Review – Pentax K-5
Note: The camera tested in this review was loaned to EtL by B&H Photo & Video store. To thank them for their generosity and allow us to receive more products for review, please consider purchasing your camera equipment through our affiliate B&H link (also found on the right sidebar). Thanks!
Pentax K-5 Main Specifications
What’s in the Box
Ergonomics and Buttons
RAW File Quality
The K-5 Sensor Stains
When Pentax introduced the K-7 in July 2009 they presented it as a camera worthy of its famous line of Limited lenses. It was noticeably smaller than its predecessor the K20D and represented not only an upgrade, but a change in camera philosophy, an attempt by Pentax to really stand out from the crowd by producing the smallest weather-sealed pro-level camera on the market. Pentax’s plans were thwarted by the camera’s sensor, a new version of the 14.6MP Samsung sensor that had already been used in the K20D. While this sensor had been considered good when it first appeared in early 2008, its high ISO capabilities were not up to scratch when compared to cameras released in 2009 around the same time as the K-7. True, at ISO below 800 in performed remarkably well and landscape shooters appreciated what the K-7 could offer them, but many other Pentax users felt let down, with the result that a lot of K10D and K20D shooters didn’t upgrade. Adding insult to injury, a few months later in September 2009, Pentax released the entry-level K-x, featuring a then new 12MP Sony sensor which turned out to have superlative high ISO image quality, prompting many Pentaxians to buy a K-x instead of the K-7. Shortly thereafter threads began to appear across various internet fora clamouring for a K-7 with a K-x sensor.
In September 2010, during Photokina, Pentax announced the K-5, looking outwardly almost identical to the K-7, but boasting Sony’s new 16MP sensor. To everyone’s dismay, the asking price was $1,600; while Ned Bunnell (President of Pentax USA) justifies the price saying the camera is worth it, Pentaxians still remembered the release price for the K-7 being $1,350.
Pentax K-5 Main Specifications
|Sensor||16.08MP (effective pixel count) CMOS APS-C (built by Sony)|
|Sensor size||15.7 x 23.7 mm (1.52x crop)|
|Viewfinder||0.92x magnification, 100% coverage|
|Focusing assist lamp||Yes (green)|
|ISO range||Standard: 100 – 12,800 (expanded: 80 – 51,200)|
|Shooting speed||7 fps|
|Shutter speeds||1/8000s – 30s|
|Exposure modes||Progam, Av, Tv, TAv, Sv, Manual, Bulb and Green|
|Metering||77 segment metering matrix|
|EV compensation||-5.0EV – +5.0EV|
|Rear screen||3 inch 921k LCD|
|RAW mode||Yes (DNG and PEF)|
|Movie mode||1080 (25 fps), 720 (25, 30 fps), 640 x 480 (25, 30 fps), mono sound (stereo sound with external mic)|
|Storage||SD, SDHC or SDXC cards|
|Image stabilisation||Yes, in-body|
|Body construction||Magnesium alloy, weather sealed|
|Dimensions (WxHxD mm)||130.5×96.5×72.5 mm|
|Weight||660 g empty, 741 g with battery and card|
|Release date||September 2010|
What’s in the Box
- Camera with body cap
- Li-ion battery D-LI90 (7.2V, 1860mAh)
- Li-ion battery charger D-BC90
- Shoulder strap
- USB and video cables
- Manual (in English) and warranty card
- Software CD (including Pentax Digital Camera Utility 4, powered by Silkypix)
- Eyepiece cover
Looking at the camera sitting in the box the first thing I thought was it’s tiny! (for a DSLR). I’d held it before, and the K-7 too, but it still surprises me every time I see this body just how compact it is. The next thing that happens (every time!) is that I pick it up and think woah, it’s dense! Yes, the camera is small, but it’s solid. It’s also comfortable to hold despite the weight, although at 741 g with battery and card, it’s still lighter than the Nikon D7000 (779 g) or Canon 7D (915 g).
Yes, it deserves its own section. Not because of the 100,000 clicks lifetime, but because of the sound (the mirror plays an important role here too), or lack thereof. This is the most pleasant, quietest shutter I’ve ever heard on a sub-$2,000 SLR. The thwik! it makes when you take a shot can best be described as a stick of butter being hit with a tennis racquet. Sublime.
While we’re speaking of shutters, I’ll address an issue that popped up with the K-7 which proved rather controversial, that of shutter-induced blur at certain shutter speeds. Falk Lumo performed an in-depth study of this problem for the K-7 showing that there was indeed some blurring in images shot at shutter speeds in the 1/50s – 1/160s range. Note that this blurring was noticeable when all other, more frequent, sources of unsharpness (badly focused lens, inadequate DoF, unstable shooting platform, subject movement, etc.) were neutralised. Falk recently looked into this issue with the K-5 and we’re happy to see shutter-induced blurring is now reduced down to 1/50s shooting speeds, below which both the K-7 and K-5 perform well anyway.
Ergonomics and Buttons
This is always very personal and I often wonder if it’s worth writing about. I like small cameras, and if they’re light, even better, so the K-5 suits my tastes. The bite in the grip really helps with holding the camera, and while I can fit my pinky, it’s a bit of a tight fit. I know for some this might be uncomfortable, but for me it isn’t (and my hands aren’t small).
The e-dials (thumb/forefinger control wheels) are rubberised, making them stick to your fingers when you’re turning them (that’s a good thing, by the way). The mode dial is some 33% taller than on the K-7, so it’s easier to grip, although the detents are a bit hard, which is surprising seeing as it has a safety button in the center that you need to press in order to turn it and change modes. The buttons spread around the body have very little travel and feel a bit stiff. The shutter release has a soft half press, and a hard full press, which is the way it should be if you ask me; there will be no accidental shots with this camera where you think you’re only half pressing the shutter release but actually press it all the way. The lens-release button seems to catch on something meaning I have to give it a serious press to unlock the lens; I suspect this has something to do with the AF screw because when the camera is in manual focus (which retracts the AF screw into the body) the lens-release button becomes easier to work. It seems to be a bit loose and at odds with the rest of the camera’s solid construction; by comparison the lens-release button on my K10D feels extremely solid, even after 3 years of use.
As for the buttons, I should point out that there are a lot of them, with most important functions being accessible directly (ISO, EV comp., WB, flash, etc). There is one customisable button, marked RAW/Fx, that you can set to file format (RAW, JPEG, RAW+), exposure bracketing, digital preview, electronic level or composition adjustment. You can set the AF button so it’s a focus lock, or it can act as the AF engage button, which is how I use it (and I disengage the AF from shutter half press). A neat function is how the 4-way dial works; by default it’s set to access one of four functions (WB, flash, drive mode and custom image), but when you set the AF point mode to SEL, the dials now perform the duty of moving the selected AF point around. If you want to go back to access one of the aforementioned 4 functions, just keep the OK button pressed for 1 second and a beep will announce they’re back to normal (if you simply press the OK button without holding it down the camera will return the AF point to the center).
The downside of having many buttons on such a limited surface is that you can easily confuse them when the camera is up to the eye. Quite a few times I’ve attempted to press the up or left arrow to move the AF point only to turn on liveview. I’m also struggling to quickly find the AF button, and I would much prefer if it were swapped with the AE-L button, which to me is located in a much better place for this function. Of course, anyone who uses the AE-L button frequently will disagree with me; so why can’t Pentax make button mapping a custom user option? I don’t know.
Maybe not part of the ergonomics is the weather sealing, but I’ll comment on it here anyway. The Pentax K-5 is has 77 seals that protect the camera from the environment; it’s not submersible, but it will survive both a thunderstorm and a sandstorm. If you’re a nature or adventure photographer you’ll appreciate the extra level of protection, especially in such a small body (check out our interview with adventure photographer and Pentaxian Kerrick James).
When it came out, shooters loved the K-7 viewfinder (0.92x magnification, 100% coverage), which the K-5 shares. I compared it to the K10D (same as in the K20D, 0.95x magnification, 95% coverage) and it looks the same size, even though mathematically the K-5′s should be larger. I will concede that it’s nice not having to guess about the edges of the view thanks to the 100% coverage—if I see it in the viewfinder, it’s in the photo.
So, as APS-C camera viewfinders go, it’s nice. But here’s a note to all camera makers (except Olympus): Give us 1.3x magnification; I’d much rather have that than 100% coverage.
I generally don’t comment on menus because each brand has their own way of doing them and I’m convinced there’s no right way, there’s just the way you like and they way you don’t. That said, I wanted to mention how easy it is to navigate the menus in this camera. There is no vertical scrolling; if a menu (Record, Play, Settings and Custom) has more options than fit into one screen (and apart from Play, they all do), the options list is divided into pages, with the page numbers shown at the top in grey and the page number you’re on highlighted in colour. To navigate between pages you rotate the rear e-dial, to navigate between menus you rotate the front e-dial. To access a particular setting you select it with the up/down keys. You can navigate pages and menus with the keypad if you really want, but it’s long-winded and not worth the effort.
Bubble levels will become a thing of the past if more cameras adopt electronic levels like the K-5 has—I just sold the shares I had in bubble-level-manufacturing companies. The K-5 incorporates a level for two axes, letting you correct camera position for both tilt and roll. Tilt is for “up/down”, so you can shoot buildings with non-converging vertical lines, for example. Roll measures “side-to-side”, allowing you to have horizontal horizons, for example. And there’s an extra option related to the roll level if you turn it on: The camera will sense when you’re a little bit off and rotate the sensor for you to make it level; not surprisingly Pentax calls this feature ‘Automatic Horizon Correction’.
The level appears as a neat graphic on the LCD showing horizontal and vertical bars divided into yellow segments indicating how far off level you are. They turn green when the camera is level. Bonus: The level works in both landscape and portrait orientation (the equivalent of having a 3-bubble level on your camera) with the added benefit that it doesn’t get in the way of the flash nor take up space in your camera bag. In the viewfinder you have a graphical representation of the roll graph using the EV comp. scale; I just wish there were also an equivalent tilt graph.
I do have one issue with the level when used on the LCD, and that’s the fact that the screen goes black when you half press the shutter release. When I’ve been shooting buildings hand-held without a tripod (as I do most of the time) this has driven me bonkers, as I will get the camera level, then half press the shutter release and pray that I don’t move the camera between then and fully pressing it to take the shot. The problem here is one of inefficient mind reading on behalf of Pentax: I’m sure the guys who designed the level thought it would be used with the camera on a tripod, which is clamped fixed once it is level. Furthermore, all menu and info displays turn off when you half press the shutter release, which must be a general camera setting, and it also affects the electronic level display; they come on immediately after the photograph is taken. Pentax, if you’re reading this, please, solve this issue in a firmware upgrade. Thank you.
Live view is not a gimmick, and those who shoot a lot of macro, astrophotography and landscape on a tripod will attest to this. The live view on the K-5 is at the level of the MILCs I’ve tried and tested, which is to say, it’s good. Apart from getting a real what-you-see-is-what-you-get framed image (or “nearly 100%” according to the manual), live view also allows you to magnify any part of the scene, superimpose a number of different grids (3), a histogram and blinkies for the under/overexposed areas. Shooting info overlay is also optional. This is the theory, anyway. Let me list a few quirks and issues with live view in the K-5 (in no particular order):
- Live view turns off if you change shooting mode.
- Live view not available while camera writes image to card (how long this is will depend on how fast your card is; with my Class 6 it took seconds).
- Live view turns off automatically after 5 minutes to keep the sensor from overheating.
- Images may be noisier when using live view due to higher sensor temperature.
- My biggest issue: The K-5 ramps up/down the live view image, meaning that it will make the image look like what the camera thinks a correctly exposed image of your scene should look like. This is fine if shooting in an autoexposure mode (Av, P, etc.), but not if you want to override settings via EV comp. or shoot in Manual. Furthermore, the histogram that’s displayed is that of the image you see on the LCD, not necessarily that of the actual image the camera will take. This means you can’t use the histogram to judge your exposure when shooting in Manual mode or using EV comp. This could, and should, be fixed via firmware.
These points notwithstanding, live view is still eminently usable and useful and many (all?) of the issues above are displayed by most (all?) digital cameras that have this capability. If you want critical manual focus and your camera is on a tripod, there’s nothing better than live view. As for autofocus, I deal with that in the next section.
Focusing through the viewfinder:
- Manual focus: I found manual focusing quite easy with this viewfinder, although it’s still not as nice as using an old-school screen designed for manual focusing and not brightness. Bear in mind I am quite adept at manually focusing lenses on an APS-C camera, so your mileage may vary. Something I did find improved was the accuracy of the focus confirm, which now seems to have less leeway; moreover, approaching correct focus from either infinity or closest focus yields effectively equal results.
