em-1 vs Pentax

So you haven't actually used C-AF or tracking focus on any of the cameras you're debating? I'd say tracking focus on my EM5 is more useful than my K5 was, although it's a far cry from the D800 which is amazing with the right lens.
Human bias is very strong. People who leave a system are apt to denigrate it in order to justify their choices. I'll stick with FNAC Lab results, thanks. I have not used an Olympus, you haven't used a K-5 II or K-3. I base my opinions on many comments I've read in forums and reviews. From the DPR OM-5 review for example:

"Focus in good light is impressively fast and continues to be pretty quick until the light gets very low. In situations where you're using the camera's highest ISO settings because you have to, you'll find the focus spends a bit of time hunting, and you'll have to start thinking about the contrast level of your target. Overall though the EM-5 turns in a good performance, especially by mirrorless standards.

However, despite Olympus' claims about improvements in continuous autofocus and subject tracking capability, we found the system too slow and unpredictable to develop much enthusiasm about. If you're careful about making sure the camera has locked onto a subject, you can expect to get a couple of sharp shots out of a burst but the results are patchy and, for the kinds of 'grab the moment' shooting you'd want it to work for, the results simply aren't reliable enough. The EM-5 is far from unusual in this respect (Continuous AF isn't a strength of contrast-detection AF), but for such an expensive model, and one for which Olympus is pushing the continuous AF capabilities, it's a disappointing result.

Tracking is a particular disappointment. If you specify an AF point, the camera will usually 'lock-on' to the correct subject but it gets very easily distracted. No matter how distinct the subject might appear (being the only red item in the frame, for instance), the camera will often decide it's much more interested in the background after a couple of frames. This, combined with the continuous AF's hunting, means you simply can't rely the system to get your shots in focus."


I'm not aware of any improvements in CDAF with the EM-1. It has PDAF, which is an improvement, but only with 4/3 lenses and even then it is not up to DSLR standards.


I had looked at some of those tests before I bought my K5 and used it as some kind of reassurance/justification, but one massive variable that is often overlooked is what lenses are being used for the tests? An AF test with a kit lens or the vast majority of primes (u4/3 being a major exception) is useless if you're testing speed. For most brands you have to buy top dollar lenses to utilize a body's AF potential - with Pentax the most expensive zoom lenses are often amongst the slowest for AF.
It's true that some of the SDM lenses are slow, notably the DA*50-135 and DA*55. OTOH, they have excellent IQ and are unique. No one else has APS-C lenses that show the classic film FOV, 75-200 & 85mm on 1.5 crop. These are mainly portrait lenses, fast focus is not as important as accurate focus. The lenses focus, then check focus, so you hear zzt, zzt, instead of zzt. It takes longer, but accuracy check vs. speed is a viable choice.

The DA*300 f4 focusses faster than the Sigma HSM 150-500mm, so it's not true that all SDM lenses are slow (I've owned both). It's true that Canon & Nikon lenses with ring motors have faster AF, but I'm not in the market for $5000-10000 lenses.
Face detection is definitely a gimmick - at least on DSLR's using live view. The only thing gimmicky about it on an Olympus is that you can shoot wide open with the 45mm F1.8 and repeatedly nail sharp eyelashes with the ~depth of field of the DA*55 @ F2.5 or F2.8, shot after shot after shot. And the gimmick works well in manual mode too ;) I shoot manual mode only indoors with Auto ISO (effectively TAV mode) with bounce flash, 90mm equiv focal length @ F1.8 and only have to worry about lighting and composition - which is what photography is all about IMO.
Larger DOF with m4/3 helps here.

I've used focus peaking with a Carry Speed LCD VF on a K-01. It's a lot larger than an EVF, and I didn't find it any more accurate than an OVF. I can see far more clearly through a pentaprism than a miniature TV. I haven't tried an Olympus EVF, but I have tried the NEX-7 and it was execrable; like watching a video game. I felt very disconnected from the scene and detail was very coarse.
I did miss ultra shallow depth of field photography though with shorter focal lengths hence the D800 (also wanted 3D AF which is killer if you've never used it before - say goodbye to focus and recompose). But even though I have the excellent 85mm F1.8 for the Nikon, more often than not I reach for the Olympus with the 45mm F1.8 when I'm doing head-shots...it's that satisfying.
I don't doubt this at all. It's not necessary to have the greatest technology for every shot. For many uses, my Q is fine, m4/3 is fine, APS-C is fine. In more challenging scenarios, different capabilities can shine. I want my DSLR for BIF, and m4/3 can't do it. Since I want pixels on the bird, 24mp APS-C suits me more that Full-Frame.
I will agree with you on one thing though, the K3 is the best camera for you and that's all that really matters. If I wasn't shooting little kids the majority of the time and was still travelling and spending as much time out in the mountains as possible, I'd likely still be in the Pentax camp as well.
I'm not sure what the mountains have to do with your choice, unless you're talking about a small weight savings. No Oly can compare with a K-5 II or K-3 for shooting little kids in dim light.

Only a DSLR is an all-purpose body. Other systems may more than meet the needs of their users. Not everyone wants to shoot wildlife in dim conditions. I need a DSLR, and the K-3 looks to be the one I've been waiting for. I will know soon, mine arrived today, and it's good'un.
 
I've owned 4 Pentax DSLRs and the continuous AF/Tracking was laughable, made even worse with SDM lenses. In fact that was a major reason I sold most of my Pentax gear and switched to Nikon and Olympus.

Olympus single shot AF on my EM5 blows away the K5 I had both for speed and accuracy - no contest.
Pentax made an unfortunate decision to use sensor shift AF on a DSLR. The captured image is stabilized but the viewfinder, AF and metering are all left with an unstabilized image to work with; this can't help but impact the ability of the AF to properly lock onto the subject quickly. The longer the focal length being used, the more jittery and 'unstable' the image is - and it is at longer focal lengths that AF tracking comes into play for things like sports.

It's like if Lexus started putting air-cooled engines in their cars to do away with cooling system maintenance - that 'benefit' cannot make up for the functional tradeoffs that result.

There's a reason the 'big boys' use optical IS on their DSLR's and it has nothing to do with making more profits on the lenses. It's because it's the logical way to do it.
I think Canon and Nikon use lens OIS because they were working on it in film days when it was the only possible solution. You can't do IBIS with film cameras.
 
Pentax made an unfortunate decision to use sensor shift AF on a DSLR. The captured image is stabilized but the viewfinder, AF and metering are all left with an unstabilized image to work with; this can't help but impact the ability of the AF to properly lock onto the subject quickly. The longer the focal length being used, the more jittery and 'unstable' the image is - and it is at longer focal lengths that AF tracking comes into play for things like sports.
So you're still flying this nonsense? The sensor stabilizes the image, in other words, it is less jittery than a non-stabilized system. Do you realize that Pentax shooters can disable IBIS if they choose to? Do you think they do that? Do Olympus shooters disable their stabilization? Or do they brag about how great it is? Your party line is ludicrous.
 
Pentax made an unfortunate decision to use sensor shift AF on a DSLR. The captured image is stabilized but the viewfinder, AF and metering are all left with an unstabilized image to work with; this can't help but impact the ability of the AF to properly lock onto the subject quickly. The longer the focal length being used, the more jittery and 'unstable' the image is - and it is at longer focal lengths that AF tracking comes into play for things like sports.

It's like if Lexus started putting air-cooled engines in their cars to do away with cooling system maintenance - that 'benefit' cannot make up for the functional tradeoffs that result.

There's a reason the 'big boys' use optical IS on their DSLR's and it has nothing to do with making more profits on the lenses. It's because it's the logical way to do it.
Never thought of that before but it's an interesting point.
No it's not, it's ridiculous nonsense. Think about it now. Which is more stable, an IBIS system or a non-stabilized system? The sensor is moving to counteract shake, in other words, it is more stable than no IS, not less. Have you ever heard anyone complain about auto-focus or metering for ILIS vs. non-stabilized? I haven't. If you're shaking so badly that metering and AF are affected, you have a serious problem. Don't take the shot, it won't be any good.
 
