[DarrenCax]

After having troubles with my K-r, I decided to upgrade to a K-5 in hopes these focusing issues have been resolved. However, now that I've gotten a chance to fully test the camera, I'm seeing almost identical results as I saw with my K-r... So people can see what I'm talking about and confirm I am getting the affected issues, I've cropped some pictures for viewing here.

All pictures were taken using a tripod and my remote. SR was automatically disabled. Lighting was fairly dim (single incontinence lamp). Using latest firmware (1.13). Aperature Priority mode, ISO-400. Spot focus was used and the lens was set to infinity then allowed to refocus before each picture. Matrix meetering. Focal poaint was the center of the quarter.

First couple are from my Pentax 35-70mm lens. Results are pretty similar but the liveview tends to be consistently sharper to some extent.





Now onto the next lens, my 40mm pancake. Very clear difference between the two. These results are consistent too.







I really don't know what to do. These results are nearly identical to the K-r. I really don't want to sell all my lenses and such but I'm seriously fed up with Pentax now. Please, tell me I'm doing something wrong and how I can resolve this with either camera!

[yucatanPentax]

Since so many others have no issues getting sharper photos than those from both the K-r and K5, it is hard for me to believe that all of Pentax is to blame.

I see you've posted some of the information with your photos, but cannot see all the EXIF info. You wouldn't think shutter speed should matter a great deal because of the tripod, but the tripod is another variable in play and the shutter speed is not listed. Are these 100% crops, or what? There's too much missing information about the images so far.

incontinence lamp

By the way, I'm a bit worried about your lamp.

[rawr]

I'm seeing almost identical results as I saw with my K-r...

Pls define your problem better. I'm not sure what you are talking about.

[imtheguy]

I feel your pain but believe its solvable.
First, and assuming the shots are not all wide open, do you get good sharpness if you have strong light? If not, have you tried adjusting the FF/FB for each lens on the K-5? Need to do that and I believe there is a forum tutorial if you are not familiar with the process.
Next, if all these shots are wide open please stop down at least 1 full stop and try it under strong light before the weak light again. If its sharp in one and not the other then at least we know the lens is good as well as the FF/BF alighnment and can eliminate that from consideration.
Many people have complained early in the K-5 life about FF (front focusing) under dim tungsten. I do not shoot under tungsten so cannot confirm if the issue still exists with FW 1.13. Someone else will likely jump in on that.
Yes, Live View and viewfinder will have differences depending on the alignment you can adjust in the menu. They use different focusing methods and hardware.
Try shooting the quarter at about a 30-45 degree angle as it lays on a yardstick. That way you can see if the true focus lies somewhere else but does exist! Then its a matter of jumping into the menu and running through the lens alighment procedure mentioned above.

Sorry for the disorganized thoughts (its 3am here) but I have have been where you are now and come out fine. Please try some suggestions and repost till we get it right.

[JohnX]

You clearly expect better, but why? I for one wouldn't be attempting to photograph anything under the conditions you describe, and if I did I wouldn't be surprised at the poor results.

If you need to shoot under such poor conditions, then maybe you're right, and Pentax is not for you.

If you don't need to shoot under such poor conditions, DON'T!

Also, there's been an interesting thread on here recently about how accurate the focus spot(s) actually are (or aren't). Worth a search and a read.

If you use any equipment outside the conditions it was designed for you shouldn't be surprised if it doesn't perform as well as you'd like it to.

[jezza323]

I believe the issue is likely related to the way you are testing. As JohnX mentioned, the focus point is quite LARGE, and could be picking any item inside the area it covers to focus on. The "Test" you have set up there will not take this into account at all

[jmg257]

The shots, especially the 2nd set, seem to be taken at an angle. What item/edge are you trying to focus on? Which edge is actually being focused on?

---

Last edited by jmg257; 03-06-2012 at 08:49 AM.

[Wheatfield]

It would help to know the actual EV being shot at. The AF is only good to a certain amount of available dark, and after that it's going to fail. The K5 was also pilloried when it came out because of poor AF performance under tungsten light, and while subsequent firmware updates have attempted to address the problem, my experience has been that this has not worked under all conditions. I still get huge front focus with my modeling lights in the studio.

[jbinpg]

The focus spot is grabbing the high contrast text either ahead or behind your focus point. The size of the center focus spot is larger than you think.

[einstrigger]

I just want to add that sometimes there is no avoiding AF adjustments for certain lenses. Most of the lenses I have were spot on, while others needed some fine adjustment. Similarly, for me the DA 40mm was the worst of the bunch. Have you tried other lenses besides the 40 and 35-70? And are your adjustments severe (i.e. +/- 10)?

[DarrenCax]

Let me get something straight before I go and bash my head against a wall... Very simply put, contrast detection focuses based on what the image appears like through the camera. The actual focal setting has no implication on the picture. Phase Detection uses sensors to determine distance and set the lens accordingly. Therefore, a mis-calbrated lens could focus fine within contrast detection but not phase detection...

Currently my only other lens is my 70-300 but I'm going to test a few things and determine what's what.... I'm beginning to think/realize the 40mm has been the issue the whole time...

[jmg257]

Phase Detection uses split images of the subjects on pairs of sensor arrays to determine the defocus amount. The defocus amount is used to determine how much the focusing lens has to move, so how much the AF motor has to spin, so how many encoder pulses have to be fed back to the CPU to countdown to zero. A focusing width value is determined to know if the defocus amount will be enough to get focus lock. Lots of places in there for calibration errors, including that the adjustable AF sensor (module) isn't the same distance from the lens as the image sensor, sub-mirrors, AF optical system, camera and lens build tolerances, motor/linkage/gear block backlash, light levels, contrasts, focal lengths, aperatures, etc. etc..


But first it sure helps to know exactly what the AF sensor is focusing on.