[derekkite]

Forgive me if I'm misunderstanding but the inability of those cameras to not focus in those situations is because they aren't based on phase detection for their auto focusing. Not sure where sensor size comes into this.

Contrast detect speed is a factor of how much data it has to process, ie larger sensor. With qualifications. As well a larger sensor requires more accurate and quicker focus due to smaller depth of field. What we have seen is contrast detect focus working very well in systems that allow it to work very well. Small sensors, short focal length.

A friend just got his D7100, and he has the very nice nikkor 300 f2.8 vr lens. Focus is almost instantaneous on that lens, and the body can take full advantage of it. His D7000 couldn't. Contrast detect can't know which way to focus, phase detect can.

Are there any pdaf mirrorless systems out there, other than the latest Canon innovation? Phase detection usually requires a separate array of sensors with the attendant light reflection mechanism.

[Rondec]

Contrast detect speed is a factor of how much data it has to process, ie larger sensor. With qualifications. As well a larger sensor requires more accurate and quicker focus due to smaller depth of field. What we have seen is contrast detect focus working very well in systems that allow it to work very well. Small sensors, short focal length.

A friend just got his D7100, and he has the very nice nikkor 300 f2.8 vr lens. Focus is almost instantaneous on that lens, and the body can take full advantage of it. His D7000 couldn't. Contrast detect can't know which way to focus, phase detect can.

Are there any pdaf mirrorless systems out there, other than the latest Canon innovation? Phase detection usually requires a separate array of sensors with the attendant light reflection mechanism.

Well, the V1 series has PDAF on the sensor. And that is the main reason why it has speedier AF than mirrorless options that have a contrast detect auto focus.

[johnmflores]

If I was going to make the motorcycle shot with my Q,
I would just have pre-focused on that point in the road,
and timed to allow for the shutter delay.

I used to do that with my MX! And I do that with my Q and the 03 Fish Eye:




But the Nikon 1 AF is so blazing fast that I don't have to resort to antique methods LOL.

I guess I am too used to the SI/ESPN style of shot, where there is isolation of a sharp player against a blurred out field. But the more narrow the depth of field, the tougher it is for the auto focus system to keep up with things.

AFAIK, Sony has introduced PDAF on one of their NEX cameras, and it's slow. And Canon's just released something using technology similar to the Nikon on-chip PDAF, so we'll see if the smaller CX sensor has in some way helped the V1 AF or if the technology can evolve to larger sensors.

And that's the reason why I brought up the V1 in the first place, to counter perceptions that mirrorless/EVF cameras can't shoot action. My experience with the Nikon V1 is that this first generation technology is already faster and more reliable than my K-5. I certainly hope that Pentax is looking into similar technology, lest they fall even further behind the AF curve.

At the end of the day, the V1's usability is quite poor for someone like me that is used to setting everything manually. If it had typically slow mirrorless AF.C I might have ditched it by now. But lightning fast AF is quite compelling, enough to keep me on board. For soccer moms and dads, fast AF may very well be the killer app.

[IchabodCrane]

delete -- redundant.

[jatrax]

delete -- redundant.

actually I suspect this entire thread is redundant......

[Barry Pearson]

You're forcing this analogy, but it's as invalid as most analogies. There's no connection between your punch cards and teletypes and the optical reflex viewfinder. And, how funny, you're talking as if the current SLR viewfinder systems don't have electronics and are not controlled by computers...

Keyboards are essentially mechanical devices, and so are mices. Printers are essentially mechanical as well, with lots of moving parts - they have to mechanically feed a sheet of paper and then imprint an image using a mechanical system (cylinders or moving heads).
For typing, as I said, you can't easily replace a keyboard: the tactile feedback you're getting is priceless (and by the way, many are preferring the old keyboards because of the better feedback). There are cool gadgets like a projector-camera system, projecting a virtual keyboard on your desk on which you can type. That would make your fingers hurt in no time.
About VoiceType (sold by IBM few years ago), what device would you use to correct the voice recognition's mistakes?

