[bwDraco]

I just came home with a K-3 II from B&H along with a battery grip. Compared to my old K-5, everything just seems more refined: the shutter, the AF, the sensor, the display, just everything. To my surprise, it actually included a Flucard for free; I haven't had a chance to try out the wireless tethering yet. I also haven't had a chance to test the PIxel Shift Resolution functionality. I expect to report back with some results soon.

The AF is much, much better than what I got on my K-5, and it's actually able to track well, with at least some degree of 3D distance prediction capability in AF.C. On a 32GB SanDisk Extreme PRO UHS-I card (purchased with the camera), the buffer clears *way* faster than what I'm used to from my K-5. UHS-I is a huge advantage...

Sadly, my DA 18-135mm lens is starting to give me infinity focus issues (again) and I will have to send it to Precision Camera for an out-of-warranty repair soon. It's still usable as it's not that bad right now, but it'll get worse with time. I'll see what I can do about it.

I hope to upload some shots soon, but I don't quite have the time to put the pictures through post right now. Sorry

—DragonLord

---

Last edited by bwDraco; 06-24-2015 at 11:23 AM.

[bwDraco]

Below are the best-case continuous shooting speeds I measured from the camera using a fully charged battery and 32GB SanDisk Extreme PRO UHS-I card. These values are based on an audio recording of the camera shooting continuously:

• Continuous H: 8.37 fps
• Continuous M: 4.58 fps
• Continuous L: 3.07 fps

I have no plans to test buffer depth, but my experience is that the specs tend to be a bit conservative when using professional high-speed memory cards. These speeds are marginally higher than manufacturer claims, but as these are best-case values, they are not intended to be representative of real-world shooting conditions.

—DragonLord

[normhead]

Congrats enjoy your purchase, that reminds me, my Sigma 18-250 quit focussing, and is in the que at Sigma. I wonder what that's going to set me back.

[dancelav]

I just came home with a K-3 II from B&H along with a battery grip. Compared to my old K-5, everything just seems more refined: the shutter, the AF, the sensor, the display, just everything. To my surprise, it actually included a Flucard for free; I haven't had a chance to try out the wireless tethering yet. I also haven't had a chance to test the PIxel Shift Resolution functionality. I expect to report back with some results soon.

The AF is much, much better than what I got on my K-5, and it's actually able to track well, with at least some degree of 3D distance prediction capability in AF.C. On a 32GB SanDisk Extreme PRO UHS-I card (purchased with the camera), the buffer clears *way* faster than what I'm used to from my K-5. UHS-I is a huge advantage...

Sadly, my DA 18-135mm lens is starting to give me infinity focus issues (again) and I will have to send it to Precision Camera for an out-of-warranty repair soon. It's still usable as it's not that bad right now, but it'll get worse with time. I'll see what I can do about it.

I hope to upload some shots soon, but I don't quite have the time to put the pictures through post right now. Sorry

—DragonLord

Try to calibrate the lens to your camera at AF Fine Adjustment!

[bwDraco]

Try to calibrate the lens to your camera at AF Fine Adjustment!

It's a mechanical lens problem, not an AF problem. Calibration would not help here.

—DragonLord

---------- Post added 06-16-15 at 01:26 PM ----------

Okay, I just did some buffer depth testing, but only to determine the absolute minimum number of shots that can be fired at full speed irrespective of memory card speed, ISO, or other camera settings. Using a slow Kingston class 4 card, RAW+JPEG, ISO 51200, and full corrections, the worst-case buffer depth is 23 shots. This is an improvement over the K-5, which has a worst-case buffer depth of 20 shots, despite the fact that the K-3 II generates significantly larger images. (Note that the physical size of the buffer was doubled in the K-3 and retained in the K-3 II.)

I already have 380 shots on the camera. Granted, this is much less of an issue with the doubled shutter life, but it's a heck of a lot of use within the first 24 hours...

—DragonLord

[bwDraco]

Just tried my DA 50/1.8. It front-focused (as it did on my K-5) and needed a -8 adjustment to give accurate focus but the autofocus is now consistent, unlike my K-5 whose AF was all over the place on this lens making it almost unusable wide-open. Those f/2.8 high-precision points are making a huge difference here...

Edit: Further refinement brought it closer to -5, and results are still inconsistent. Light and contrast levels seem to be playing a major role here. The K-3 II is still better than the K-5 with this lens with visibly less variance in AF behavior.

—DragonLord

[bwDraco]

More test notes:

• The AF really works at very low light. Ricoh wasn't joking when they said the AF works down to EV -3. Using my DA 50/1.8, I was able to get reasonably accurate focusing down to that very level for as long as there is a high-contrast subject to focus on.
• The SR has definitely improved. In the same low-light conditions, the K-5 would have practically no chance of giving even a somewhat sharp shot at about 1/4s. With good technique, the K-3 II would give shots that are only slightly soft most of the time. (SR on the K-5 was originally specified to be effective to 4 stops but is closer to 3 stops in reality as that number was not CIPA spec. SR on the K-3 II is rated to 4.5 stops, a value which is CIPA spec.)
• Live View has changed considerably. The control layout changes keep tripping me up. Pressing OK now magnifies the image, not INFO which will always pull up the control panel. The contrast detection AF is much smarter, racking back and forth fast for a first pass and doing a slower second pass only if required to confirm accuracy.
  • Edit: Live View now has significantly lower lag and runs at 60 fps provided there is enough light, unlike the K-5 which ran Live View at 30 fps. The maximum sensor gain allowed during Live View has increased as well, allowing higher frame rates even in poor lighting conditions. Live View also is no longer subject to interruption when a lens is removed or attached, and resumes automatically after exiting from the menu or control panel or from playback mode.
• Pixel Shift Resolution brings a huge boost in resolution when used correctly. See the image below to get a sense of what you can expect from this technology. For full-size images, see these Flickr links: PSR off, PSR on

Pentax K-3 II with D FA 50mm macro lens. Av, 5s f/8 ISO 100. Image on left is with Pixel Shift Resolution OFF; image on right is with Pixel Shift Resolution ON.