- Autofocus: Much has been written about how the updated SAFOX IX AF engine is much better than previous versions; I’m afraid I haven’t noticed this with my lenses. Bear in mind that I’ve used only screwdrive lenses, and comparing speeds between the K-5 and my K10D I find no appreciable difference either in low or good light. Accuracy is still as good as it ever was, but that has never been an issue with recent Pentax cameras. I’m guessing the screwdrive motors in both cameras have the same power, although I don’t think they’re the same motor because they sound quite different. I’ve been told the shape and mass of the body can change how the motor sounds, so it might be the same model; I just don’t know.
With screwdrive AF, speed will be limited by how fast the motor can move the focusing elements rather than the camera’s AF software engine; might SDM focusing be faster? I borrowed a DA* 16-50mm f/2.8 from Mark Roberts, who uses it on a K20D, and put it to the test, comparing its AF speed with my K10D. In bright light I found the K-5 focused faster than the K10D, and Mark reported the same when compared to his K20D. The difference wasn’t abysmal, but it was there. Where most of the gain (maybe all?) seemed to come from was in the last part of the focusing, where my K10D would stutter a bit until it locked focus, while the K-5 was a lot surer in locking focus, sometimes not even stuttering (and yes, the images were correctly focused). However, in low light (think 50mm, f/2.8, 1/25s at ISO 2,000), I couldn’t notice any difference between the two cameras. I never tested the K-5 with the AF assist beam, which might have improved its focusing speed in low light.
An improvement I did notice, even with screwmount lenses was in AF-C; it’s not earth-shattering, but I noticed it. At the Millennium Park ice rink in downtown Chicago I spent a while photographing the skaters; the camera had no problem keeping up with even the fastest ones. The chap in the photo above was speeding away in style (who wears a tie to skate?) and all the photos I took of him were in focus. This is good enough for me! Please note that shooting a Formula 1 car driving towards you might not work as well.
- Catch-in Focus: For those not familiar with the term, it refers to a method of shooting in which the camera is set in such a way to only release the shutter when it detects something in focus. This is useful for macro, wildlife and sports photographers, and street shooters too. The K-5 allows this functionality if you enable it in Custom function #25 (default is ‘off’), but there’s a catch (if you pardon the pun): It only works with manual focus lenses, or autofocus lenses with an AF/MF switch. A shame, because catch-in focus can be very useful in certain situations.
Some notes of warning: I don’t have any Pentax lenses with AF/MF switches, only Sigmas, and catch-in focus didn’t work with them. When using manual focus lenses, if you have AF mapped to the AF button instead of the shutter release button, you must hold down the AF button while pressing the shutter release in order for catch-in focus to work.
Focusing with live view:
- Manual focus: One of the strong selling points of live view is its usefulness for manual focusing, especially when camera makers allow you to zoom in the display. Luckily, Pentax have included 5 levels of zoom (2x, 4x, 6x, 8x and 10x) accessed sequentially via the info button. As you zoom in a small graphic in the lower left corner tells what size relative to the full frame the image being displayed is. You can also use the 4-way controller to move this zoomed area around anywhere on the scene, which is useful if you’re subject is off-center, as it often is. At 10x magnification the zoomed rectangle takes forever to move, so I recommend zooming in to 4x, moving the rectangle to your subject, then zooming in further.
- Autofocus: One thing I found curious is that there are 3 AF options available in live view: Contrast detect, smile detect and phase detect. If selecting the last option, the mirror will flip down during focusing, thus blacking out the LCD view; I’m not sure why anybody would prefer this method. One immediate drawback I found with live view AF is the lack of auto focus point select, which every MILC I’ve ever used provides. In practical terms, this means you have a fixed AF region only, which granted, you can manually move around almost to the edges of the screen. You cannot, however, change its size (boo!). As for speed, it’s slower than non-live-view phase detect, and slower than other contrast detect cameras I’ve tested (such as the Panasonic G2 or Samsung NX10). It’s usable, thought not for anything that’s moving fast.