What you're forgetting is that the movable mirror's role is to allow a TTL optical viewfinder (and this is the important part, not the mirror). You're not arguing for the replacement of the mechanical parts, but also for the optical ones.
By the way, how many digital cameras can survive for 40-50 years?

My K-5 IIs is able to shoot at 7 fps and it has a 11 point AF system (9 being cross type, able to work down to -3EV). 7fps is not exactly impressive, and the 11 point AF system is much less so. Could you spot any improvement from your S1A and SFX-n?
And... there is no Moore's Law. You're trying to extend something which doesn't exist to the EVFs, nice try but it won't work.

See Moore's Law in Wikipedia! If you mean "it is an observation, not a "physical law", true, but so what? It makes no difference here that it is a reliable observation rather than a physical law! Here is a quote from Wikipedia which reveals how relevant and important it is here:

"Moore's law is the observation that, over the history of computing hardware, the number of transistors on integrated circuits doubles approximately every two years. The period often quoted as "18 months" is due to Intel executive David House, who predicted that period for a doubling in chip performance (being a combination of the effect of more transistors and their being faster).

"The law is named after Intel co-founder Gordon E. Moore, who described the trend in his 1965 paper.The paper noted that the number of components in integrated circuits had doubled every year from the invention of the integrated circuit in 1958 until 1965 and predicted that the trend would continue "for at least ten years".His prediction has proven to be uncannily accurate, in part because the law is now used in the semiconductor industry to guide long-term planning and to set targets for research and development.

"The capabilities of many digital electronic devices are strongly linked to Moore's law: processing speed, memory capacity, sensors and even the number and size of pixels in digital cameras.All of these are improving at (roughly) exponential rates as well (see Other formulations and similar laws). This exponential improvement has dramatically enhanced the impact of digital electronics in nearly every segment of the world economy.Moore's law describes a driving force of technological and social change in the late 20th and early 21st centuries."

The analogy I gave is good - because it is an analogy, not equality or identity! ("Analogy": "A relationship of resemblance or equivalence between two situations, people, or objects, especially when used as a basis for explanation or extrapolation"). It is my observation (and also obvious to the world) that we are step by step replacing mechanical actions with electronic ways of achieving the requirements, and that this brings down per-instance costs, increases the scope of innovation, and typically increases reliability. I am using my personal experience of computers since I first programmed one in 1965 to extrapolate to other electronic subsystems, especially EVFs. It is common to start by supplementing the mechanical actions with electronic ones, but that often isn't the end. Disc drives have more electronics associated with them than when they were first introduced. But solid state discs do away with the mechanical bits altogether. The same is happening to viewfinders - of course OVFs have been supplemented by electronics over the years. But now EVFs do away with the mechanical bits.

See also "Why Moore's Law Applies to Photography".

[lytrytyr]

"The capabilities of many digital electronic devices are strongly linked to Moore's law: . . . and even the number and size of pixels in digital cameras. All of these are improving at (roughly) exponential rates as well. This exponential improvement . . ."

It's debatable as to whether more and smaller pixels may be considered to be an improvement anymore.
Especially if our lenses can't keep up.

[lytrytyr]

actually I suspect this entire thread is redundant.

As "K-3 speculation", almost certainly.

But for me, the value of threads like this
lies in the various side issues
that get discussed along the way.

It's like the proverbial "rock soup":
Put a rock into boiling water,
add various herbs, leftovers, and other ingredients,
let it simmer for several hours,
then remove the rock before serving.

[Winder]

It's debatable as to whether more and smaller pixels may be considered to be an improvement anymore.
Especially if our lenses can't keep up.

There are many parts to this equation. Yes, there are the MPs of the camera and the quality of the glass on the capture side. There is the quality of the RAW processor or the JPEG engine on the processing side. Then there is the quality of the monitor or printer on the output side.

If the user only shoots JPEG, uses a super-soft, super-slow, super-zoom, and shares pictures on Facebook with the occasional Wal-mart 4x6 print, then 6MP is overkill.

