G3 heralds the end of APS-C DSLRs

Started May 13, 2011 | Discussions thread
Anders W Forum Pro • Posts: 21,468
Re: and that's before we talk AF micro-adjust! ;-)

ljfinger wrote:

Anders W wrote:

Why on earth should I consider the DPR an authority on this.

I am still waiting for your evidence with regard to the following two claims of yours:

1. AF evaluation frequency does not matter with respect to CDAF speed.

2. AF motor does not move continuously in CDAF but stops and starts at each evaluation.

Of course, it's not irrelevant because "static subjects" are often less than perfectly static.

Then the CDAF system fails.

On what grounds?

It can't keep up with or predict focus position.

Does focus change when a person changes his or her facial expression?

If the facial expression is what changes, they they are "perfectly static" in this context (focusing performance).

Let me try to explain this once more since you seem to have difficulties getting the point. If a person changes his or her facial expression and nothing else, then he or she is perfectly static from a focus point of view. But he or she is not static with regard to facial expression. Hence AF-S time is important, which is what you originally disputed. Do you understand or should I explain again?

Third, most compacts can achieve similar numbers. They're all dirt-slow.

What compacts go equal with the GH2. And which DSLRs are faster?

Most compacts issued in the last 5 years or so are similar, and all DSLRs are faster.

Lots of hot air. Claims and claims but not a shred of evidence. Where are your figures and measurements? I am waiting.

I make my own. All three of my compacts can achieve initial focus in less than half a second.

Which three compacts do you have? And which DSLRs do you claim are faster?

The GH2 achieves focus in about a quarter of a second, not a half, based on the tests conducted by Imaging Resources. Now am I waiting eagerly for your proposal as to which cameras are faster. Some are, undoubtedly. But it remains to be seen which.

Focusing speed is important for tracking moving subjects, not for achieving initial focus.

See above.

Reports on the forums are not a good overall cross section.

Why not?

Because a much higher percentage of people with problems complain than people with no problems report success.

And? Please carry out the argument to its completion and I'd be happy to prove its lack of validity.

If a hundred million Toyotas are sold, and a hundred report "sudden acceleration", you hear about it all over the news. The other 99.9999 million are silent. And we know how that turned out.

Yup. Toyota had to recall millions of cars to fix the problem. Manufacturers with no complaints about sudden acceleration didn't have to recall any.

On any DSLR forum, a considerable proportion of the threads deal with AF problems of one kind or another. On this forum and the NEX forum, you are hard pressed to find any such threads. Why is that?

The fact that these cameras have sold in the tens of millions over the years indicates that they are staying sold, which they wouldn't if they didn't work. So they do work.

It just indicates that until now there has been nothing better.

There still isn't.

Please try to stay on the subject. You have great difficiulties with that. In what way does your argument stay valid in view of my counterargument. Old cars and old cameras have sold in the millions. How does that prove that they are superior to modern cars or modern cameras?

You're the one wandering. You're claiming PDAF is unreliable. If it were as unreliable as you claim, people wouldn't use it. But they do, as evidenced by the sales numbers as well as the usage of all those people.

Then it's broken or horribly poorly designed, or you don't know how to use it.

Nope. There's systematic, publicly available evidence to the contary, and you already know which.

Never seen any, and it's contrary to my personal experience.

I thought you looked at at least one of the tests conducted by optyczne.pl. There are at least a hundred or so more more telling the same story about the comparison between PDAF and CDAF as far as accuracy is concerned.

PDAF is worse for accuracy than CDAF because it is a) subject to more systematic error ( the multiple path problem), b) subject to more random error (because its sensors and and control system aren't as accurate as those of CDAF). I have a lot evidence for both conclusions. Where's yours?

I have used both extensively under lots of conditions (hundreds of thousands of shots from 6mm to over 3000mm from sunlight to EV-6) and PDAF is far, far more reliable. CDAF is accurate when it hits, but it misses badly far more often. I'd much rather get a nearly-perfect shot nearly all the time with slight misses on rare occasions than a dead-perfect shot sometimes and total misses other times, which is what I get with CDAF. And with CDAF, I get virtually 100% misses on fast-moving subjects.

Based on everything you have said here (and in previous threads about similar topics), you are averse to CDAF cameras in general and m43 cameras in particular. Your profile doesn't list any m43 camera. So I doubt that you have much experience using a modern, high-performance CDAF system like that on the GH2. And I doubt even more that you have given it a fair chance if indeed you tried it.

But this is not the main point. It's still just your subjective impression. So what do you have to show in support of your contention in terms of publicly available evidence. I have pointed out mine. You have still to show yours.

 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 +28 more
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow