What is the best m43 for focus tracking?

Started Mar 31, 2013 | Questions thread
Anders W
Forum ProPosts: 17,073Gear list
Like?
Re: what is the best m43 for focus tracking?
In reply to PerL, Apr 2, 2013

PerL wrote:

Anders W wrote:

The only site I am aware of that actually tries to test AF-C performance in a systematic fashion is the French Fnac labs.

In one of their tests, they simulate a car approaching the camera at a speed of 50 km/h. The camera is fired in burst mode with a 200 mm (EFL) lens. At the outset, the car is fifty meters away. Obviously, it gets more and more difficult for the AF system to keep up, the closer the car gets. So one significant measure is the distance of the car from the camera when the last reasonably sharp shot is caught. Here are the results from this year's roundup:

CDAF

Panasonic GX1: 2.5

Panasonic G5: 8

Panasonic GF5: 8

Panasonic GH3: 9.5

Sony NEX F3: 11

Sony NEX 7: 12.5

Sony NEX 5R: 13

Sony NEX 6: 15

Samsung NX1000: 19

..........................

CDAF/PDAF

Nikon J2: 6.5

..........................

PDAF

Canon 650D: 6

Canon 7D: 6.5

Nikon D4: 7

Nikon D800: 8

Canon 5DIII: 8.5

Nikon D5100: 8.5

Nikon D600: 8.5

Canon 1100D: 9

Nikon D3100: 9.5

Pentax K-5: 9.5

Pentax K-30: 9.5

Nikon D3200: 10

Sony SLT-A57: 10

Nikon D7000: 10.5

Sony SLT-A65: 10.5

Canon 600D: 11.5

Sony SLT-A37: 11.5

Canon 60D: 12.5

Sony A-77: 13.5

As expected, the PDAF systems do better on average on this test. However, the difference is not all that big and the faster/better CDAF cameras beat the slower/worse PDAF cameras.

Source:

http://multimedia.fnac.com/multimedia/editorial/labo/reflex_2013_v8.pdf

Note three things:

First, as with most tests, there is a margin of error involved here. There are also some strange "outliers" like the extremely good (and therefore probably incorrect) results for the GX1. So I tend to look for broader patterns and not make too much of the results for individual cameras.

Second, I have excluded the results for the Olympus E-PM2 and E-M5 since the results indicate that the lab made the mistake of shooting both cameras in high-speed burst mode where they won't even try to AF between shots. The buffer has then filled rapidly, which explains the slow rate at the end of the series.

Third, apart from the distance at which the last sharp shot is caught, another measure of performance in this test is the total number of sharp shots the camera manages. My little table above does not include that measure but you can easily look at it yourself by means of the pdf I link to.

I have been critical about these tests before.

And I have dealt with that criticism in some detail. Those interested find my latest response in the exchange between us here:

http://www.dpreview.com/forums/post/50851802

If you Per have anything further to say, I suggest you start where we left off (the post I linked to) so that we don't have to repeat everything from the beginning again.

To many odd results IMO (even if one excludes the most strange ones like G2X)  that seems to be at odds with common sense and practical user experience. A few samples:

"Common sense" and "user experiences", both often rife with confirmation bias, are hardly infallible. That's precisely why we need systematic and objective tests like this, isn't it?

G5 and GF5 faster than G3H?

As I pointed out, a test like the one at issue here must by necessity involve a statististical error margin and you shouldn't put too much stock in each individual data point but look at broader patterns. The difference between the G5 and GF5 on the one hand and the GH3 isn't all that large and all three are in the latest generation of Panasonic MFT cameras. So I don't find too much to be surprised about here.

G5 and GF5 in the same ballpark as Nikon D4? (8 vs 7)

Yes. Isn't it amazing how quickly MFT has made progress, now rivalling the D4 in a test of AF-C?

Consider the fact that in the FNAC report of two years ago (the one where they tested the GH2 that you have personal experience with), the average value for the mirrorless camera tested was 16.25 meters. Two years later, that figure is down to 11. For MFT cameras specifically, the value two years ago was 13.5. By now, the average is down to 7.  In the meantime, the results for DSLRs haven't changed much. PDAF is a rather mature technology. CDAF has been developing rapidly in the last few years and is now catching up.

If we look in a broader view - the low end m43 Panasonics matches the semi pro cameras like Nikon D800 (same AF system as D4) and Canon 5D3?

See above.

The low end m43 Panasonics and the also slower GH3 clearly beat the prosumer and middle of the line DSLRs like Nikon D7000, Canon 60D, Sony A77?

See above.

Even the results in the DSLR group is confusing with entry level models beating the more expensive ones.

The only really unexpected result here is for the 650D. Other than that, the expected cameras (recent pro/semipro cameras) top the chart.

You can pick out a lot more strange stuff here, I would not put much trust in this.

So you trust only results that fit your preconceptions? A fool-proof way of never putting one's prejudices to any real trial.

 Anders W's gear list:Anders W's gear list
Panasonic Lumix DMC-G1 Olympus OM-D E-M5 Olympus E-M1 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G Vario 7-14mm F4 ASPH +21 more
Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow