E-M1 II: Testing PDAF Single point precision - long post, many pics

Started 7 months ago | Discussions thread
ForumParentFirstPrevious
Flat view
Messier Object Veteran Member • Posts: 6,225
E-M1 II: Testing PDAF Single point precision - long post, many pics
23

I've mentioned a couple of times that I'm not happy with the size of the Single AF point, and that its effective area seems to be larger than the indicated box. This leads to focussing errors and a lower keeper rate than I get with my E-5 and Canon DSLRs which all have a Single Point Small option. Others here have also made comments regarding the loss of the Small point option from the E-M1 to the Mark II.

I only shoot with 4/3 lenses and according to the manual PDAF is automatically selected by the camera when shooting the 4/3 lenses, but I have my doubts about that.

Here are some results from recent testing.

E-M1 Mk II Firmware 1.1, ZD300mm F/2.8, S-AF + MF, tripod mounted

My test target was a whiteboard set at about 13m with hand drawn targets.

My first test was to establish the Effective size of the single AF box.
I drew a rectangle on the centre of the board such that the AF box in the viewfinder overlaid the target. After confirming that S-AF was working and locking onto the target I re-aimed the camera on to the blank part of the whiteboard to the left of the target and tried the AF.
Moving progressively back towards the target, I tested the AF to find the point at which AF locked the target.

Here are the results.

The images are in pairs: The first of each pair shows a dotted outline of the indicated AF Box in the viewfinder, ie this is where the camera was aiming. The second of each pair has the dotted box erased before the shot was taken using S-AF. Focus was reset before each second shot

All images are uncropped

1.1 far left

1.2   S-AF hunted and failed - expected result

2.1 medium left

2.2 Almost focussed. No hunting, but the AF was slow. This result was repeated several times - unexpected unless CDAF is in play

3.1 close left

3.2 Effective AF edge Left. fast Sharp Focus. No hunting. Did you expect this ?

I repeated this above, below and to the right to make sure the results were symmetrical

here is the 'below' result

5.1 Close Below

5.2 Effective AF edge below. fast Sharp Focus. No hunting

I was then able to draw a box around the original AF target showing the enlarged Effective size of the Single AF box

I'm sure that many here will find this result unexpected. But that's not the end of it because each of the above tests were done with the camera's indicated AF box on a blank part of the whiteboard, with nothing on which to focus. That there is some 'leakage' to an available target just outside that blank area is not surprising if we assume that CDAF is also in play - despite what the user manual says about 4/3 lenses using only the PDAF

The practical up shot of this is whether the larger Effective AF box area comes into play when shooting real targets like birds in a tree with surrounding foliage. Does the AF go for the foreground branch that's inside the Effective box rather than the soft feathers of the bird in the centre of the indicated box.

In order to test this I setup a foreground object (cardboard carton) such that the edge of the carton intersected the Effective AF box but not intersecting the smaller Indicated AF box.

Firstly I tried it with a nice high contrast target - a nice fat cross drawn on the whiteboard

The AF was not distracted by the foreground if the target was an easy one. Repeated result = Good !

next I tried a harder AF target. I replaced the '+' with a group of dots. First I made sure that the dots were not too hard a target for the AF system. AF nailed it repeatedly and quickly without any problems.

Sharp focus, repeatedly and fast = good result

Next I moved the foreground object back inside the Effective AF box.
Here it is Manually focussed to show you the placement of the foreground object

Manually focussed on the target

.

And this is what I got with S-AF

Not a good result. A more attractive target outside of the indicated AF box has taken AF priority

I repeated this several times. Sometimes the foreground object was sharp, other times it was a tad out of focus like in the above image, but in all cases the AF system ignored the real target on the whiteboard that was in the middle of the indicated AF box, that  very same target which it was able to reliably and quickly lock into sharp focus without the foreground object.  But as soon as I moved the foreground object back outside of the Effective box the AF system was once again able to lock onto the intended target.

.
CONCLUSIONS

When using 4/3 lenses (the 300mm f/2.8 specifically):

1.   the Single AF box is  imprecise. Its effective size is very much larger than that indicated in the viewfinder

2.  When using the Single AF point the AF system can be easily distracted by a more attractive target outside of the indicated AF box. This is troubling and is the probable cause of  the abnormally high number of soft bird shots I get shooting with this camera/lens combo at f/2.8

I'd be happy to hear from anyone who can explain these results or show me how to better configure the camera to avoid this issue.
Peter

 Messier Object's gear list:Messier Object's gear list
Olympus E-5 Canon EOS 5D Mark III Canon EOS 7D Mark II Olympus E-M1 II Olympus Zuiko Digital ED 50-200mm 1:2.8-3.5 SWD +11 more
Olympus E-5 Olympus E-M1
If you believe there are incorrect tags, please send us this post using our feedback form.
ForumParentFirstPrevious
Flat view
Post (hide subjects) Posted by
ForumParentFirstPrevious
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow