• Welcome to BirdForum, the internet's largest birding community with thousands of members from all over the world. The forums are dedicated to wild birds, birding, binoculars and equipment and all that goes with it.

    Please register for an account to take part in the discussions in the forum, post your pictures in the gallery and more.
ZEISS DTI thermal imaging cameras. For more discoveries at night, and during the day.

E-M1 Mark2: C-AF a disappointment (1 Viewer)

Tord

Well-known member
This weekend I tested E-M1 M2 and also m4/3 300/F4.

When used in S-AF the E-M1 coupled with the 300/2.8 and the metabones + Canon 400/5.6 produced encouraging results. Fewer frames suffering from OOF compared with the E-M1. Still not 100% reliable, but OK.

The 300/4 was spot on when used in AF-S. So gave the setup a good workout in C-AF. However, having read all the praise of the improved AF system I must say I was really disappointed with the C-AF performance. Out of some 500 frames and after review in the computer I came to conclusion that I got zero (sic) keepers, all frames OOF.

Focus acquisition was fast, only at few occasions the setup would hesitate or fail. However quick review in field showed that something was wrong.

I tried the following settings:
AF points: All, central 9, central 5, even single AF.
C-AF, C-AF + tracking.
IS-1, IS off.

Finally, in desperation, I tested a static subject. S-AF locked as expected, whereas using C-AF/C-AF+Tr on same subject resulted in OOF result.

Attached:
The Tern: this is the best shot of 500+. At a first glance it looks semi-OK but closer inspection reveals it shows quite pronounced OOF.

The tulip is shot with single AF point, positioned so that it aims at the central tip of the central leaf
Shutter release prio OFF.
Camera locked focus immediately, no hesitation.
1/250s, F/4

The first is S-AF. More or less spot on.
The second is with C-AF. Notice the severe OOF (front).

Anyone else?
 

Attachments

  • P5260320.jpg
    P5260320.jpg
    242.7 KB · Views: 183
  • P5260320-2.jpg
    P5260320-2.jpg
    503.7 KB · Views: 218
  • P5270414.jpg
    P5270414.jpg
    440.8 KB · Views: 228
  • P5270413.jpg
    P5270413.jpg
    355.6 KB · Views: 218
Tested on a focus chart. The first is S-AF, the second C-AF + TR
 

Attachments

  • P5280001.JPG
    P5280001.JPG
    413.3 KB · Views: 196
  • P5280011.JPG
    P5280011.JPG
    411 KB · Views: 188
I am getting similar results to yours on a different chart: S-AF (and C-AF) is spot on, C-AF+Tr focuses too close. I used the 12-100mm F4 PRO lens for the test. I will retest it with my 75-300mm telephoto.

Camera firmware version 1.2, lens firmware version 1.1
 
Have had acceptable C-AF results with E-M1 M2 and 40-150 f/2.8 PRO with and without MC-14.
 

Attachments

  • 165062489.hlNSKgrf.jpg
    165062489.hlNSKgrf.jpg
    81.4 KB · Views: 213
  • 165088606.guDFI6pR.jpg
    165088606.guDFI6pR.jpg
    72.8 KB · Views: 232
Last edited:
I redid the tests today, trying to be as scientific as possible. C-AF+Tr fails consistently on 12-100 (failed 6 times out of 6), and works almost perfectly on 75-300 lens (failed 1.5 times out of 7 - the 0.5 is due to a photo with focus slightly off). Tested on 75mm focal length. The status of Image Stabilisation doesn't matter.

12-100mm F4 PRO example:
12_100_CAF_Tr.jpg

75-300mm example:
75_300_CAF_Tr.jpg
 
No idea what is going on there. One would expect it to work at that price.
If you add +1, does it affect both S-AF and C-AF and/or C-AF tr?
I updated my Mark II to Firmware 1.2 and the 400+Metabones seems a good bit faster! I don't think it is my imagination. EVF looks better too, more natural.
 
One would, right? But manufacturers keep surprising us. I wrote to Olympus and asked them about it.

The change only affects C-AF+Tr.

Improvements to EVF display were listed as a feature of this upgrade.
 
