[Mattox]

Staff note: This post may contain affiliate links, which means Pentax Forums may earn a small commission if a visitor clicks through and makes a purchase. If you would like to support the forum directly, you may also make a donation here.

---

Given that Nikon use Sony sensors what would be the rush to Nikon over Pentax?

Also, while Pentax have had some hardware issues with new cameras Nikon have a pretty bad recent history. Google about D600, D800 and D750 and you will know what I mean.

Lenses or flash system are more persuasive arguments to move to Nikon.

Very very true. If Nikon and Pentax use the same Sony sensors on full frame then maybe that answers my question. Wait. If Pentax can keep there price down to say the 750d price with a nice 24-70 2.8 at around 1000€ they've got me

[Vihmameister]

Very very true. If Nikon and Pentax use the same Sony sensors on full frame then maybe that answers my question. Wait. If Pentax can keep there price down to say the 750d price with a nice 24-70 2.8 at around 1000€ they've got me

Looking OP works, i would say that pentax/ricoh must jump over his pants to fulfill OP-s need to go ahead. There is nothing in current pentax lineup to go further.
No FF camera, no lenses. Is only big "IF" pentax will do: 1,2,3 ... and etc with good price.

From one point - yes FF in half of year or so, maybe is worth waiting and go another half of year will to get knowledge - what it really gives (and solve/sort out all errors developed with FF). But there is still no lens for it.
With no lens i mean - covered is mostly tele end- 70-200 and 150-450 line.
No FF UWA, no FF good standard lens. And do not start bully with DA this or DA* that. There is no lens in range 10-135 mm for full frame with real resolving power for landscape. No pentax lens go over 20-21 P-mpix, while canon seldom do above 30 p-mpix and nikon best glass go 40-44 P-mpix. smc DA 55 alone is not enough to cover FF. With most pentax crop lenses you can magnify till 30-50% from real size. Seldom 70-100% if shot as macro withs macro lens. No 100% printout with pentax lens on APS-c landscape. And Sigma does not help out here much too.

@ the moment best resolving power offers nikon glass on D810 followed closely by nikon D750. Decent metering, good AF and tracking (again no need for landscape). Looking pictures on OP-s web page i would say that single nikon 24-70 F2.8 - with stabilizer covers all OP-s needs. (yeah - there are new 50 megapixel canon -s, but they look more like freak-line, as they are lacking anything improved but resolving power over 5D mark III). Even 6D will be serious improvement over current K5 or K3 offer on landscape area.

one more thing on "IF"-s. looking pentax current lineup - no pentax camera has decent ISO handling. There is no such thing as real ISO performance above 1270, specially in poor light conditions, outdoors. You get nothing but ISO noise above 1250 in fog. As landscape shooter, this is serious drawback.
K5 does have some 1230 ISO limit , K3 1250-1270, 645Z sits in similar hole. Why would next FF perform much better in this area? I am very pessimistic in this area, as long i see on next FF that it really performs on ISO. I do not care how good it is indoors, taking selfies or birthday cakes.
What about K3, i would say it is very inconsistent camera. On some days it just excels, while some hours later it just can not make any decent picture, not having anything in focus on many pictures.
Some tell -nikon use same sony sensors as pentax. I see this more like problem, not as solution for my needs. And i proceed to think this way as long they prove opposite in practice.
Whatever they are lacking in engineering, problems persist in current lineup. I would say, that best cameras pentax has made are K20 and K5. There is no serious improvement after those ones. (have not used K-S2, but i bet it is as good or bad as K3 is).

So there is just 2 plain ways to go. You or stay and wait (as i do - because financial reasons) or you just do plain switch and can tell us how you improved after.

[Mattox]

I really know what you mean by inconsistent. In some situations the k5 for example has great results it performs extremely poor, and I am not talking about white balance and temperature. I am talking AF. Even under live view. It's not dark, you've got some decent light, you think you are precise on marking your AF point and then when you get home and see it in the big screen it's all out of whack, and yes to answer your q's. The lens is calibrated and tripod used. What is the most inconsistent is the schizophrenic AF. In many ways my kit lens and the K10 performed better

[Vihmameister]

I really know what you mean by inconsistent. It some situations the k5 for example has great results it performs extremely poor, and I am not talking about white balance and temperature. I am talking AF. Even under live view. It's not dark, you've got some decent light, you think you are precise on marking your AF point and then when you get home and see it in the big screen it's all out of whack, and yes to answer your q's. The lens is calibrated and tripod used. What is the most inconsistent is the schizophrenic AF. In many ways my kit lens and the K10 performed better

You description resembles exactly my experience - on what "inconsistent" means. Worse is, that it it behaves similar on full manual /MF drive also. And tripod does not help here sometimes. It's just or is in focus sometimes, or it is not. And you do not know that before looking results on big screen at home. Have tried different focus points, avoiding linear (not cross) points, tried different exposure area metering in combo with different WB settings. Still no clue why it sometimes does not focus, even on still meadow.

