since SQF is based on MTF, it does not measure anything which MTF doesn't measure.

If you search on "Photodo" you'll find that some photonet members find that its tests are useful as one among several indicators, while others distrust them entirely.

As others have pointed out above, lenses are like cars: what is best for you depends on your budget and how you intend to use what you buy.

Vince Farnsworth , Jan 10, 2000; 11:43 a.m.

As pointed out above, these tests are done at infinity focus and are pretty useless for comparing lens performance in other conditions. I noted that the Contax (Zeiss) 85/1.4 Planar lens had spectacular numbers and was significantly "better" than the 85/2.8 Sonnar lens. I decided to test both in real-world situations and came to the conclusion that the 85/1.4 was significantly sharper at infinity focus but the 85/2.8 was at least as sharp just about everywhere else, and significantly sharper at close range. My results were consistant with the MTF graphs but revealed much more about the character of each lens. The differences would only be noticed by very picky viewers or in big enlargements. Real world testing is the only reliable way to choose between lenses. The MTF graphs are only one point of reference.

    1   |   2     Next    Last

Back to top

Notify me of Responses