When ‘smile detect’ is selected as the AF method, it works like ‘contrast detect’ until a face enters the scene, at which time the camera will detect it automatically and surround it with a box; it can deal with up to 16 faces. Each will have a box around it, with the main face (chosen for esoteric reasons) with extra marks in its box corners. When focus is achieved the box turns green.
When in ‘contrast detect’ or ‘smile detect’ you can select AF-C to enable AF tracking of either a subject in the center of the frame (when in ‘contrast detect’) or a face (when in ‘face detect’).
- Catch-in Focus: Not available in live view!
RAW File Quality
One of the winning qualities of this sensor was meant to be its dynamic range (14 stops according to DxO Mark); it’s not easy to test DR without some specialised equipment, but I went out to see what I could achieve as a practical example, if not necessarily precise. To this end I shot the image of the building you see above; not a prize winner by any means, but it combines a dark foreground subject in shadow with a bright textured sky. I say “textured” because although it might not seem like it, those clouds did have texture!
I set out to check something that’s important to me, which is overexposure headroom. Basically this tells me how much I can overexpose an image (or portion thereof) and still recover detail in postprocessing. After loading the image into ACR I turned on the ‘highlight clipping warning’, which colours all blown out pixels red. I then began lowering the exposure and the red clipped area started shrinking; the point when the blown out area stops shrinking tells you when you’ve reached the recovery limit. In the case of the K-5 it was at -1 EV, meaning it has a 1 stop overexposure headroom. You can see in the images below what you can recover from those areas that were not completely blown out.
Now that I knew how much room I had in the highlights, I wanted to know how the K-5 performed in shadow areas. There will be no technical talk here as I think the images speak for themselves, so just let me ask you to look at the two photos below and tell me the main differences you see (apart from the different DoF, for which I apologise); don’t be shy, click to download the full 16MP resolution files and pixel peep all you like.
If you’re still not sure of the differences, I’ll give you a clue: Before I massaged the second one with ACR it looked like this out of the camera. I’m serious! I was able to recover a full 6 stops of underexposure. I’m not claiming both images look the same, but this extreme case underlines exactly how little noise the K-5 exhibits and how well it handles underexposure; perhaps better than it handles overexposure. Needless to say, I’m impressed.
A few days ago I was lucky enough to be on the beach with the K-5. I took this shot thinking, again, about testing the dynamic range. Not only is it prettier than the building above, but I think it exemplifies the K-5′s capability to reproduce high DR scenes. In this case we have bright sand and fluffy clouds, contrasting with the shadow in the foreground and the dark vegetation. The clouds aren’t blown out and there is plenty of detail in both the shadow and the vegetation. I shot this image in RAW and converted to JPEG using my standard ACR settings, meaning I didn’t manipulate exposure nor did I add fill light or highlight recovery; blacks were set to 5. Oh, and contrast was set to ‘strong contrast’ in point curve.
The K-5 Sensor Stains
Word from Pentax USA and Pentax Japan is that the origin of the stains has been found and steps taken to avoid them appearing in new sensors. All sensors being shipped out in cameras now should be stain free. Reports from recent buyers support this. Nonetheless, I have decided to keep this section here for informational purposes.
A WORD FROM PENTAX IMAGING USA REGARDING THE K-5 SENSOR STAINS
I’ve been in contact with Pentax Imaging President, Ned Bunnell, and here is what he had to say concerning this matter:
I can confirm that Pentax Japan is fully aware of this situation and is diligently investigating the cause of these stains. I’d like to remind our US customers that their K-5 is covered under the 1-year warranty and that we’ll of course honor that. However, until we find out which batch(es) are affected by the stains, we recommend that photographers keep their cameras and refrain from exchanging them for another unit if purchased from an online retailer. Once we fully understand the origin of the stains and how to deal with them we will encourage affected K-5 owners to send in their cameras for warranty repair.
January 26th 2011: Read the Official Statement from Pentax USA regarding the stains.
Every product ever released in the whole history of manufacturing has had some sort of issue with it—I can guarantee you the first wheel ever made was not quite round. Some problems are bigger than others, despite which they all seem to get blown out of proportion in the same way. The problem with the K-5 is that some units have stains on their sensors (or maybe the stains are underneath the IR/UV filter, but the result is the same). Is this an epidemic? Not at all, but those who find themselves with a stained sensor make the most noise and you might be lead to believe every camera is faulty, but this isn’t the case. We still don’t know how rampant this issue is, but if you own a K-5 you should look to see if yours is affected.
One thing that caught my attention on forums posting about the stains was the large number of people lamenting that their K-5 sensor was stained when all they had was some dust. I know this because they were posting examples of these “stains”. I wonder how many of those claiming stains and not posting examples actually had dust bunnies; this is likely skewing the numbers of cameras claimed to be affected. In order to help curb this trend, I am including instructions on how to properly test and identify marks that you may find on your sensor.
Types of marks on sensors:
- Smudges: This is a very light, generally circular mark that is only visible against uniform light backgrounds such as a clear sky. They often only show up at small apertures and most often than not will go unnoticed. To clean them you will have to use a wet-cleaning kit.
- Dust (and pollen): Depending on the size of the mote its effect can range from minimal to a tiny black hole in your photograph. Larger motes will show up clearly even at large apertures. Sometimes the camera’s dust-cleaning feature will get rid of them, but more often than not you’ll need a blower. I use and recommend the Giotto 7.5″ Rocket Blower, which is incredibly cheap given how useful it is.
- Filaments: They resemble short hairs and are more annoying to clone out than smudges or dust specks. They tend to respond better to the camera’s dust-cleaning feature than dust.
- K-5 stains: They tend to come in groups and are perfectly circular in shape, with a halo around a dark center. Interestingly, many users report finding them close to the center of sensor. They are not affected by in-camera dust-cleaning or air blowing. I don’t think anyone has been able to clean them with a wet-cleaning method, but if they have, I’d like to hear about it.
Dust and filaments that are not cleaned promptly can become permanently attached to the sensor and will be impervious to blowing; pollen attaches itself faster than dust. When this happens the only way to remove them is via wet-cleaning.
How to find marks on sensors:
- Mount your longest lens on the camera. You can run the test using an 18mm lens, but longer is more convenient; I use either 300mm or 400mm.
- Set the camera to manual focus and focus the lens to infinity.
- Select matrix metering on the camera, set the shooting mode to Av and choose the smallest aperture (highest f number) available, set the ISO to its lowest value, set EV comp. to +1EV.
- Point the camera at a piece of blank sky (either blue or grey, just choose a uniform spot). If you don’t have access to a daylight sky, then point at an evenly lit, light-coloured wall, standing as close to it as possible without casting a shadow on your image.
- Take the shot! Don’t worry if the shutter speed chosen by the camera seems too slow for the focal length you’re using; you’re not taking a sharp picture of whatever is in front of your lens, but rather an image of what’s on the sensor, which will not be affected by motion blur.
- Load the image into your favourite viewer and zoom in to 100%; increase contrast if you want. Scan the whole image looking for marks.
Below I’ve blown up the dust spec and the K-5 stain from the above image to 100%. The dust spec ignored the dust-cleaning feature on the K-5 (I ran it 3 times), but disappeared after some blowing from my Giotto Rocket. The stain was unaffected by any of these cleaning methods. I made this image by shooting a blue patch of sky at 400mm and f/32; I didn’t increase contrast.
If you are unlucky enough to find a stain, I propose that you go back and redo the testing, only this time take pictures at various apertures (f/32, f/22, f/16, f/11, f/8…) and see how the appearance of the stains changes as aperture increases in size; I expect them to become barely visible by f/8. Now go out and take some real pictures, at the apertures you normally use, and then go back to check for the stains—even knowing where they are, can you see them in your real photos? I’m not saying they’re not an issue, I’m just saying that stains are not the end of the world and you can still use the camera for a few weeks until Pentax figures out how to fix the problem.
For more than you ever wanted to know about stains on sensors, read Falk Lumo’s Advanced Stainology post.
All photos: ©Miserere (unless otherwise noted).
|Please consider purchasing the Pentax K-5 from B&H. We’ll get a small commission (at no extra cost to you) and you’ll get our eternal gratitude. It’s one of the ways we keep this site running!|
- Review – Pentax K-5, Part 2
- Review – Pentax K-x Part 1: First Look
- Review – Pentax K-x Part 2: Features, Menus, Ergonomics
- Review – Pentax K-x Part 3: Conclusions
- Review – Olympus Pen E-PL1, Part 1: First Impressions