—DragonLord

---

Last edited by bwDraco; 06-17-2015 at 04:13 PM.

[bwDraco]

Flash metering also seems to be much more consistent on the K-3 II. Using my Metz mecablitz 58 AF-2 digital flashgun, I'm able to get consistently correct exposures even with bounce flash. The Auto ISO behavior with flash is different; it only shows "Auto" which may indicate that the camera cannot determine the correct sensitivity until the monitor pre-flash is fired. Setting the ISO manually to a low value, as I normally do with flash, does not cause any exposure issues unless the required flash output exceeds what is available, unlike my K-5 which would consistently underexpose (fire flash with lower than correct output) with low ISO settings unless I applied exposure compensation.

What a difference the 86K RGB metering sensor makes... 👍😃

—DragonLord

[tomtor]

Flash metering also seems to be much more consistent on the K-3 II. Using my Metz mecablitz 58 AF-2 digital flashgun, I'm able to get consistently correct exposures even with bounce flash. The Auto ISO behavior with flash is different; it only shows "Auto" which may indicate that the camera cannot determine the correct sensitivity until the monitor pre-flash is fired. Setting the ISO manually to a low value, as I normally do with flash, does not cause any exposure issues unless the required flash output exceeds what is available, unlike my K-5 which would consistently underexpose (fire flash with lower than correct output) with low ISO settings unless I applied exposure compensation.

What a difference the 86K RGB metering sensor makes... 👍😃

—DragonLord

I did also various flash tests today with my Pentax flash. It worked flawless. Next time I will try my Metz.

Very pleased with the k3II. Also shake reduction is working MUCH better than on my k-x.

[howieb101]

Thanks for the feedback Dragonlord.

Nice example of pixel shift.

I have the K3 and flash exposure was largely fixed on that model.

Improved shake reduction is welcome as the 24Mpx sensor is unforgiving to camera shake.

Thanks again,
Howie B

[bwDraco]

Mirror blackout time is down as well. Based on audio recordings. from the time the mirror starts moving up to the time the mirror is reset, blackout time is down from 110ms from the K-5 to 102ms on the K-3 II. (Measurement was difficult due to significantly different shutter recock sounds. These figures should be accurate to about +/- 2ms.)

Haven't had to pull out my Sensor Swabs and Eclipse cleaning fluid for a while, but just cleaned the sensor on the K-3 II, after about 2250 shots. The stubborn spots had likely accumulated due to the frequent lens changes. It took four swabs this time. (Direct sensor cleaning is hard and tedious because of the tendency of the swabs to leave particles and residue behind, necessitating multiple attempts.)

—DragonLord

[bwDraco]

The AA filter simulator works as intended, with Type1 giving a weak AA effect and Type2 noticeably stronger. Maximum continuous shooting speed is reduced to 6.18 fps in both Type1 and Type2 modes, as determined from audio recordings. 500 Hz pure sine wave noise is audible, especially in Type2 mode where the sensor vibration is strong enough to be felt in the camera body (and likely capable of removing dust from the sensor). The AA filter simulator does not operate at shutter speeds below 1/20s. I don't run into moire patterning often enough to need to turn on the AA filter simulator on a regular basis—the high pixel density of the sensor means that moire isn't going to appear all that often in the first place.

The shutter count is now 2512.

—DragonLord

[RobA_Oz]

The shutter count is now 2512.

Wow! 200 shots per day average is what I might do when I'm travelling abroad. You'll be rivalling Ron Hendricks and his 100k+ bodies before you know it!

[bwDraco]

Most of these shots are experimental and get thrown out en masse. Things should start to slow down over the next few days.

I'm still not used to the new Canon-style AF mode selection controls (hold AF Mode button, front dial to change AF mode, rear dial to change AF point selection). Nonetheless, the new AF system is remarkably sure-footed, with far less hesitation especially in suboptimal lighting conditions. The higher AF point density gives me a good deal more control over where to focus, with good accuracy across all points. I haven't had much of a chance to shoot sports with the K-3 II (it's summer term so the CSI Dolphins aren't going to be playing much for a while), but now that we have a rudimentary area AF system with a 5x5 center grid, I expect better results with more keepers.

I'm also trying out back-button AF as it allows better control over focusing.

—DragonLord

[bwDraco]

Here's an example of how the camera handles a backlit subject with shadow and highlight correction on auto:

Pentax K-3 II with DA 50mm f/1.8 lens. HyP-Tv, 1/500s f/5.6 ISO 200.


The camera actually has the intelligence to turn on dynamic range corrections when it is appropriate to do so. In doing so, it had unexpectedly bumped the ISO to 200, the minimum required for highlight correction. Another advantage of the Real-Time Scene Analysis System and its underpinning 86K RGB metering sensor...

—DragonLord