Hi LTZ470,

I use focusing peaking on my K-01 all the time. I even have it on with AF lenses just to verify the point of focus before capture and to free the one customization button too. Not that CDAF is not accurate but I am not sure of the size of the CDAF sensor. Sometimes, after locking focus I may move just a little closer or further away in close up shoots.

For me, MF lenses work much easier with Catch in Focus but only when I use the central AF point only. For all other AF points and for weird shooting positions LV+FP is a great asset that I do miss on my K5ii.

K3 seems to sum it all in one camera :)

Great shoots, very detailed! Did you use this lens?

http://slrgear.com/reviews/showproduct.php/product/1567/cat/14

It looks fantastic.
 
Pentax made an unfortunate decision to use sensor shift AF on a DSLR. The captured image is stabilized but the viewfinder, AF and metering are all left with an unstabilized image to work with; this can't help but impact the ability of the AF to properly lock onto the subject quickly. The longer the focal length being used, the more jittery and 'unstable' the image is - and it is at longer focal lengths that AF tracking comes into play for things like sports.
So you're still flying this nonsense? The sensor stabilizes the image, in other words, it is less jittery than a non-stabilized system.
You missed his point. The sensor on Pentax is stabilized, but the light that is reaching the AF sensor (different location) is not stabilized. Neither is the light that is reaching you via OVF. The first one (jittery light that reaches AF sensor) affects AF performance. The second one (jittery light that reaches OVF) affects viewing experience.

Optical stabilization doesn't have that issue.

Neither does EM1 as the CDAF is done by the main sensor (which is stabilized) and also the view in EVF is stabilized.
 
Last edited:
Pentax made an unfortunate decision to use sensor shift AF on a DSLR. The captured image is stabilized but the viewfinder, AF and metering are all left with an unstabilized image to work with; this can't help but impact the ability of the AF to properly lock onto the subject quickly. The longer the focal length being used, the more jittery and 'unstable' the image is - and it is at longer focal lengths that AF tracking comes into play for things like sports.

It's like if Lexus started putting air-cooled engines in their cars to do away with cooling system maintenance - that 'benefit' cannot make up for the functional tradeoffs that result.

There's a reason the 'big boys' use optical IS on their DSLR's and it has nothing to do with making more profits on the lenses. It's because it's the logical way to do it.
Never thought of that before but it's an interesting point.
No it's not, it's ridiculous nonsense. Think about it now. Which is more stable, an IBIS system or a non-stabilized system? The sensor is moving to counteract shake, in other words, it is more stable than no IS, not less. Have you ever heard anyone complain about auto-focus or metering for ILIS vs. non-stabilized? I haven't. If you're shaking so badly that metering and AF are affected, you have a serious problem. Don't take the shot, it won't be any good.
 
Pentax made an unfortunate decision to use sensor shift AF on a DSLR. The captured image is stabilized but the viewfinder, AF and metering are all left with an unstabilized image to work with; this can't help but impact the ability of the AF to properly lock onto the subject quickly. The longer the focal length being used, the more jittery and 'unstable' the image is - and it is at longer focal lengths that AF tracking comes into play for things like sports.

It's like if Lexus started putting air-cooled engines in their cars to do away with cooling system maintenance - that 'benefit' cannot make up for the functional tradeoffs that result.

There's a reason the 'big boys' use optical IS on their DSLR's and it has nothing to do with making more profits on the lenses. It's because it's the logical way to do it.
Never thought of that before but it's an interesting point.
No it's not, it's ridiculous nonsense. Think about it now. Which is more stable, an IBIS system or a non-stabilized system? The sensor is moving to counteract shake, in other words, it is more stable than no IS, not less. Have you ever heard anyone complain about auto-focus or metering for ILIS vs. non-stabilized? I haven't. If you're shaking so badly that metering and AF are affected, you have a serious problem. Don't take the shot, it won't be any good.

