[quh86]

Hey folks,


no test pictures at this time, sorry!
I am just asking for the best way to assess sharpness on the pentax 16-50 f/2.8 ? What would you recommend?
I got this lens for a steal basically and on the outside it does not have a scratch, optics seem perfect aswell. Im just pondering whether it has a problem with element-alignment.

The pictures seem uniformly sharp at f8 and above (even f/18 seems to be very alright at the pixel level, diffraction might not be a big issue here) but below f/4.5 Im not so sure anymore. Sometimes the pictures turn out quite good, and sometimes not so much. However all pictures are okay with 50% crop, if i zoom in further, details degrade heavily, so is this lens supposed to be sharp at f/2.8 ?

At which focal length does the lens naturally perform best?
The autofocus does give a very ever so slightly squeaking sound. is this normal? (it might be that the lens hasnt been used in years, does it get better with time?)

thanks for your time

---

Last edited by quh86; 11-27-2015 at 04:09 AM.

[Unregistered User 8]

Hi. A little more information might be needed to help you. Which camera body are you using, are you hand holding or tripod. When I got my copy of this lens, it was back focusing on my K3. A +8 in camera correction fixed that, but after the correction, like you some of my photos were sharp and others not so good. After much practice of my hand holding technique the photos improved. I hope this is helpful.

[WPRESTO]

Almost all lenses perform best @ f5.6~8. Most zoom lenses perform best at intermediate FL settings (some exceptions here, but I've never seen a zoom lens test where the IQ was best at the longest FL for the lens). Many picture elements contribute to apparent IQ: light, contrast, camera motion, subject motion, focusing accuracy. Test your lens, each FL & aperture, under consistent conditions, same detailed subject, on a sturdy tripod, preferably using flash illumination (variable output manual flash most useful), preferably using manual focus and focus peaking. Then evaluate.

[asp1880]

I have the same feeling about my own 16-50. Sometimes it seems to disappoint, with no apparent reason. Focus was allright, aperture and focal length was middle of the road, shutter speed was adequate, the light had nice directionality, and yet - WTF? Dreamy pictures with fuzzy edges.


Other times it produces stunning results.


Perhaps mine is defective in the same manner yours are?


Regards,
--Anders.

[quh86]

thanks for the quick replies guys

so Im shooting with my old trusty K-5. However this body does have some focus-errors! Even on my manual lenses (50/1.7; 85/1.8) it shows wrong focus confirmation.
So I am aware of these issues.
The problem with a zoom of course is that you can only dial in the fine adjustments that applies to all focal lengths(FL).
I have a tripod, a flash, not a cable release but Im trusting in the 2 second delay-option

Possible explanations for my mixed results may be:
not adequate holding of the lens
too slow shutter speeds
In body stabilization going rampant
wrong AF fine adjustments

So I am aware of these things, the solution to get reproducable results is of course using a tripod. the questions im having are:
What are the FL that i should test (16,28,35,50) ?
what subjects should i use to test for sharpness ?
how do i go about the whole testing procedure ?
and lastly, did anyone find (if there is any) a sweetspot for the lens, where it is sharpest?

cheers,

Sebastian

[Rondec]

You really need to post some photos. My experience with the 16-50 is that it has very good (but not excellent) sharpness at f4 and is excellent at f5.6. Weakest area of the lens in my experience is in the 16 to 18mm range.

For testing purposes, I would do a range of tests -- brick wall combined with just a lot of real world shots and see what it looks like.

Biggest negative with regard to auto focus on the K5 is that is will not be consistent in tungsten light and a lens that focuses normally in other lights will be off with tungsten lighting (this has been fixed in subsequent cameras).

[DeadJohn]

...Sometimes the pictures turn out quite good, and sometimes not so much...

Samples of the not-so-good would help.

Some reviewers say the 16-50 is soft wide open. I think it's reasonably sharp at f2.8 and like most lenses sharpness improves when you stop down. My 16-50 is prone to fringing and flare in bright light at f2.8. The fringing is fixable in Lightroom but the flare is tough to correct. I wonder if some users mistake fringing for softness.

I've learned to use f2.8 sparingly. In full sunlight f2.8 causes a lot of fringing and extra work during processing. For night cityscapes, f2.8 gives ugly flare if a bright streetlight is near the corner of the frame. I treat my 16-50 like it has a maximum f4.0 aperture most of the time, with f2.8 only used for special cases.

FYI here's a set from my 16-50 https://flic.kr/s/aHsjP33xCE.

---------- Post added 11-27-15 at 12:45 PM ----------

I have the same feeling about my own 16-50. Sometimes it seems to disappoint, with no apparent reason. Focus was allright, aperture and focal length was middle of the road, shutter speed was adequate, the light had nice directionality, and yet - WTF? Dreamy pictures with fuzzy edges...

The fuzzy edges could be fringing and CA.

[Rondec]

This is the 16-50 at f54.5 and 34mm.



16mm and f4

[WPRESTO]

For a test subjectrinted pages with fine (small) font; or several dollar bills, the newer the better.

[clackers]

Test both a detailed surface at medium range and a scene at infinity.

[Mattox]

Funny enough I have been shooting with my k5 and sdm 16-50 for 6 years now and have had the lens and camera repaired in various occasions all for af issues. The lens never, ever is sharp enough for my liking. I have pretty much had it with this lens and it's gone caput again after a small (ahhhhhh). I know there are bad batch with this lens and am pretty sure mine is one.

[quh86]

thank you guys very much for your posts. it gave me some good ideas what to pay attention to when testing
I will post some pictures over the weekend, when i find time to do proper testing.
thanks again, love this forum

cheers,

sebastian

[quh86]

UPDATE:

so i spent some time making the AF adjustments according to the tutorial on the mainpage:

"The adjustment process is quite simple:
Choose a clear subject that has something in front of and behind it (lens alignment tools work best)
Take a photo with autofocus enabled
Check if focus is accurate on your computer monitor at 100% magnification
If things in front of your subject are sharper than the subject itself, apply a -1 focus correction and start over from step #2
If things behind your subject are sharper than the subject itself, apply a +1 focus correction and start over from step #2
If your subject is perfectly in focus, you are done!


Read more at: https://www.pentaxforums.com/reviews/lens-align-front-back-focus/the-remedy.html#ixzz3smunlkYL"

and i had to apply a full +10 to get it to an okay level:

now the extreme FLs 16mm and 50mm are a tad (maybe -1) too front focusing, while the medium range (20,28,35mm) are still back focusing by about +2. However i cant test this as I am at the maximum with +10.

That seems a little odd maybe. however i used a common ruler. The focus distance was about 0.6 meters or just a hair below 2 feet, according to the scale on the lens. I havent done any infinity testing yet.

See for yourself below (Focus Point is between the 70mm and 71mm mark):

50mm f2.8 AF +10


28mm f2.8 AF +10

[Unregistered User 8]

Hi Sebastian, what you have done so far is good but may I suggest a couple of things to you. At 0.6 mtrs, the depth of field of a 50mm 2.8 lens is about 1.5 cms. At 0.6 mtrs a 16mm 2.8 lens has a dof of about 15.7 cms so you can disregard the wide end. Might I suggest that you do the test shot again at 50mm with your ruler set up at about a 45 degree angle in good light and focus on the 70 again and then check your focus on the computer. I hope this helps.

[Rondec]

I tend to use live view for evaluating focus with a technique like this: AF microadjustment for the 1Ds mark III, 1D Mk3, 5D Mk2, 7D, 1D X. It isn't perfect, but it certainly seems to work pretty decently.