Is C-AF actually improved with v3.0?

[brian1208]

Its early days yet Ron and I haven't had the chance to try capturing birds against a complex background yet (don't expect much luck using all points though )

What I have found is that if it has got a lock before a gull passes in front of or behind foliage its holds lock very well (CAF lock set to Normal).

As I say, roll on some decent light and subjects so I can give it a real work out

[brian1208]

An interesting moment this evening when I lifted the camera to my eye and pressed focus in portrait mode, CAF at 9fps and the system hunted like a md thing, not once but on successive attempts - panic!

then I checked my settings and realised I had left it in large single point focus, switched it to all points and bingo, back to its usual rapid lock and capture

The Martins still aren't coming close yet, this is a 100% crop at around 110 pixels on the longest side

martin, CAF portrait 9fps by Brian Wadie Photographer, on Flickr

[David Morison]

Originally Posted by brian1208

too true David but if its essential to get a camera with cross type sensors to gain the best CAF then Olympus isn't the tool to do it.

I'm happy to compromise on this as, having tried other brands in the "Light-weight" end of the spectrum ,I have returned to Olympus as it provides me with the best compromise on price, performance and enjoyment when shooting with it

We mustn't forget Brian that you are an acknowledged expert in this field but you will accept that it is a bit of a minefield for those who are new to wildlife photography and haven't cut their teeth on DSLRs like us. It would be easy to be taken in by marketing and imagine that BIFs are going to be straightforward with the OMD series.

Pretty good Martin shots!

David

[brian1208]

thanks for the compliment David, but I would suggest that any photography is a bit of a minefield to start with, whether it be trying to shoot BIF with DSLR, macro with a dedicated macro lens etc.

I don't think that m4/3rds for birding is inherently "difficult" but does require technique to be learnt.

I have found it more difficult to teach people moving from DSLR to m4/3rds than I have people who have never shot a DSLR in the first place to use an m4/3rds for shooting BIF or sport because the "similarities" between them confuse as, in my opinion, totally different camera handling techniques are needed to get similar results

[pvasc]

I picked mine up this weekend. On the way back I had a stopover in Milton Keynes so what better opportunity to find out the tracking problem than a train coming head on at 125mph. No problem there. I was using Box 9 focus points, 6fps high speed sequential, C-AF, and the 12-40mm. Now at first it DID lock on to the weeds on the side of the tracks, and we know why. So I though a big old train is not a small bird in flight. I have managed to catch some, much better than with the M-10, but, of course, there were not many out flying while I was out wanting to shoot them. I did manage to get this. I thought that was pretty good, just not sharp because I moved, it looked like he was going to fly right into me.



I went out later and there were some small birds flying, and I did notice it lock then hunt, then lock.... I was using the 40-150mm......F4-F5.6.

Earlier while nothing was flying I had on the 50-200mm lens, and a little bug smasher flew by. The lens struggled to lock, turned to portrait orientation, instant lock.

It's a very nice camera, leaps ahead of the M-10, I shouldn't be surprised but I am.

[brian1208]

its not just me finding this works then

nicely done

[pvasc]

Thanks

That Goose was with the 75-300mm held landscape orientation but the plane, with the 50-200mm, it hunted and hunted, then I turned the camera 90 degrees and BAM lock and track. I haven't had the bird availability to check out the other lenses. But if it works for you....I can't wait to see.

[Gwyver]

Please could those of you who are achieving good results for BIF & aviation photography when using SeqH + C-AF recommend the optimum settings for the following:
CogC Release

• Image Stabilization (formerly named Sequential + IS Off)
• Half Way Rls With IS

I'm confused as to whether these settings have any relevance if you have specified IS=Off (as might be expected when doing rapid panning if you aren't convinced that IS=Auto will give optimum results).

Many thanks

[brian1208]

I can only comment on what works for me (I believe others have different settings?)

- Release setting so it will release the shutter without confirmation of focus (I can never remember whether that is "Off" or "On" and my EM-1 is packed away for an early start tomorrow)
- IS auto
- IS set to work between sequential shots
- body in portrait mode
- set to square format as I find this helps my mind cope with framing a moving object with the camera in portrait orientation
- all squares active (or for larger subjects, 9 box active) focus points
- 9 fps

[brian1208]

Well, I am totally convinced that CAF works beautifully at 9fps, so long as the camera body is in portrait mode.

