Pen F users - some quick questions...

MEDISN wrote:

whumber wrote:

Raist3d wrote:

You win. That’s what matters

Important to note than in MEDISN's example the subject is moving at approximately 2mph. That said, it's still a very good example for the Pen-F, I never saw results that good. I don't know if he's able to get results like that consistently or not but given that there's not much to configure on the Pen-F maybe he's just better than pressing the shutter button than everyone else, =). That or maybe Olympus made some stealth FW updates.

Sounds about right. 2mph = 3ft/sec, so roughly 7 inches per frame in the burst. The DOF at the beginning of the sequence was ~ 10 inches and narrowed to ~3 inches by the end. You would think the two OOF shots would be more likely to occur at the end where DOF is much shallower but they occurred at shot 16 and 26 so probably just random misses.

16 I can see being a little out because the AF point grabbed onto the hair on the left but then I would have expected 13 to be out as well. In 26 the AF point is right on the face so it could be a miss, or if Olympus is doing some kind of temporal smoothing it could be the af point miss at 23 finally catching up. Pure conjecture though as I don't know what they're actually doing behind the scenes. In any case, none of them are really far out, at worst they're slightly soft. Even at the speed the subject is moving this is a solid performance for a CDAF system. View the image at 100% to get the animated GIF to play.

Notice I wasn't rack zooming either as hit rate drops with all of my mirrorless including the A9 and EM1X. Not so bad when the subject is coming toward the camera but everything starts to struggle when the subject is moving away from the camera. Or may just be my technique.

The only mirrorless camera I've had good success with while racking zoom was the X-T3. If Olympus could fix that issue in the E-M1X I think I'd be pretty much set. That's kind of disappointing to hear that the A9 suffers the same issue, the only experience I have with the A9 was with the 85 GM so never got a chance to test that in person. I do find it kind of hilarious that nobody other than Michael the Maven on youtube seems to think that this is an AF behavior worth testing. BTW, how are you liking the E-M1X? I think last time we discussed C-AF behavior you were still shooting the E-M1ii. I don't know if you shoot any winter sports or other activities, but if you haven't tried it with winter gloves on yet I think you're in for a really pleasant surprise.

If you really want to make a camera hurt, try the swingset torture test. Get in close and shoot wide open with a fast prime, eye-AF enabled. Hold the shutter down until it starts buffering. When you sort through in post you might get 15-20% sharp on the subjects face even with the A9 and EM1mkII

Yeah, that's a seriously brutal test, I had recommended that test to the DPReview guys a year or so ago in one of the review comments and they seemed to think it was a great idea but I guess not enough to start implementing in their testing. When I've run these kinds of tests with DFD cameras like the GH5 and now G9 they seem to just give up and the only in focus shots are when the subject happens to pass through the current focus plane, kind of a broken clock thing, =).

The 34 shot PEN-F sequence was firmware 3.0 but I don't recall experiencing any dramatic change in AF going back to 2016 when the camera was released.

My experience with the Pen-F was never quite as bad as what TCSTV posted when they shot with the 300 PRO, but it was very inconsistent and I don't remember ever getting more than half or so in focus, even with relatively simple motion. Certainly nothing like what you posted above with any form of reliability, but then again I never invested too much effort into the Pen-F C-AF based on my initial experiences coupled with the reviews.

As others have posted here it may well be lens dependent. My assumption would be that many PEN-F's are paired with compact budget primes most of the time which may yield disappointing results. Then factor in the other variables like settings, target contrast, light levels, and user ability and it's easy to see how results can vary.

This is a good point, I tend not to save sequence with lot of OOF shots so I can't really dig in but most of the time I was using either the 17 1.8 or 75 1.8 with the Pen-F. I did do some testing with the 40-150 and don't remember it being noticeably better but maybe there was a camera or lens firmware update that made some difference.

The types of motion I shoot CAF with the PEN-F tend to be street/lifestyle/travel candid's and portraits plus kids at play as in the sequence above. For that type of shooting the PEN-F doesn't disappoint. I've never tried motorsports or surfing with it as it isn't weather sealed and the big lenses tend to balance better on the larger bodies.

I really liked it as a travel camera, I only got rid of mine because I wanted to shoot with the PRO prime lenses more and it really didn't balance well in my hand with those.