[buessi]

After many months waiting for Pentax/Ricoh to come to its senses, I finally jumped into the pool just before Christmas and purchased a red (!) K-50 at a very good price via an Amazon vendor. I am still learning to use all of its well-regarded features, but one stands out: the kit 18-55mm lens is quite soft in AF-S mode and Auto. When compared with my other Pentax or K-mount lenses, it's quite apparent. My Sigma 28-70 and another Sigma 28-80 lens are sharper, and so is even a Tamron 100-300mm! For the past several weeks I have tried everything to maximize the focusing capability of the kit lens. The AF Fine Adjustment for the lens seems to settle on -5 as the best image (I magnify the images in Photoshop to 600% or more). My other lenses check in at other values: Tamron zoom goes at 0 (no Fine Adjustment needed), and the Sigma zooms run to around +2. The best focus is achieved with a Pentax 50mm prime, a kit lens from some time in the SF1N era. It also needs noFine Adjustment. I have downloaded two different lens testing images, done the 5-battery lineup described elsewhere as the easiest way to check focus and - of course - use a tripod and a cable release. I do want to find the best autofocus setting, but I am hesitant to assume the problem is with the camera itself. What do you think?

[lytrytyr]

I do want to find the best autofocus setting, but I am hesitant to assume the problem is with the camera itself. What do you think?

The contrast-detect autofocus (CDAF) that you have in live view
will sidestep many of the potential front/back focus inaccuracies
that are inherent to phase-detect autofocus.

So I would recommend seeing how your images look when using CDAF.

That said, the kit lenses are mass-produced cheaply,
so there is a wide variation in the quality of the samples that are sold.
You may have been unlucky with your copy.

[Na Horuk]

Hey there, welcome and great that you got an awesome red K-50!

But I think you might be equating some different concepts here. For example, the lens being soft is not the same as AF problems. To test its "sharpness" (quality), you need to use a tripod, 2 sec timer, set lens to 35mm, camera to Av mode at f8, ISO 100, and manually focus on an object at a reasonable distance (maybe using live view). This should be done in bright, natural light. This is how to test if the lens is "sharp"

AF is another issue. In live view, you can use CD AF (contrast detect), which is precise, but slow. just make sure its not focusing on something with more contrast, for example a contrasty background behind the object.
Outside live view, the camera uses PD AF (phase detect). The AF adjustment only applies to this one, as it may misfocus, but it is fast and works even in dark conditions. The problem people often have is that they don't know which AF point is being used. And the other problem is that the AF areas are actually much bigger than the little AF overlay in the viewfinder (the little red light that lights up). so maybe the camera is locking focus, just not on the object you want it to. Anyway, this is a good link to start with. This is common accross all brands, which is why DSLRs have AF adjust function and test charts and such.

The other thing to keep in mind are shutter speed and ISO. A high ISO can cause so much noise, the photo looks unsharp. The other problem might be slow shutter - as then the handshake will be noticeable and the photo will be blurry. This is why for tests you need to use lowest ISO and a sturdy tripod.

It would be great if you could post some photos so we can help you diagnose the problem. And keep the exif info in the photo, so we can see aperture and everything.

[buessi]

Knowing that I needed absolute sharpness, all photos were shot on a tripod using a cable release (non-electronic). All in viewfinder mode. Usually these were shot in bright daylight, ISO 100, and I ran through the entire gamut of AF options. So there should be absolutely no difference in focus, other than the lens.

The standard included here is a Sigma 28-80 mm on the K-50 body, then a Sigma 28-70mm (S70), the prime Pentax 50 mm (P50) and the new Pentax 18-55mm.
No AF adjustments. Hand-held photo in all cases


I can provide comparison photos done on a tripod with cable release, if that helps.

Attached Images

View Picture EXIF

File Date

 

View Picture EXIF

File Date

 

View Picture EXIF

File Date

 

View Picture EXIF

File Date

 

[OldNoob]

my 18-55kit lens seems to do best with fairly close objects, however it lags in sharpness in landscape images , though i understand there is reported quality variance with kit lenses.
A picture of a cat, the second file is 100% crop
Although the forum compression takes away sharpness so its hard to tell sometimes.

[Oldbayrunner]

I have been unhappy with my 18-55mm WR up until today. I finally took the time to use the auto fine focus adjustment on it. So with some time spent and adjusting it to a -8 due to severe front focusing with my camera I have finally got it to where it has really acceptable sharpness. Funny thing I tried to use the focus charts and was having a tough time getting it right. Then I decided to grab my ladies Minora, set it at a 45 degree angle and poof that worked. Now everything from near to far has really good focus.

Of course it's good weather replacement I ordered will arrive today, a new Sigma 17-70mm F2.8-4 DC Macro HSM (Contemporary) and also new Tamron SP AF 90mm F2.8 Di Macro.

I learned a lesson if one is not happy with a lens, try and find out what it's doing and if you can correct it then do that.

[crewl1]

In the sample shot of the sunflower sculpture the lens is set at f5.6 which is wide open at 55mm.
That is probably the worst case scenario for the kit lenses.
Shoot it at f8-f11 and pull it back a few mm's and it will improve (a little).

[paulh]

As Larry said, the kit lens is weak at 55mm wide-open. It performs best stopped down to f8-f/11 in good light. This was taken with the mk 1 kit lens at f/11 on a K10D.

[buessi]

You are correct - a wide open f-stop isn't a good way to establish proper depth of field. However, I wanted to give this a start by going fully auto, and then seeing how the lens performs. I have a good number of test photos I took (both inside and outside in bright light), all on tripod with cable release. I will post some of these since my first effort to establish front or back focusing didn't satisfy. For those I used two test targets I downloaded from dpreview.com, and then used a recommended method where you line up 5 AA batteries at a 45-degree angle to the camera. Each battery is the same distance from the other (3/4 in.) and the test should be conducted with the lens quite open. It did reveal differences among my several lenses. The third effort involved focusing on a more distant object at F8 and 100ISO. For each photo I made an AF adjustment by just one point. The only result I found was that the 50mm Pentax was sharpest at zero AF adjustment, while the two Sigma zooms (28-70mm and 28-80mm) checked in at -5 and +5, respectively.
This is a new world for me, since I had been using a Fuji Finepix 7000S prosumer cam for years that never had such a focusing problem, And so I got lazy and used a lot of auto settings, since this one handled it so well. A point well made: we used to look at 35mm film and slides that were crisp and sharp until they were enlarged. Now we have a large photo on the monitor and can see every imperfection quite large.


Next, I will post a small selection of sequential test photos to have you take a look at my contention that this particular kit zoom is less sharp than it needs to be (or that my lovely K-50 has a similar problem with autofocus.

---------- Post added 02-18-14 at 09:48 PM ----------

The four photos of the sunflower plaque also were hand-held. I didn't expect crystal sharp images (except for the one shot with the Pentax 50mm), but I also wanted to be assured that I didn't need to go manual every time a photo opp presented itself.

[MSL]

I'm not going to argue the kit lens is sharp - I struggle when I use it to get images I like and I mostly try to use it where it is weak (below 28mm) where I don't have any other lenses. But before I would draw any conclusions from your test I will raise one objection. The lighting on the last image of the 4 is totally different from the other 3 and in an awkward way where half the image is much brighter than the other half. Also, you wrote that the images are

The standard included here is a Sigma 28-80 mm on the K-50 body, then a Sigma 28-70mm (S70), the prime Pentax 50 mm (P50) and the new Pentax 18-55mm.

, but the EXIF data indicates that the second one is the 18-55 image. They were also all taken with different focal lengths. So you may be right that it isn't sharp, but I don't feel I can use these images to make any comparisons.

[Oldbayrunner]

I'm not going to argue the kit lens is sharp - I struggle when I use it to get images I like and I mostly try to use it where it is weak (below 28mm) where I don't have any other lenses. But before I would draw any conclusions from your test I will raise one objection. The lighting on the last image of the 4 is totally different from the other 3 and in an awkward way where half the image is much brighter than the other half. Also, you wrote that the images are , but the EXIF data indicates that the second one is the 18-55 image. They were also all taken with different focal lengths. So you may be right that it isn't sharp, but I don't feel I can use these images to make any comparisons.

To add, also different ISO, Apertures & Shutter speeds.

[buessi]

Correct - the test with the sunflower plaque was one that did not involve tripod, perfect lighting, or a cable release. And as one of you surmised, one photo was shot earlier in the morning, when the sun fell on that side of the fence. I needed to see if the handheld camera on auto, using the viewfinder, would give me different results. The small focal distance variation is thus understandable; it should not have affected results all that much.

My acid test here was shooting on tripod, with cable release, all 20 options of AF fine adjustment capable by the camera. The photos were taken in wan sunlight (it's Seattle in winter after all) at f11 and 1/100. Please compare these five and tell me your impressions. Having both vertical and horizontal lines (as well as the shingle detail) in the photo gave me some ideas where I should set the AF. The focus always was on the tower-like structure of this residence. Note also the sign on the bottom corner for image sharpness. Although reducing the file size considerably due to the forum limits, I believe some photos indicate which way I need to go with Fine Adjustment.

[stevebrot]

Please compare these five and tell me your impressions.

My impression is that the depth of field at f/11 is sufficiently deep that all five images at the point of focus look about the same.

Might I suggest doing your AF calibration at moderate distance (say 20x the focal length) using a tripod and a lens resolution test target. I suggest using a method similar to that described in this video (the "dot tune" method):

Youtu.be

Youtu.be

Steve

[buessi]

Thank you for the hint. This 18 min. video was very helpful (as were the comments by other forum members who recommended Live View to calibrate the kit lens (and others in my collection). From another website I did download the EIA Resolution Chart 1956 some time ago, but I found its use not as helpful as lining up batteries as recommended and then taking a series of photos of these at different settings. My error there was that I did that indoors, so the lens was pretty well wide open. See sample photo shot on tripod with Pentax kit 18-55mm.
Although zooms do have a deserved reputation for softer image generation, I always had been satisfied with my Sigmas - used often for macro photography - and my long-barrel Tamron. It would really disappoint me if I could not use a kit lens as intended, since I figured that these would be best matched to my new K-50.

[stevebrot]

Thank you for the hint. This 18 min. video was very helpful (as were the comments by other forum members who recommended Live View to calibrate the kit lens (and others in my collection). From another website I did download the EIA Resolution Chart 1956 some time ago, but I found its use not as helpful as lining up batteries as recommended and then taking a series of photos of these at different settings. My error there was that I did that indoors, so the lens was pretty well wide open. See sample photo shot on tripod with Pentax kit 18-55mm.
Although zooms do have a deserved reputation for softer image generation, I always had been satisfied with my Sigmas - used often for macro photography - and my long-barrel Tamron. It would really disappoint me if I could not use a kit lens as intended, since I figured that these would be best matched to my new K-50.

I appears that you have decided to forge your own path here. Good luck.

As a parting set of advice:

• Focus peaking in live view against a flat target is your best standard for the actual point of focus
• Moderate distance (a few meters) is best for resolution testing
• Before you write off a lens as bad, do an objective comparison (all settings and conditions the same) with a known good lens

Steve