Something must be buggy with Tord's test setup. What I have seen of the C-AF tests is impressive indeed. Can't believe that it could get through production being that far off.
 
Very strange...

From the manual:
Press the shutter button halfway to focus; the camera then
tracks and maintains focus on the current subject while the
shutter button is held in this position.
• The AF target is displayed in red if the camera can no longer
track the subject. Release the shutter button and then frame
the subject again and press the shutter button halfway.
• Tracking range will be narrow when using a Four Thirds system
lens. Autofocus does not work while the AF target is displayed in
red even if the camera is tracking the subject.

Also page 111
C-AF Lock
Sets the tracking sensitivity for C-AF.
 
Last edited:
My first observations of firmware 1.2 are:

Faster and more direct S-AF with both FT lenses and my Canon 400/5.6/Metaqbones setup. C-AF still useless with the Metabones, but that is not Oly's problem! ;-)

Single AF point more selective with FT/adapted lense. Better able to pick LBJs out of the branches. Some people wanted a smaller single point, but the fact that it is more selective makes up for it.

EVF more like the Mark I, that is, more "natural" looking.

Early days yet, but so far I am pleased. Lets remember how much the Mark I was improved with all the firmware updates.
 
Have had acceptable C-AF results with E-M1 M2 and 40-150 f/2.8 PRO with and without MC-14.

Neither of those two shots look in focus to me. Dog, water in lower right is sharp, but nothing on the dog. Maybe too much movement to tell for sure.
Bird, focus seems to be on the twigs behind the bird. The bird is generally unsharp.
 
So, I have spent the last couple of hours trying to replicate what you guys are observing. As I have never bothered with C-AF (more trouble than it is worth) I haven't really tested it. The only mFT lens I have is the 75-300 II, so I put it on the tripod and set up Jenny Lind. ;) Distance about 5 meters. I observed:
The CDAF is in general not as consistent and accurate as PDAF, that is a known, and that is the first thing that struck me. There is quite a bit of variation (about +/- 1 or even more on my scale on the left = about 1cm) between shots at any given setting. I de-focused the lens between shots. The AF-Adjust helps, but only somewhat. Minus 1 on the long end seems to be more or less what works, but really more or less. The other thing is that there seems to in fact be a slight amount of front focusing in C-AF compared to S-AF. Not a huge amount, but more than there should be, but there again, the inconsistency in the CDAF makes it a little hard to judge. There were times when it was the other way around. It certainly doesn't come close to the bang on consistency I get from the Canon/Metabones/PDAF. Here are a couple of sort of average shots...
34621905060_ad409a30a9_k.jpg 34165597064_a6bf63650a_k.jpg
 
Last edited:
I saw no difference in the behavior between C-AF and C-AF+TR, at least on this static target.
By the way, all done with single AF point here:
34878624361_fa3a0d9517_c.jpg

We should also keep in mind that there is no such thing as a 100% reliable AF system. There are SO many variables. You can get 100% with the right setup on the right target, but that has little to do with the real world. I get 100% with the Mark II/Canon 400/Metabones on this kind of target though, and I certainly do NOT get it with the mFT CDAF.
 
Last edited:
Ah, you used 75-300II for this test! I had no issues with this lens either. But both me and dwever had issues with 12-100mm F4 PRO. It must be something lens-specific.
 
After advise from other EM-1 Mark 2 users I switched to L-FPS (10 fps) when using C-AF. The results I get are now very good, much better than any I got with the E-M1 Mark 1. I now consistently get frame/s with focus spot on in almost every sequences, even when photographing challenging subjects. How many you get depend on how good you are at tracking the subject.

Still experimenting with the settings. 5 or 9 AF points seem to give the best results.

Moreover the AF system is prompt in re-acquiring focus after losing it.


Pity that the user manual is not that clear in that C-AF behaves like S-AF once focus has been acquired, as focus is locked throughout the sequence.
 
Warning! This thread is more than 7 years ago old.
It's likely that no further discussion is required, in which case we recommend starting a new thread. If however you feel your response is required you can still do so.

Users who are viewing this thread

Back
Top