[bm75]

I don't totally rely on AF. Shooting landscapes with tripod on my k5 I always use MF (also MF lenses) and liveview magnification to check precisely what I'm focusing on. With k3Wifi support I suppose you'll be able to remote control your camera and check risults with a tablet or something so.
My son is a mtb biker and I always take lots of pictures during mtb races and I've taken my best shots with panning or rifle shooting, both with M or Tav mode with AF on or off, depends on situation. I NEVER used AF-C . AF is permanently setted to central point (with average central exposure measure) and I never used expanded AF. So having 39 or 51 focus points (that is in Nikon camera bodies) really doensn't matter at all. You will say my setting style is "old and basic", but it works, being me a non pro. My "static" images are 90% in focus with AF-S and Pentax gear, but I realized better risults are aquired with more personal skill and a clever study of the shooting conditions, then relying on the camera performance. So, what I'm telling is that , IMHO , the best thing is invest in your skill and trust more in that than relying on supposed features, especially of other brands (if you never used that brand's gear). Best regards.

[stevebrot]

A few notes regarding AF:

• AF cannot read minds
• Multi-point AF will not detect and limit focus to eyes only. The same is true for ears, feet, breasts, or buttocks. (see first point above)
• AF may or may not pick the subject over the background (see first point above)
• AF (even PDAF) depends on detail and subject contrast for accuracy, precision, and ability to acquire focus
• In low light near the specification lower limits, all AF systems will suffer degraded accuracy, precision, and ability to acquire focus
• PDAF precision (different than accuracy) is poor at apertures wider than f/2.8 when using the center area focus points*
• PDAF precision is poor at apertures wider than f/5.6 when using other than the center area focus points
• AF fine adjust cannot correct poor focus precision
• The AF technologies used by the various brands are proprietary and patent protected. If Nikon dominates in predictive AF, it is not because the other makers are stupid or lazy. More likely, it is because current technology favors the method Nikon has patented and other makers hands are tied until those patents expire.**
• PDAF time to lock is determined by both lens and body
• There is a trade-off between time and accuracy for focus acquisition. Rumor is that Canon does not confirm focus before lock.
• Good as some AF systems are, many (most) photographers who rely on AF for difficult subjects also rely on high frame rates as a form of insurance (spray and pray)
• Manual focus using focus confirm or catch-in-focus suffer from the same limitations of the underlying AF system
• Accurate focus is no guaranty of acceptable image sharpness

I have been fielding question and concerns about AF on this site for several years and the vast majority involve a lack of understanding of one or more of the above points. I won't make apologies for the Pentax AF system. It is what it is and I regularly suggest that Nikon or Canon may better suit a particular shooting style or expectation, though perhaps with some trade-off of features at a given price point.


Steve

* AF points capable of focus sensitivity at f/2.8 are generally not present and if there are fairly few in number and clustered toward the center. This is true across brands. The other AF points have a focus sensitivity of f/5.6 meaning that the ability to detect out-of-focus is the same at wider apertures as at f/5.6. What this means is that the probability of an OOF result with the K-3 and an f/2 lens is fairly high when shooting wide open using the center focus points and even higher with those to the sides.

** Patents are often worded somewhat broadly such that the feature itself (e.g. predictive AF) may be protected regardless of how it is implemented.

[Vihmameister]

Manual focus using focus confirm or catch-in-focus suffer from the same limitations of the underlying AF system Accurate focus is no guaranty of acceptable image sharpness

As i shoot 99% manual, often with focus confirm (specially when it is dark enough), only those points are valid most time. Inconsistent K3 is seen at apertures F5-F13 with different lenses. I mean good weather, strong light with F5,6 -F8 apertures (normally strongest area on most lenses) may have very variable results, depending program line you use. And yes - there is no option else but just to make several shots to hope one hit. Light can change fast and it usually does, sometimes you have no more light than for 3- 4 shots and that it. And IF only central points are capable to wide apertures, then this should be written also in manual. More problematic is fact, that very often there is no focus nowhere while camera thought it was in focus...

If this is just missing "knowledge and skill" then it is hard to explain why one has it while shooting with K-30, K-50, K-5 and lacks same skills with K-3?

[bm75]

Thanks Steve for the explanation about AF . my opinion is not that Pentax AF is bad, but simply that I have always to use my head, check for what the system is doing and eventually take manual control over that. the same applies to metering and autoexposure. that's with any brand or system. Thanks and best regards.

[stevebrot]

As i shoot 99% manual, often with focus confirm (specially when it is dark enough), only those points are valid most time. Inconsistent K3 is seen at apertures F5-F13 with different lenses. I mean good weather, strong light with F5,6 -F8 apertures (normally strongest area on most lenses) may have very variable results, depending program line you use. And yes - there is no option else but just to make several shots to hope one hit. Light can change fast and it usually does, sometimes you have no more light than for 3- 4 shots and that it. And IF only central points are capable to wide apertures, then this should be written also in manual. More problematic is fact, that very often there is no focus nowhere while camera thought it was in focus...

If this is just missing "knowledge and skill" then it is hard to explain why one has it while shooting with K-30, K-50, K-5 and lacks same skills with K-3?

I am not sure that I am reading your comment above accurately. This is my interpretation:

• You use mostly manual focus using a mix of both the viewfinder and focus confirm
• You use focus confirm more often when the light is dim
• You are getting poor results (soft images?) at the aperture settings where the lenses should be at their sharpest
• You were unaware that there is a difference in the AF point sensitivity since it is not mentioned in the published K-3 specs or user manual
• Your results vary according to program line
• The camera will frequently acquire focus when nothing in the frame is in focus
• You did not have these problems when shooting with the K-5, K-30, or K-50.

It is the last three points that are most troubling.

Here are a few bullet points with my comments:

• I also own and use mostly manual focus lenses
• I never use AF or focus confirm for critical focus with any of my lenses
• The stock focus screen is not particularly good for manual focus. The focus sensitivity is only about f/4. Likewise the apparent DOF is more than what is really present.
• Due to limitations of the stock screen (see above), I have an aftermarket screen with split-image focus aid
• I almost always get acceptable manual focus
• Focus confirm will often light when the split image is clearly not matched. In those cases, there would be missed focus if I took the camera's advice.
• Focus confirm uses exactly the same sensor as the AF system. The difference is that it uses the center focus point only. I suggest using the center point with focus confirm since it is cross-point and has f/2.8 sensitivity.
• For general shooting where focus is not critical, I often use my AF lenses and am usually satisfied with the results, even in low light
• I almost always use AF-S, center point only for AF
• Program line should not affect the AF or focus confirm performance directly, though use of wider apertures may make mild missed focus more obvious. Similarly, a longer exposure time might cause blurring due to camera/subject motion.
• Only a gross failure would result in missed focus across the full frame
• Even at narrow apertures (for greater DOF), there is still only one plane of focus. With high magnification (pixel peeping at full resolution) this is more obvious
• With a high resolution sensor and pixel peeping, missed focus is more evident than with a lower resolution sensor. The same is true for softness due to poor quality optics and camera/subject motion.

You are correct that the focus point sensitivity specs are not in the user manual, though they have been published by Ricoh/Pentax in their promotional materials and widely quoted by many reviewers. The diagram below was taken from the review on this site



It is hard to tell what is causing your poor results. The AF performance (focus confirm included) on your K-3 should be better than that of the other three cameras you listed. You might want to try some controlled shots on tripod using focus peaking in magnified live view (SR off). That will at least allow you to eliminate focus and camera motion as the cause.


Steve

[Mattox]

I am glad that this post has shifted slightly to AF concerns with Pentax, as it is the major problematic issue that we have with these bodies. Thanks for your input everyone with regard to it. Being a landscape photographer I do find that it is a huge issue. For obvious reasons, mainly landscape photographers shoot in low light situations (dawn or dusk) for better lighting results and due to poor lighting conditions the Pentax AF system struggles. As pointed out when using manual focus in landscape your can acquire better retults but I find that on the APSC system with such a small viewfinder and LCD it is so hard to get so accurate. The naked eye can't really fine tune that well within small detail. For me at the moment I rely on selecting my AF subject in liveview and shoot. I cannot see a better way to get better results at this stage.

[redcat]

Nikon D750 Digital SLR Camera Body Brand New Full Frame DSLR 1 Year Warranty | eBay <= buy now if you really want

[bm75]

You want him to jump ship ?!!!!
For static objects or landscapes I would switch to Sony a7r II that will be on the market in Italy soon. More mpx, more DR, good AF and 5 axis SR (and its viewfinder can compensate for darkness/brightness making really more simple focusing in the dark) . If the Op wants more, perhaps he has to seriously consider Medium Format cameras. No way Nikon d750 would be more performant than a7r II for landscape and , as a plus, if he wants, he can put EVERY KIND of lens into Sony mirrorless cameras via adapters. I say twice: want more for landscapes and don't trust in Pentax?Go to SONY !!!
I'll wait for the new Pentax FF !!!!!

[Vihmameister]

I am not sure that I am reading your comment above accurately. This is my interpretation:

Very thorough answer, thank you ahead for that.
You interpretation sounds pretty correct. Worry-some are last 2 points. I used to be 70-90 % accurate with K-30 (my previous main camera). Now with k-3 i am barely at 25-30% accuracy. (I try to avoid K-3 s¤%& words)

The stock focus screen is not particularly good for manual focus. The focus sensitivity is only about f/4. Likewise the apparent DOF is more than what is really present. Due to limitations of the stock screen (see above), I have an aftermarket screen with split-image focus aid

How big is this - aftermarket screen? What is it, some 8-10" tablet? How you carry or connect it? while it can be OK for macro shooting and maybe even for tripod-landscape - waiting for moment - this style is absolutely nonsense for guy like me bicycling to places and walking 5-15 km in rampant or bog area. How can one use such thing for birding? My point is - decent on-camera focus system alone should be sufficient to make picture, without any aids.

Focus confirm will often light when the split image is clearly not matched. In those cases, there would be missed focus if I took the camera's advice.

Sometimes this phenomenon happens to me too - but what camera advice here?? I shoot almost all pictures via optical viewfinder. Not using live view -because this one was more inaccurate for me as focus confirm. Focus confirm method did work very well with K-30 and K-50. Simple caught focus sound and red dot on focus point were reliably sufficient with them. Didn't matter which focus point you used.

Focus confirm uses exactly the same sensor as the AF system. The difference is that it uses the center focus point only. That center point is cross-point and has f/2.8 sensitivity.

This one really confuses me. let's say, i am taking landscape in park, and there is close object in left lower area of frame and i want to keep focus lock on that object By reading you statement, that means while i set focus via appropriate lower left point on focus area, - focus confirmation is still done via central 3 points only? Sounds mad.
In reality, in situation above i often get what i want - meaning that point is in focus and sharpest. This is when usually K-3 does not fail.
But it fails very often, if there is no such strong object in focus area and you do not want to move angle so much to have one. Lets say, hayfield with forest far on horizon. Hay-field in evening fog. Or bird sitting on hay, reed-beds and etc. Here you miss often. Or shooting over plants level under the trees - with F6,3 -11 it may fail. or it misses when focus point is far away.
Problem is actually focus locking. I use manual regime with shutter releases anytime - is locked or not setting. That because hayfield tend not to stand still. Nature is not still most time at all.
and yes, shots may be quite long - i can handheld 1/13 1/15 with ease and it was enough with K-30. With K-3 even tripod with remote shutter does not help and it's better to shoot near 1/30-1/60 at least. If you ask why so slow times, then - forest bog and such areas before and after dawn. You do not want to have High ISO here, and even with 640-1000 you still fall quickly to this time frame - if not using tripod.
But i am not complaining inability to hold, but failure to focus outdoors: against bright light, in good light, on meadow barely moving and etc.
(while i can make very good results at night-time)

With a high resolution sensor and pixel peeping, missed focus is more evident than with a lower resolution sensor. The same is true for softness due to poor quality optics and camera/subject motion.

As much i understand - to have anything meaningful on frame after - focus MUST lock before/at shot moment. If you press shutter without camera "caught focus", there is just muted or blur image, no focus anywhere. And here is point of K-3 failure. It says it is focused, but is not. And it happens in good conditions too. (And i am not speaking pre- or back focus issues here.)

I almost always use AF-S, center point only for AF

Hmm, i tend to use AF-C in manual. Should try out AF-S in manual.
But i still refuse only central point usage, otherwise all this AF system is just plain junk.

Program line should not affect the AF or focus confirm performance directly, though use of wider apertures may make mild missed focus more obvious.

i mostly do not have "unexpected" focus issues with narrow DOF - as wide apertures give. But with auto program line it happens sometimes. So i use Normal or Macro regimes only.

