[dosdan]

I've commenced my K-3 Long Jump AF testing. I like to use LJ as an AF test as it's not easy maintaining focus on someone running towards the camera. I started around 5:30pm. There were some long shadows on some parts of the track. I used the Sigma 70-200/F2.8 II HSM. To stress the AF I only used f/2.8-3.2. With the K-5 I would have been too scared to do this with someone running directly at the camera and would have chosen f/6.3-7.1.


The results were inconsistent. Some sequences were good with every shot in focus, some had 1 or 2 OOF, some had most OOF. I'm not that disappointed. I have to work on the AF settings and my technique.

But what I immediately noticed was that the good shots were better than I'd ever seen from the K-5. I did not expect the difference to be that significant, but it was.

There were a number of changes that could have contributed:

1. This was entering golden hour shooting time.
2. I had to upgrade from SilkyPix Developer Studio Pro 4 to 5 to handle K-3 shots.
3. I was using shallow DOF.
4. I was using a 24MP camera without an AA filter.

If anything, the shots are a bit under-sharpened, and I could probably apply more sharpening. The K-5 would not have responded well to this level of sharpening – it would have cookie-cutter edges and increased harshness. But the K-3 seems to accept more sharpening without problems. The sharpening is different in PSP 4 vs 5, so this complicates matters, but my initial impression was that this new baby really can offer a visible improvement. I'll see how this initial impression holds up when I try the DA* 60-250/F4 with the K-3.


I've not tried any AF micro-adjustment yet either. So it's early days. But I'm pleased so far.


Here's a good shot, both in the original framing and with a 100% crop. I've applied the equivalent of the SP Pro 4's Natural Fine sharpening preset. As you can see in the 100% crop, the edges of the face are not unnaturally separated from the background and it looks reasonably natural. I thought the visual quality was good for a 100% crop with minimal PP.


1/1000s, f/3.2, ISO200, 128mm FL.

[JinDesu]

I think most any top level camera would struggle to have 70% of a sequence in focus, much less 100%, with someone running and jumping at the camera. The k-3 is doing pretty well at the right settings, it seems.

[andre-mz5]

This is going to be an interesting topic to follow for me since dosdan is a well known member who shoots sports.
Dosdan I'm curious if the 60-250 can deliver the same kind of shots since the AF of the Sigma 70-200 is among the fastest in the Pentax lens world.

[dosdan]

'm curious if the 60-250 can deliver the same kind of shots since the AF of the Sigma 70-200 is among the fastest in the Pentax lens world.

Yes that will be interesting.

I've noticed, after a lot of K-5 processing, that the DA* 60-250/F4 shots look harsh when you apply heavy cropping. I presume that's because lens aberrations become more obvious as you approach the 100% cropping level. So I'll be particularly looking out for this when I get to test the K-3 with this lens. The next few days are expected to be heatwave conditions (37°C), so unless I can get my son to do this early morning when it will only be about 24°C, I may have to wait a while before proceeding with further testing.

Dan.

---

Last edited by dosdan; 01-02-2014 at 06:49 AM.

[pearsaab]

can you comment on what photos settings you used (single point, sel, etc) ..thanks so much!!

[jb09ss]

This looks like good news for the tracking ability of the k-3. The hardware needed is all in the camera, with firmware updates and faster focusing lenses, my guess is that the k-3 should be able to be in the same league as the 7d for tracking.

Interesting fact, at the opposite side of the world (eastern Canada) this morning the indicated temperature in my car was -37C.

[RonHendriks1966]

That image looks very good.

[Miguel]

my guess is that the k-3 should be able to be in the same league as the 7d for tracking

Well, maybe in the same league, but most likely in the second division. Based on my limited experience shooting 6,000 sports frames with the K-3 and four years shooting over 100,000 frames with the 7D, I'd say that the K-3 is about 60% as capable. Forgive my generalization here. Unlike Canon, Pentax hasn't released a super-telephoto with fast focusing technology, so I'm hoping that lens release will improve the K-3's performance. I think the maker's own lenses should outperform a third party's--if not optically, then at least performance wise. So in my mind the K-3's action AF picture is quite incomplete.

In the meantime, I've gone back to using my 7D for well-lighted sports venues, but will continue to use the K-3 for darker, more challenging environments as its sensor is newer and better.

M

[dosdan]

Another shot from yesterday. Pulled back (darkened) -0.5 EV in PP. Sharpening: Emphatic in SilkyPix.

Portrait mode. Cropping: kept 1/3rd towards the bottom of the frame; cut 20% off the RHS.

1/1250s, f/2.8, ISO125, 105mm FL

First version - WB as shot. Bringing out the golden hour aspect:



2nd version - dropped the colour temp almost 600°K. This is the Sunset preset in SilkyPix:




Dan.

[normhead]

Well, maybe in the same league, but most likely in the second division. Based on my limited experience shooting 6,000 sports frames with the K-3 and four years shooting over 100,000 frames with the 7D, I'd say that the K-3 is about 60% as capable. Forgive my generalization here. Unlike Canon, Pentax hasn't released a super-telephoto with fast focusing technology, so I'm hoping that lens release will improve the K-3's performance. I think the maker's own lenses should outperform a third party's--if not optically, then at least performance wise. So in my mind the K-3's action AF picture is quite incomplete.

In the meantime, I've gone back to using my 7D for well-lighted sports venues, but will continue to use the K-3 for darker, more challenging environments as its sensor is newer and better.

M

Yes, that thing about not having fast focussing lenses is certainly note worthy. You get the feeling with a K-3 and 60-250, that it's the lens that's holding you back. When you get to focus it locks very quickly, but the time it takes to get to focus is the same as a K-5. On the other hand, I doubt I could afford a high performance lens, and what's available in Canon land isn't changing my mind at all.

I used the Sigma 70-200/F2.8 II HSM. To stress the AF I only used f/2.8-3.2. With the K-5 I would have been too scared to do this with someone running directly at the camera and would have chosen f/6.3-7.1.

I am not sure I understand what you mean by this statement. Is it not so that the AF measures with the lens fully open no matter the choosen f-stop?

[dosdan]

I am not sure I understand what you mean by this statement. Is it not so that the AF measures with the lens fully open no matter the choosen f-stop?

Yes AF is performed with a fully open aperture.

What I meant was, the AF performance is "stressed", in the sense that a minor AF miss is not concealed by a deep DOF. So if the AF isn't accurate, you will more easily notice it.

Dan

[Miguel]

what's available in Canon land isn't changing my mind at all.

Well, once a certain standard of AF tracking quality is achieved (which I'd say the K-3 meets), the "argument" then switches to lens systems--and that's were I'm frustrated with Pentax. I'm probably going to order the Tamron 150-600mm f5-6.3. It comes out in a few weeks, but sadly is not offered in K-mount. I wouldn't be surprised if it performs similarly to the Bigma. What has caught my attention in addition to the magical 600mm focal length, is that it has built in IS and seemingly complex optical components. Could be a plastic entombed POS, but at $1100 I'm game. My usage goal would be more perched birds and shorebirds, rather than BIFs or sports, but we'll see.

This is the kind of lens Pentax should have released simultaneously with the K-3. My 100-400L f4.5-5.6 has been around since 1998, surely that's been enough time for Pentax to come up with something similar.

M

Yes AF is performed with a fully open aperture.

What I meant was, the AF performance is "stressed", in the sense that a minor AF miss is not concealed by a deep DOF. So if the AF isn't accurate, you will more easily notice it.

Dan

Ok, thanks. That makes sense.