I spent yesterday at or local Bournemouth Air Festival in pretty dire light conditions with sea haze, shooting into the sun, only relieved when the cloud finally rolled in.

I tried SAF+ M and CAF in both portrait and landscape camera positions and the one that most consistently got me the sequence of in focus / sharp shots was CAF in portrait mode, up to and including the Typhoon barrelling toward me at 600+ mph.

What surprised me was how often SAF + M hesitated in gaining focus lock under these conditions, it seemed to grab first lock then shuttle back and forth a few times before confirming lock.

With CAF in portrait mode there appeared to be no similar effect, with more or less instant lock on and hold (focus lock release set to Normal )

I'm only part way through looking at all the many pics but most CAF approach sequences seem consistent (this does however mean that if it started firing before lock was acquired the whole sequence is OOF )

My only problem with these fast jets was keeping the darned things in the right part of frame when shooting in portrait mode, easy when coming toward me but really tough when panning at around 600mph

[damianmkv]

Why does it work better in portrait than landscape mode ? It unnatural I think

[pvasc]

Originally Posted by brian1208

Well, I am totally convinced that CAF works beautifully at 9fps, so long as the camera body is in portrait mode.

I spent yesterday at or local Bournemouth Air Festival in pretty dire light conditions with sea haze, shooting into the sun, only relieved when the cloud finally rolled in.

I tried SAF+ M and CAF in both portrait and landscape camera positions and the one that most consistently got me the sequence of in focus / sharp shots was CAF in portrait mode, up to and including the Typhoon barrelling toward me at 600+ mph.

What surprised me was how often SAF + M hesitated in gaining focus lock under these conditions, it seemed to grab first lock then shuttle back and forth a few times before confirming lock.

With CAF in portrait mode there appeared to be no similar effect, with more or less instant lock on and hold (focus lock release set to Normal )

I'm only part way through looking at all the many pics but most CAF approach sequences seem consistent (this does however mean that if it started firing before lock was acquired the whole sequence is OOF )

My only problem with these fast jets was keeping the darned things in the right part of frame when shooting in portrait mode, easy when coming toward me but really tough when panning at around 600mph

What lens? I noticed a night and day difference with the 50-200mm.

[brian1208]

Why does it work better in portrait than landscape mode ? It unnatural I think

as I understand it, its a consequence of the PDAF sensors being linear rather than cross format, things like birds and plane present an horizontal image to the sensor and is not well seen. By turning to portrait mode the planes cross the PADAF sensor and are easily "seen" and focused on (I vaguely remember this being mentioned in the manual?)

What ever the reason the effect is very clear

What lens?

Sorry - should have said, all shot with 40-150 f2.8 + 1.4X TC

I've had a quick look in the manual and the only reference I can see at first glance is that it may be difficult to auto focus on subjects that contain only horizontal lines (so logic would suggest that turning the camera from landscape to portrait mode will resolve that issue by converting the horizontal lines into vertical lines )

[Ross the fiddler]

Originally Posted by pvasc

What lens? I noticed a night and day difference with the 50-200mm.

This guy loves it too. http://www.furnfeather.net/Reviews/3.1_Firmware.html

[Gwyver]

Originally Posted by brian1208

as I understand it, its a consequence of the PDAF sensors being linear rather than cross format, things like birds and plane present an horizontal image to the sensor and is not well seen. By turning to portrait mode the planes cross the PADAF sensor and are easily "seen" and focused on (I vaguely remember this being mentioned in the manual?)

What ever the reason the effect is very clear

I also thought the advice about turning the camera to portrait mode to achieve AF appeared in the manual - though I'm unable to find it now. What it does say (p140) is: "It may be difficult to focus with auto focus in the following situations."..... Item 3 = "Subject containing no vertical lines"

It seems as if, from the user's perspective, the design decision regarding orientation of the PDAF sensors was flawed - though this may be constrained by the direction of the cell wiring on the sensor chip. A PD sensor re-orientation would benefit not only C-AF where horizontal motion seems to be most common, but also many landscapes which only 'need' S-AF have more horizontal than vertical tonal transitions.

Maybe Olympus will rectify this next time - or perhaps it was a deliberate marketing choice to encourage 43 users to transition to m43?