With a high resolution sensor and pixel peeping, missed focus is more evident than with a lower resolution sensor. The same is true for softness due to poor quality optics and camera/subject motion

What some call pixel peeping, i call hi-quality printout ability. For family or travel album 30-50 % magnification is sufficient. For 100% size printout - not.
And here we are more at particular lens and sensor complex separation limits, than focusing. With current DA lens-line and 24 Mpixel APS-c pentax is not exactly there yet (and nikon D810, D750 and even D610) are km-s ahead on landscape. Canon 6D separates also far better on its 20 MP than K-3... (having only one meaningful cross-focus point at center).

Overall i try out this AF-S thing and lets see.


---------- Post added 07-27-15 at 12:17 PM ----------

You want him to jump ship ?!!!!
No way Nikon d750 would be more performant than a7r II for landscape and , as a plus, if he wants, he can put EVERY KIND of lens into Sony mirrorless cameras via adapters. I say twice: want more for landscapes and don't trust in Pentax?Go to SONY !!!
I'll wait for the new Pentax FF !!!!!

about sony: Camera Database - DxOMark

What sony- were is that sony???
D750 is one of the best landscape cameras for not too high price. (if you think dixo is just pixel-peeping with no real world contact - look flickr user tizziano results with D610 and D750 or do your own search on web).

And if there is no pentaprism viewfinder - there is NO camera.

[bm75]

Ok guys....we could say we don't like Pentax AF but some words have to be said before that conclusion.

1- for shhoting single frames AF-S is the setting, being said that AF-C is for tracking. So, always use AF-S for single or rifle shots without predictive purposes to gain better Af performance.

2- the optical viewfinder always suffers from a certain degree of front/backfocus. Only liveview offers you real sensor image . No way it could be otherwise with OVF. if you like EVF (like in fuji or sony cameras) this would equal liveview and give you a real image.

3-on charts k3 and canon 6d equals in -3 Ev AF capability . for knowing what that means, check ISO 100 EV charts to find what f\ and exposure it means (very low level). you have to check your shooting conditions and if you calculate your shooting means -5Ev there's no way any AF system would be useful in thatcondition.

4- no way a 20 mpx sensor on FF will give better sharpness than a 24 mpx on APS C (talking about simple resolution - check image dimensions) given an adequate resolution power of the lenses. More mpx means more blur appearence (nikon d800 users know that) and this means you have to be really careful in setting the camera and using a tripod and mirror lock up for the most of the shots. in this regards better have a Sr on the sensor (Nikon hasn't). Ckeck for blur before saying AF performance is low

4- I was a Canon user and , trust me, Canon Af is really good but you'll always need some MF operation or adjustment in low light conditions. And I don't regret about my choice of switching to pentax in the matter of AF.

5- I don't try to apologize anything about Pentax, just trying to point out some technical questions that could affect any brand AF performance but just due to incorrect use of the camera settings.

If you're about to make some trials, let as know your findings. Best regards. Matteo

[aleonx3]

I am not sure that I am reading your comment above accurately. This is my interpretation:

• You use mostly manual focus using a mix of both the viewfinder and focus confirm
• You use focus confirm more often when the light is dim
• You are getting poor results (soft images?) at the aperture settings where the lenses should be at their sharpest
• You were unaware that there is a difference in the AF point sensitivity since it is not mentioned in the published K-3 specs or user manual
• Your results vary according to program line
• The camera will frequently acquire focus when nothing in the frame is in focus
• You did not have these problems when shooting with the K-5, K-30, or K-50.

It is the last three points that are most troubling.

@Steve, good summary here... I don't want to join into arguments about inadequate AF in different Pentax models...but IMHO those points are tie to user errors more than anything else. I also think that k-5 or even k-5IIs are still a bit behind compare to other brands, but most the problems have been addressed in k-3.

---------- Post added 07-27-2015 at 09:34 AM ----------

Hmm, i tend to use AF-C in manual. Should try out AF-S in manual.
But i still refuse only central point usage, otherwise all this AF system is just plain junk.

Do you always use AF-C? Are you using an old legacy manual focus lens? You don't have to use the center focus point, but if your lens is an old lens, then you have no choice but the center point. I have used the D750, and 6D from a close relative and friend, sure they can fire several shots like my k-3, but are they all in focus? the answer is no. With my k-3 I have more keeper....