[BigMackCam]

A short while back I was searching for a list of Pentax lenses along with a summary of the optimum aperture ranges to use with each to obtain the best resolution. It seems one didn't exist, so I decided to pull something together for the lenses I own, using data harvested from a number of lens review websites. Then, having put some effort into that, I extended it to include a wider range of DA and FA lenses. I figured it would be useful if I should ever think of buying any of these lenses, and may also come in handy for some forum members.

What these tables summarise is how well each lens resolves detail across the aperture settings, highlighting the optimum apertures to use for center and edge sharpness respectively. The source data often went no further than f/8, but as we know, performance will gradually tail off at this point due to diffraction.

This is a work in progress. So far I've covered most of the DA and FA primes, where data is available, and I've made a start on some of the zooms. My intention is to cover all DA and FA lenses over time.

I'd welcome any constructive feedback (including corrections - it's possible I've made an error or two!) from those of you who might find this useful or interesting. I appreciate many of you won't, in which case I'm happy to provide a full refund of the asking price

---

Last edited by BigMackCam; 11-26-2015 at 06:37 AM.

[AquaDome]

Thank you for this list. I spent way more time looking at it that I expected I would.

[pathdoc]

Fascinating! It would be good to know what your cutoffs for the various performance categories are. I'm not asking you to justify them per se, heaven knows that starts enough endless debates; just that it would be good to know how and where the bar has been set so we can make our own judgements.

[normhead]

That's a lot of work... thanks for the effort....

[bertwert]

Useful list, I see why you wanted the table now

[BigMackCam]

Fascinating! It would be good to know what your cutoffs for the various performance categories are. I'm not asking you to justify them per se, heaven knows that starts enough endless debates; just that it would be good to know how and where the bar has been set so we can make our own judgements.

That's a perfectly fair and relevant request. Broadly speaking (because the websites I used are not in sync on this), the following resolution bands apply:

Poor: up to 1150 lw/ph
Fair: 1150 - 1450 lw/ph
Good: 1450 - 1750 lw/ph
V.Good: 1750 - 2050 lw/ph
Exc.: better than 2050 lw/ph

In some cases I was able to harvest specific figures for each aperture - in others, I had to interpolate as best as possible. Also, where a measurement was *very* close to the next band (eg. 2043 lw/ph is V.Good but right on the border of Excellent) and the measurement for the previous aperture was considerably lower (say, 1800 lw/ph), I've rounded the higher measurement up, as it gives a better appreciation of the performance curve for the purposes of the reader. I hope that makes sense!?

---------- Post added 11-26-2015 at 03:13 PM ----------

To all - thanks for the nice feedback so far, folks. Glad it's of interest

[normhead]

The lw/ph of the lenses varies pretty drastically with the sensor resolution... did you normalize for that?

[BigMackCam]

The lw/ph of the lenses varies pretty drastically with the sensor resolution... did you normalize for that?

Hi Norm... good point. So, I think it's important to say that this is not a relative comparison of the various lenses, but a table of numerous *individual* resolution profiles. Some of the tests were carried out with older, lower resolution sensors. The example resolution bands that I just gave were from a review of the FA31, based on a 10Mp sensor, others were based on 7MP if I remember correctly (still others may have been with higher resolution sensors). Furthermore, the various analyses within a single review site vary considerably in terms of the sensor resolutions used. Normalising across these, and then across the various websites, would be a bigger task than I have appetite for

In most of the analyses, the Poor, Fair, Good, V.Good, Excellent rating at a particular aperture is either explicit, or can be interpolated from a smaller number of ranges (one site typically has graphs broken down into four ranges, as I recall - broadly representing Fair, Good, V. Good, Excellent, while another site uses bar charts that are typically - but not always - broken down into five bands, from Poor to Excellent). I've done my best to normalise this to some extent. Hopefully this makes sense?

[normhead]

Sounds like the best anyone could have done with what's out there, thanks again.

[timb64]

Great job,BigMack.

[pathdoc]

Methodology duly noted and gratefully accepted in the context of its limitations.

I know there are charts for much older lenses (M, A, even K), but they are given in terms of lp/mm, with the best topping out around 100. How does this translate to lw/ph, or is there no reasonable comparison?

[fwcetus]

Thank you very much for your work. I am impressed with what you have so far, and I think you have handled the variables involved the best you can.

However, one question (or actually two questions) that I have -

1. Are the results shown from FF, cropped, or both formats?

2. How are you handling the difference (primarily in the "Border" data) between FF and cropped sensors?

[Rondec]

Interesting collection of data. I think there may be an error under the DA 560 where there is data listed for f4 and above, but I believe it is an f5.6 lens?

DA *200 seems odd to me. My experience was that it got to excellent at f4, but data shows a very narrow window where it has excellent performance.

Thanks for sharing this.

---------- Post added 11-26-15 at 11:45 AM ----------

Thank you very much for your work. I am impressed with what you have so far, and I think you have handled the variables involved the best you can.

However, one question (or actually two questions) that I have -

1. Are the results shown from FF, cropped, or both formats?

2. How are you handling the difference (primarily in the "Border" data) between FF and cropped sensors?

I imagine it is all based on cropped cameras of various vintages.

[bertwert]

I think there may be an error under the DA 560 where there is data listed for f4 and above, but I believe it is an f5.6 lens?

That is correct. It's an f5.6 lens.

[btnapa]

Very informative. Must have been a lot of work. Thanks again.