--
Dan
Jitter will not affect AF speed or functionality
Sure it can have negative effect on AF and also on metering.

EM1's IBIS of course would be immune, as the AF, metering, and also EVF image all happens via the main sensor, which is stabilized.
 
So you haven't actually used C-AF or tracking focus on any of the cameras you're debating? I'd say tracking focus on my EM5 is more useful than my K5 was, although it's a far cry from the D800 which is amazing with the right lens.
Human bias is very strong. People who leave a system are apt to denigrate it in order to justify their choices. I'll stick with FNAC Lab results, thanks. I have not used an Olympus, you haven't used a K-5 II or K-3. I base my opinions on many comments I've read in forums and reviews. From the DPR OM-5 review for example:

"Focus in good light is impressively fast and continues to be pretty quick until the light gets very low. In situations where you're using the camera's highest ISO settings because you have to, you'll find the focus spends a bit of time hunting, and you'll have to start thinking about the contrast level of your target. Overall though the EM-5 turns in a good performance, especially by mirrorless standards.

However, despite Olympus' claims about improvements in continuous autofocus and subject tracking capability, we found the system too slow and unpredictable to develop much enthusiasm about. If you're careful about making sure the camera has locked onto a subject, you can expect to get a couple of sharp shots out of a burst but the results are patchy and, for the kinds of 'grab the moment' shooting you'd want it to work for, the results simply aren't reliable enough. The EM-5 is far from unusual in this respect (Continuous AF isn't a strength of contrast-detection AF), but for such an expensive model, and one for which Olympus is pushing the continuous AF capabilities, it's a disappointing result.

Tracking is a particular disappointment. If you specify an AF point, the camera will usually 'lock-on' to the correct subject but it gets very easily distracted. No matter how distinct the subject might appear (being the only red item in the frame, for instance), the camera will often decide it's much more interested in the background after a couple of frames. This, combined with the continuous AF's hunting, means you simply can't rely the system to get your shots in focus."


I'm not aware of any improvements in CDAF with the EM-1. It has PDAF, which is an improvement, but only with 4/3 lenses and even then it is not up to DSLR standards.
I had looked at some of those tests before I bought my K5 and used it as some kind of reassurance/justification, but one massive variable that is often overlooked is what lenses are being used for the tests? An AF test with a kit lens or the vast majority of primes (u4/3 being a major exception) is useless if you're testing speed. For most brands you have to buy top dollar lenses to utilize a body's AF potential - with Pentax the most expensive zoom lenses are often amongst the slowest for AF.
It's true that some of the SDM lenses are slow, notably the DA*50-135 and DA*55. OTOH, they have excellent IQ and are unique. No one else has APS-C lenses that show the classic film FOV, 75-200 & 85mm on 1.5 crop. These are mainly portrait lenses, fast focus is not as important as accurate focus. The lenses focus, then check focus, so you hear zzt, zzt, instead of zzt. It takes longer, but accuracy check vs. speed is a viable choice.

The DA*300 f4 focusses faster than the Sigma HSM 150-500mm, so it's not true that all SDM lenses are slow (I've owned both). It's true that Canon & Nikon lenses with ring motors have faster AF, but I'm not in the market for $5000-10000 lenses.
Face detection is definitely a gimmick - at least on DSLR's using live view. The only thing gimmicky about it on an Olympus is that you can shoot wide open with the 45mm F1.8 and repeatedly nail sharp eyelashes with the ~depth of field of the DA*55 @ F2.5 or F2.8, shot after shot after shot. And the gimmick works well in manual mode too ;) I shoot manual mode only indoors with Auto ISO (effectively TAV mode) with bounce flash, 90mm equiv focal length @ F1.8 and only have to worry about lighting and composition - which is what photography is all about IMO.
Larger DOF with m4/3 helps here.

