Japanese blogger tests K-3 II AF.C performance against K-3: Measurable improvement

bwDraco · 08-14-2015, 03:00 PM

https://translate.google.com/translate?sl=jp&tl=en&js=y&prev=_t&hl=en&ie=UTF...-text=&act=url (via RiceHigh)

To summarize, the K-3 II exhibits a measurable improvement in the keeper rate over the original K-3 when shooting subjects moving directly towards the camera. Most critically, the K-3 II is much faster to recover after losing focus than the K-3. In the Sobu Rapid test, the K-3 lost focus for 11 consecutive frames in the middle of the sequence, while the K-3 II missed four consecutive frames towards the end.

AF.C 9-point expanded-area, focus-priority first frame, FPS-priority thereafter. Hold AF Status is off. Both cameras tested with DA* 300mm lens.

Questions, thoughts, or comments?

—DragonLord

Last edited by bwDraco; 08-14-2015 at 03:08 PM.

stevebrot · 08-14-2015, 03:37 PM

Originally posted by DragonLord

Questions, thoughts, or comments?

What is the Sobu Rapid test?

Steve

cali92rs · 08-14-2015, 03:43 PM - **1 Like**

Originally posted by DragonLord

Questions, thoughts, or comments?

—DragonLord

My thought is...cool, I am glad I shelled out the extra money for something

I live in an area that is pretty far from good night sky so I don't get to use the astrotracer

. I played around with pixel shift resolution and while it is cool, both your tripod and subject have to be dead nuts stable otherwise you get weird results.
The increase in AF.C performance is definitely welcome

bwDraco · 08-14-2015, 03:43 PM

Originally posted by stevebrot

What is the Sobu Rapid test?

Steve

The subjects used for the test are trains from two lines in Japan. Sobu Rapid is one of them.

—DragonLord

**jpzk** · 08-14-2015, 04:27 PM

I wonder why this is?
Why would the K3II be "so much better" at AF-C than the K3 ?

It was quite hard to understand Google translate.

**Spodeworld** · 08-14-2015, 04:34 PM

If true, and if AF hardware is the same, makes me wonder why Pentax doesn't update the K3's AF.

aleonx3 · 08-14-2015, 04:37 PM

Originally posted by DragonLord

The subjects used for the test are trains from two lines in Japan. Sobu Rapid is one of them.

—DragonLord

Siebu (I think it is a translation typo) is one of the railway lines connecting suburbs to the Tokyo subway and JR rail system.

Originally posted by jpzk

I wonder why this is?
Why would the K3II be "so much better" at AF-C than the K3 ?

It was quite hard to understand Google translate.

It is possible that the improved AF-C tracking in k-3II requires some adjustment that a simple firmware upgrade can not achieve in k-3.

Adam · 08-14-2015, 05:39 PM

Interesting. I bet they are just unwilling (to spend resources) to add the fix to the K-3 so they can say the K-3 II is better. But at least there's a measurable improvement

**Mikesul** · 08-14-2015, 07:07 PM

I saw this test and found it interesting. I also have had a K-3ii intermittently (returns for SN) since late June. I have not done controlled tests but I have had the definite impression that it holds focus better than my K-3. I have several times acquired and kept a swallow moving erratically along a water course with, for me, an amazing number of keepers. The swallow was very small in the frame so they are not great photos but they were clearly in focus through most of the shots. Nothing objective but it does appear more decisive. GPS is a big deal for what I do most of the time so I am a bit predisposed to like the K-3ii.

derekkite · 08-15-2015, 04:58 PM

I get the impression with the K3 that it is resource limited; it cannot complete in a timely way the calculations required when a subject is moving faster than it can handle, essentially calculating a focus point where it was rather than where it is. There was some hardware changes in the AF if I remember correctly, so maybe they changed a component that caused a bottleneck somehow.

The tests of the K3 correspond to my experience. It loses it's mind, then is slow to recapture, and it is worse close in.

I'd like to try one.

bwDraco · 08-15-2015, 06:48 PM

It seems the AF algorithm improvements were aimed at increasing the maximum tracking speed so that the camera can maintain locks on fast-moving subjects better. If anything, this makes the K-3 II the best Pentax DSLR yet for sports photography. It's similar to the differences shown in this diagram (source):

—DragonLord

derekkite · 08-15-2015, 07:14 PM

I was out late this afternoon, diffuse light, and there were gulls flying by. My goal were osprey, but I have to wait for them. 500mm Sigma f4.5. When I did AFC continuous shooting it would get confused, shoot four or more out of focus shots, then find itself. When I followed a gull without shooting, it kept very good focus even against mountain background. It adjusted constantly and accurately.

The interval between mirror up and mirror down gives no information to the PDAF sensors nor the metering sensor. It has the short time between shots to sample and detect movement. The sample rate for the calculations needs to be quite high to get usable data. Without knowing, that is where the change between the K3 and K3II has occurred, and it isn't simply a matter of firmware. 1.20 did change AF for the better, probably as good as the K3 hardware can be.

I may try a slower continuous shutter rate to see if I get more usable shots.

bwDraco · 08-16-2015, 05:18 AM

Part of the secret to the incredible AF tracking performance of top-of-the-line C&N cameras is their extremely short mirror blackout time. Unlike most cameras, the Nikon D4 and Canon EOS-1D X can provide a nearly uninterrupted viewfinder image. The EOS-1D X, in particular, has a blackout time of just 60 ms, while the D4 blacks out the finder for 74 ms. (My own testing found that the K-3 II black outs for about 102 ms in single-shot, though faster for subsequent shots in a continuous sequence.)

—DragonLord

audiobomber · 04-30-2016, 03:28 PM - **1 Like**

Originally posted by bwDraco

AF.C 9-point expanded-area, focus-priority first frame, FPS-priority thereafter. Hold AF Status is off. Both cameras tested with DA* 300mm lens.

Questions, thoughts, or comments?

I have a comment. If you aren't using AF-C, Expanded Area focal point and AF Hold, the camera is not tracking, it is simply refocusing for every shot. The Japanese blogger is missing an important tool for true tracking, i.e. AF Hold "ON". I suggest AF Hold Medium as a starting point. NOTE: The first shot must be perfectly focussed before you release the shutter, otherwise you will have a string of misses.

I only use Focus priority, never FPS, because I have no use for out-of-focus shots.

**jpzk** · 04-30-2016, 03:35 PM

Originally posted by bwDraco

https://translate.google.com/translate?sl=jp&tl=en&js=y&prev=_t&hl=en&ie=UTF...-text=&act=url (via RiceHigh)

To summarize, the K-3 II exhibits a measurable improvement in the keeper rate over the original K-3 when shooting subjects moving directly towards the camera. Most critically, the K-3 II is much faster to recover after losing focus than the K-3. In the Sobu Rapid test, the K-3 lost focus for 11 consecutive frames in the middle of the sequence, while the K-3 II missed four consecutive frames towards the end.

AF.C 9-point expanded-area, focus-priority first frame, FPS-priority thereafter. Hold AF Status is off. Both cameras tested with DA* 300mm lens.

Questions, thoughts, or comments?

—DragonLord

Wonderful .... for those lucky owners of the K3II !
I fell for the K3, thinking that I might need the on board flash but I don't.

Just starting to get acquainted with the K3, especially for birding, so I should have picked the K3II instead ...

08-14-2015, 03:00 PM	#1
bwDraco Veteran Member Join Date: Jul 2010 Location: New York Photos: Albums Posts: 1,071	Japanese blogger tests K-3 II AF.C performance against K-3: Measurable improvement https://translate.google.com/translate?sl=jp&tl=en&js=y&prev=_t&hl=en&ie=UTF...-text=&act=url (via RiceHigh) To summarize, the K-3 II exhibits a measurable improvement in the keeper rate over the original K-3 when shooting subjects moving directly towards the camera. Most critically, the K-3 II is much faster to recover after losing focus than the K-3. In the Sobu Rapid test, the K-3 lost focus for 11 consecutive frames in the middle of the sequence, while the K-3 II missed four consecutive frames towards the end. AF.C 9-point expanded-area, focus-priority first frame, FPS-priority thereafter. Hold AF Status is off. Both cameras tested with DA* 300mm lens. Questions, thoughts, or comments? —DragonLord Last edited by bwDraco; 08-14-2015 at 03:08 PM.

08-14-2015, 04:27 PM	#5
jpzk Site Supporter Join Date: Jan 2009 Location: Québec Photos: Gallery \| Albums Posts: 8,251	I wonder why this is? Why would the K3II be "so much better" at AF-C than the K3 ? It was quite hard to understand Google translate. Last edited by jpzk; 08-14-2015 at 04:32 PM.

08-14-2015, 05:39 PM	#8
Adam Administrator Site Webmaster Join Date: Sep 2006 Location: Arizona Photos: Gallery \| Albums Posts: 51,597	Interesting. I bet they are just unwilling (to spend resources) to add the fix to the K-3 so they can say the K-3 II is better. But at least there's a measurable improvement Adam PentaxForums.com Webmaster (Site Usage Guide \| Site Help \| My Photography) PentaxForums.com server and development costs are user-supported. You can help cover these costs by donating or purchasing one of our Pentax eBooks. Or, buy your photo gear from our affiliates, Adorama, B&H Photo, KEH, or Topaz Labs, and get FREE Marketplace access - click here to see how! Trusted Pentax retailers:

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
AF.C on K-5 II vs K-3	StigVidar	Pentax DSLR Discussion	2	06-22-2015 09:57 AM
lots of different creative work- K-5, K-5 II/s, or K-3	yas_min	Troubleshooting and Beginner Help	15	01-23-2014 08:05 AM
A tale of 3 sensors K-3/K-01/K-7	Painter	Pentax K-3 & K-3 II	21	11-16-2013 09:38 PM
Proposed K-3 AF test against the real comparable alternatives	RockvilleBob	Pentax K-3 & K-3 II	9	11-13-2013 11:33 AM

08-14-2015, 03:37 PM	#2
stevebrot Otis Memorial Pentaxian Join Date: Mar 2007 Location: Vancouver (USA) Photos: Gallery \| Albums Posts: 42,007	Originally posted by DragonLord Questions, thoughts, or comments? What is the Sobu Rapid test? Steve

08-14-2015, 03:43 PM - 1 Like	#3
cali92rs Veteran Member Join Date: May 2011 Location: Long Beach, CA Posts: 3,354	Originally posted by DragonLord Questions, thoughts, or comments? —DragonLord My thought is...cool, I am glad I shelled out the extra money for something I live in an area that is pretty far from good night sky so I don't get to use the astrotracer . I played around with pixel shift resolution and while it is cool, both your tripod and subject have to be dead nuts stable otherwise you get weird results. The increase in AF.C performance is definitely welcome

08-14-2015, 03:43 PM	#4
bwDraco Veteran Member Join Date: Jul 2010 Location: New York Photos: Albums Posts: 1,071 Original Poster	Originally posted by stevebrot What is the Sobu Rapid test? Steve The subjects used for the test are trains from two lines in Japan. Sobu Rapid is one of them. —DragonLord

08-14-2015, 04:34 PM	#6
Spodeworld Site Supporter Join Date: Aug 2011 Location: New Joisey Photos: Gallery \| Albums Posts: 1,370	If true, and if AF hardware is the same, makes me wonder why Pentax doesn't update the K3's AF.

08-14-2015, 04:37 PM	#7
aleonx3 Veteran Member Join Date: Jul 2007 Location: Brampton, Ontario Photos: Gallery \| Albums Posts: 3,996	Originally posted by DragonLord The subjects used for the test are trains from two lines in Japan. Sobu Rapid is one of them. —DragonLord Siebu (I think it is a translation typo) is one of the railway lines connecting suburbs to the Tokyo subway and JR rail system. Originally posted by jpzk I wonder why this is? Why would the K3II be "so much better" at AF-C than the K3 ? It was quite hard to understand Google translate. It is possible that the improved AF-C tracking in k-3II requires some adjustment that a simple firmware upgrade can not achieve in k-3.

08-14-2015, 07:07 PM	#9
Mikesul Loyal Site Supporter Join Date: Sep 2009 Location: Phoenix, AZ Posts: 7,594	I saw this test and found it interesting. I also have had a K-3ii intermittently (returns for SN) since late June. I have not done controlled tests but I have had the definite impression that it holds focus better than my K-3. I have several times acquired and kept a swallow moving erratically along a water course with, for me, an amazing number of keepers. The swallow was very small in the frame so they are not great photos but they were clearly in focus through most of the shots. Nothing objective but it does appear more decisive. GPS is a big deal for what I do most of the time so I am a bit predisposed to like the K-3ii.

