[vodanh1982]

Pentax SMC DA 70mm f/2.4 Limited - Review / Lab Test - Analysis

SMC Pentax-DA 70mm f/2.4 Limited Lens Review

Look at MTF50 chart, how are they so different? They are both tested on K-5.

[JinDesu]

They are quite different, not only in values, but also in pattern..

[Na Horuk]

I mostly stopped relying on tests when I noticed that there are huge differences in these measurements from site to site. Things like resolution, distortion, fringing.. meh, the lens reviews on this forum are more reliable, because you get real life experiences and sample photos.

[Docrwm]

You have to be able to analyze and synthesize information to read the reviews. Everyone has their own biases and they show in the reviews often. Both reviews linked to come to the conclusion that the DA70 is a stellar lens, one even said it was hard to give it back at the end of the review. I try to read all the reviews here on K-mount lenses and I try to pay attention to what camera people were using too. The old review of the DA70 was with a K10D as I recall, somewhat different than with a K-5 or K-30. All those factor into evaluating a lens for me - oh, and price too

[vodanh1982]

Other user reviews are not quantitative. It is hard to compare two lens without numbers. The MTF50 is a standard so I expect it is consistent with slightly errors. These two tests are huge different.

[DSims]

meh, the lens reviews on this forum are more reliable, because you get real life experiences and sample photos.

True - there's no substitute for actual results.

Learn to see the differences for yourself. And if you can't see the difference, than save your money and get the cheapest lens you like (which you should do anyway).

[kiwi_jono]

To me the MTF charts look pretty similar (in a number of ways). While MTF is a standard way of expressing optical performance, readings are greatly effected by test setup and which body / sensor is being used. For numerical value you can only really compare the MTF figures for tests done with the same test setup and the same body. And then there is sample variability as well. All the review sites come up with pretty different numbers generally (except comparing other lens numbers for same body on the same review site).

I tend to look at these charts in the following ways:
1. The shape of the chart. How does wide open compare to f5.6 and down to f16? Is the lens sharp over a wide range?
2. Center chart readings versus border / frame edge readings. Gives you a good idea how sharp it is across the frame.
3. Compare charts from other lenses on same model camera and same test technique (same review site generally).

[Docrwm]

To me the MTF charts look pretty similar (in a number of ways). While MTF is a standard way of expressing optical performance, readings are greatly effected by test setup and which body / sensor is being used. For numerical value you can only really compare the MTF figures for tests done with the same test setup and the same body. And then there is sample variability as well. All the review sites come up with pretty different numbers generally (except comparing other lens numbers for same body on the same review site).

I tend to look at these charts in the following ways:
1. The shape of the chart. How does wide open compare to f5.6 and down to f16? Is the lens sharp over a wide range?
2. Center chart readings versus border / frame edge readings. Gives you a good idea how sharp it is across the frame.
3. Compare charts from other lenses on same model camera and same test technique (same review site generally).

Exactly. Many of the tests of K-mount lenses have been done with older cameras like the K-10. Then you get a test with a K-5 and you have a host of differences, even if the other parameters of the test setup were the same. Apples and Oranges even when the "Test" itself was supposedly the same. That's why I also look at real users photos - lots of them.

[civiletti]

It's interesting that PZ found maximum corner resolution at f5.6, while ePHOTO found it at f11. Perhaps it's sample variation.

[civiletti]

I mostly stopped relying on tests when I noticed that there are huge differences in these measurements from site to site. Things like resolution, distortion, fringing.. meh, the lens reviews on this forum are more reliable, because you get real life experiences and sample photos.

How is a small sample image useful in determining the resolution and micro-contrast of a lens?

[vodanh1982]

http://www.ephotozine.com/articles/smc-pentax-da-35mm-f2-4-al-lens-review-16...35mm_mtf50.png

I think ePhoto is not right. The center sharpness is lower than the corner.

[cali92rs]

I had this lens for a while...it was great and wonderfully sharp. However, I never got used to the focal length. Too long in most cases indoors, and outdoors i feel like i want something longer.
But if you like the focal length (totally subjective) then it really is an awesome lens. Both articles highly recommend it, by the way.

[Na Horuk]

How is a small sample image useful in determining the resolution and micro-contrast of a lens?

How is a number describing resolution or micro-contrast useful in taking photos?

[lytrytyr]

This comment from the ePHOTOzine review is really mixed up:

"The MTF50 graph looks at the fine resolution and ability to record a very high standard of detail.
When MTF measurements started to be used in the 1970s many magazines started off using MTF40 data
and quickly reduced that to MTF30 as the figures always looked low."

The "MTF50 graphs" show what lp/mm (or lp/ph) resolution gives a 50% contrast,
while the older graphs mentioned show the percentage of contrast at 40 or 30 lp/mm.

It was firms like Sigma that started (and continue) using the 30 lp/mm graphs,
presumably to try and make their lenses look better in comparison
with the products of other firms that followed the 40 lp/mm standard.

[lytrytyr]

How is a number describing resolution or micro-contrast useful in taking photos?

It helps you decide which lenses to buy, and which lenses to avoid.