I've used focus peaking with a Carry Speed LCD VF on a K-01. It's a lot larger than an EVF, and I didn't find it any more accurate than an OVF. I can see far more clearly through a pentaprism than a miniature TV. I haven't tried an Olympus EVF, but I have tried the NEX-7 and it was execrable; like watching a video game. I felt very disconnected from the scene and detail was very coarse.
I did miss ultra shallow depth of field photography though with shorter focal lengths hence the D800 (also wanted 3D AF which is killer if you've never used it before - say goodbye to focus and recompose). But even though I have the excellent 85mm F1.8 for the Nikon, more often than not I reach for the Olympus with the 45mm F1.8 when I'm doing head-shots...it's that satisfying.
I don't doubt this at all. It's not necessary to have the greatest technology for every shot. For many uses, my Q is fine, m4/3 is fine, APS-C is fine. In more challenging scenarios, different capabilities can shine. I want my DSLR for BIF, and m4/3 can't do it. Since I want pixels on the bird, 24mp APS-C suits me more that Full-Frame.
I will agree with you on one thing though, the K3 is the best camera for you and that's all that really matters. If I wasn't shooting little kids the majority of the time and was still travelling and spending as much time out in the mountains as possible, I'd likely still be in the Pentax camp as well.
I'm not sure what the mountains have to do with your choice, unless you're talking about a small weight savings. No Oly can compare with a K-5 II or K-3 for shooting little kids in dim light.

Only a DSLR is an all-purpose body. Other systems may more than meet the needs of their users. Not everyone wants to shoot wildlife in dim conditions. I need a DSLR, and the K-3 looks to be the one I've been waiting for. I will know soon, mine arrived today, and it's good'un.

--
Dan
Either way I think a Pentax user debating with an Olympus user about which system has superior Continuous/Tracking AF is kind of like Seattle Mariners fans debating with the Chicago Cubs about who has the better odds on the World Series next year :) Like your excerpt said the EM5 C-AF is kind of choppy but the on screen tracking box is pretty cool but it will always get at least a few shots out of a series in decent light and they will always be sharp and in focus. If I know I'm going to be needing the feature though, I grab the D800, especially in low light. I'd be willing to be my entire camera collection that the K3 with any lens will not touch that in tracking mode :)

And you don't need to spend 5k+ on to get a very fast Nikon lens, in fact the new 70-200 F4 is not much more expensive than the aging DA* 50-135! I agree human error can effect judgement because of bias, but next time you're in a camera shop ask for the D800 with the affordable 70-200 F4 and see if you're still biased towards birding in low light with an SDM lens...

My hiking comment was pro-Pentax - if I didn't have fast subjects to shoot (ie pre-kids) I probably couldn't have justified the switch from Pentax and would likely be drooling over the K3 right now (and preferred it to the EM5 as the IQ is definitely better) wondering how long I'd have to wait for the beta testers to verify the camera worked the way it should. I was usually pretty happy with Pentax when it came to landscape shooting, extremely happy with static scenes, not so happy when I had a quick few seconds to shoot a bear, elk, mountain goat etc.
 
Either way I think a Pentax user debating with an Olympus user about which system has superior Continuous/Tracking AF is kind of like Seattle Mariners fans debating with the Chicago Cubs about who has the better odds on the World Series next year :)

LOL! OK, but I'm expecting the K-3 to put that to bed. I should know by Monday. My K-3 arrived yesterday, no opportunity to test tracking yet.
 
I follow the EM-1 threads from time to time, and there seem to be three areas of concern: 1) focusing on horizontal lines (camera has to be tilted 30-45 deg to get a lock)
I switched from Pentax to M43 3 years ago (to GH2, to GH3, and E-M1 now), I am yet to find a case when I was really trying to focus on something w/ only horizontal lines within the focusing area... while it is true, in real life shooting its not not happening.
 

Keyboard shortcuts

Back
Top