08-15-2015, 04:58 PM	#10
derekkite Pentaxian Join Date: Sep 2011 Location: Nelson B.C. Posts: 3,782	I get the impression with the K3 that it is resource limited; it cannot complete in a timely way the calculations required when a subject is moving faster than it can handle, essentially calculating a focus point where it was rather than where it is. There was some hardware changes in the AF if I remember correctly, so maybe they changed a component that caused a bottleneck somehow. The tests of the K3 correspond to my experience. It loses it's mind, then is slow to recapture, and it is worse close in. I'd like to try one.

08-15-2015, 06:48 PM	#11
bwDraco Veteran Member Join Date: Jul 2010 Location: New York Photos: Albums Posts: 1,071 Original Poster	It seems the AF algorithm improvements were aimed at increasing the maximum tracking speed so that the camera can maintain locks on fast-moving subjects better. If anything, this makes the K-3 II the best Pentax DSLR yet for sports photography. It's similar to the differences shown in this diagram (source): —DragonLord

08-15-2015, 07:14 PM	#12
derekkite Pentaxian Join Date: Sep 2011 Location: Nelson B.C. Posts: 3,782	I was out late this afternoon, diffuse light, and there were gulls flying by. My goal were osprey, but I have to wait for them. 500mm Sigma f4.5. When I did AFC continuous shooting it would get confused, shoot four or more out of focus shots, then find itself. When I followed a gull without shooting, it kept very good focus even against mountain background. It adjusted constantly and accurately. The interval between mirror up and mirror down gives no information to the PDAF sensors nor the metering sensor. It has the short time between shots to sample and detect movement. The sample rate for the calculations needs to be quite high to get usable data. Without knowing, that is where the change between the K3 and K3II has occurred, and it isn't simply a matter of firmware. 1.20 did change AF for the better, probably as good as the K3 hardware can be. I may try a slower continuous shutter rate to see if I get more usable shots.

08-16-2015, 05:18 AM	#13
bwDraco Veteran Member Join Date: Jul 2010 Location: New York Photos: Albums Posts: 1,071 Original Poster	Part of the secret to the incredible AF tracking performance of top-of-the-line C&N cameras is their extremely short mirror blackout time. Unlike most cameras, the Nikon D4 and Canon EOS-1D X can provide a nearly uninterrupted viewfinder image. The EOS-1D X, in particular, has a blackout time of just 60 ms, while the D4 blacks out the finder for 74 ms. (My own testing found that the K-3 II black outs for about 102 ms in single-shot, though faster for subsequent shots in a continuous sequence.) —DragonLord

04-30-2016, 03:28 PM - 1 Like	#14
audiobomber Veteran Member Join Date: May 2008 Location: Sudbury, Ontario Photos: Albums Posts: 6,806	Originally posted by bwDraco AF.C 9-point expanded-area, focus-priority first frame, FPS-priority thereafter. Hold AF Status is off. Both cameras tested with DA* 300mm lens. Questions, thoughts, or comments? I have a comment. If you aren't using AF-C, Expanded Area focal point and AF Hold, the camera is not tracking, it is simply refocusing for every shot. The Japanese blogger is missing an important tool for true tracking, i.e. AF Hold "ON". I suggest AF Hold Medium as a starting point. NOTE: The first shot must be perfectly focussed before you release the shutter, otherwise you will have a string of misses. I only use Focus priority, never FPS, because I have no use for out-of-focus shots.

04-30-2016, 03:35 PM	#15
jpzk Site Supporter Join Date: Jan 2009 Location: Québec Photos: Gallery \| Albums Posts: 8,251	Originally posted by bwDraco https://translate.google.com/translate?sl=jp&tl=en&js=y&prev=_t&hl=en&ie=UTF...-text=&act=url (via RiceHigh) To summarize, the K-3 II exhibits a measurable improvement in the keeper rate over the original K-3 when shooting subjects moving directly towards the camera. Most critically, the K-3 II is much faster to recover after losing focus than the K-3. In the Sobu Rapid test, the K-3 lost focus for 11 consecutive frames in the middle of the sequence, while the K-3 II missed four consecutive frames towards the end. AF.C 9-point expanded-area, focus-priority first frame, FPS-priority thereafter. Hold AF Status is off. Both cameras tested with DA* 300mm lens. Questions, thoughts, or comments? —DragonLord Wonderful .... for those lucky owners of the K3II ! I fell for the K3, thinking that I might need the on board flash but I don't. Just starting to get acquainted with the K3, especially for birding, so I should have picked the K3II instead ...