Lots of people will download RAW files of a D800 and a K-5, process them on a 8-bit 1280 x 1040 monitor and proclaim that the K-5 is just as good. Or the make prints with their 10 year old Kodak Easy-share all-in-one printer to demonstrate that there is no difference in image quality.

For people with the right equipment the added MPs do indeed make a big difference. When I upgraded to a 2560 x 1440 monitor it was like getting new glasses. I realized how much detail that was being captured by my K-5 that I was missing with my old monitor. It improved my post processing which improved the quality of my prints.

I would welcome 24MP APS-C with the same S/N ratio and DR as the 4 year old 16MP sensor of the K-5. A 32MP would be even better. Processing power and memory/storage are pretty cheap these days, so I don't view that as a problem.

4096 × 3112 displays (12MP) are not very far off.

[VoiceOfReason]

.

4096 × 3112 displays (12MP) are not very far off.

I almost bought a 50 inch 4k TV to use primarily as a secondary or tertiary display. I still may do that. I do pixel peep!

[Kunzite]

See Moore's Law in Wikipedia! If you mean "it is an observation, not a "physical law", true, but so what? It makes no difference here that it is a reliable observation rather than a physical law!

It's not a law, period; it never was, but an empirical observation. And in the recent years I have the impression that, instead of observing the reality, it's the reality which adjust itself to follow the "law"
By the way, I work in the IT industry so I'm quite familiar with the concept. Thanks for the explanations, though.

The analogy I gave is good - because it is an analogy, not equality or identity!

I'm sorry, but it doesn't works this way. An analogy is not good because it's an analogy; it has to be pertinent- and this one just isn't. The modern DSLRs are not the equivalent of teletypes and punched cards, far from it.

I am using my personal experience of computers since I first programmed one in 1965 to extrapolate to other electronic subsystems, especially EVFs. It is common to start by supplementing the mechanical actions with electronic ones, but that often isn't the end. Disc drives have more electronics associated with them than when they were first introduced. But solid state discs do away with the mechanical bits altogether. The same is happening to viewfinders - of course OVFs have been supplemented by electronics over the years. But now EVFs do away with the mechanical bits.

Except that SSDs offers low capacities for high prices, and would wear out with constant writes (which is an issue, unless it's SLC-based). Like EVFs, they can't yet replace their mechanical counterparts. Unlike EVFs, they offer better performance.

I would rather say the SLRs are now computerized beasts, and I would see the viewfinder as a HID. And like in computers, some things are better left mechanical (optical), until a properly better technology will appear.

See also "Why Moore's Law Applies to Photography".

That's more than a little bit forced, there's a long way from those transistors to a broad, generalized statement like "Moore's Law Applies to Photography". I say it doesn't; that would be something like "Images are doubling their size every 18 month", or "people are taking double the number of pictures they did 18 months ago".

[LamyTax]

What release date are you guys betting on? Is there a pot?

[Aristophanes]

It's debatable as to whether more and smaller pixels may be considered to be an improvement anymore.
Especially if our lenses can't keep up.

Canon has already commented about the need to update their L glass to improve fine resolution precisely because of this.

[Aristophanes]

What release date are you guys betting on? Is there a pot?

Early September for a K-5 Mk III with 24 MP, better video, and some other stuff.

I also call a new flash with an outlier that this flash may work with the Ricoh GR.

And 1 new lens. Not sure the FL but it will be a zoom. I bet WA and WR.

[Winder]

Something will be announced in September. No idea what we will see.

Sony has released a new 20MP APS-C, but it is not really impressive.
The Sony sensor in the Olympus OM-D would scale to 26-27MP.
There is a rumored 32MP APS-C.

Toshiba has a very nice 24MP ASP-C sensor that is currently being used by Nikon. It wouldn't surprise me to see Toshiba getting aggressive with its sensor sales.

Whatever sensor they decide I hope we see a much faster & more accurate AF. A larger and brighter VF.