[audiobomber]

Good post, Greg. It looks to me like the cost difference if you compare bodies with similar features is a myth. The cost saving on MILC only plays if you leave out the EVF.

[IchabodCrane]

One problem with going with only contrast detection is that it can't decide if it's front focused or back focused, it only knows if it's focused or not by moving back and forth then deciding on the peak. This is brutal for any kind of Z-axis tracking and need some complementing AF system to work that out.

Exactly... and probably why there's such a thing as on-sensor PDAF now. It's an infantile technology (though Canikon have started using it) and currently inferior to DSLR PDAF but the key word is "infantile". You have to think it's going to get better and of course this allows focus tracking to become possible on MILCs.

[IchabodCrane]

I wonder how true, in practical terms, the cost advantages of MILCs really are. In any case, we have not actually seen the cost advantages as yet. Compare like for like, and the MILCs tend to be more expensive. The Olympus E-M5 is $999. The Pentax K-30 is +$200 less expensive. Is the E-M5 really +$200 better than the K-30? The E-M5 has a tiltable touchscreen display and slightly better IS. The K-30 handles better, enjoys a better thought-out interface, has more dynamic range, does a better job of balancing larger glass and more powerful flashes, produces slightly cleaner, less noisy images, and is attached to a more mature, richer system. If you're an economy shopper, the K-30 will give you more bang for your buck.

Even if, further down the road, the theoretical cost advantages of MILCs kick in, how significant will that really be? The most expensive component in a camera is the sensor, and an APS-C sensor used in a DSLR will be just as expensive as an APS-C sensor used in an MILC. So where are the cost advantages? Is a mirror box really that much more expensive than an EVF? In 2003 I bought a brand new SLR camera for $140. A quality accessory EVF can easily run several hundred dollars. Is an EVF really signifcantly less expensive to make than a mirror box? Will EVFs be significantly less expensive in the future?

Northcoastgreg, if it was just the EVF vs mirror box and motor tradeoff, I'd agree. But it's also the pentaprism and dedicated PDAF hardware that is needed on DSLRs and not MilCs.

One thing that will keep DSLRs around and maybe not mentioned yet is the greater battery longevity. Certainly that's going to be of value to some subset of users (mostly professionals, presumably).

[VisualDarkness]

Exactly... and probably why there's such a thing as on-sensor PDAF now. It's an infantile technology (though Canikon have started using it) and currently inferior to DSLR PDAF but the key word is "infantile". You have to think it's going to get better and of course this allows focus tracking to become possible on MILCs.

Yeah, on-sensor PDAF solves lot, even though it needs some developing still, most cameras with it aren't using it well. I just wanted to point out why contrast detection haven't taken off as a pro tool yet.

[jon404]

What I took from the interview was that they would like to make a full-frame mirrorless -- and yes, they value viewfinders -- but that they hadn't decided yet what to do. New owner, some normal product planning & integration confusion. Actually, it sounded good -- Mr. Kitazawa is obviously a photographer himself, not an empty marketing suit.

As for size. I can speak to that! Have found that there is a HUGE difference between the shirt-pocket pocketability of my little Olympus XZ-1 -- and EVERYTHING else. Once a camera can't fit in my shirt pocket, it's in another world. And -- to me -- there is no difference in that world between a K-01, a K-5, an old film Nikon... or my 645N. None fit in my pocket; they all have to be worn on a strap around my neck or shoulder. If one of these cameras is a pound or two (a kilo) heavier than others, who cares. Weight just doesn't matter, once you're on a strap. In context, they all weigh about the same as a quart of milk, no big deal to carry.

[audiobomber]

What I took from the interview was that they would like to make a full-frame mirrorless -- and yes, they value viewfinders -- but that they hadn't decided yet what to do.

Are we reading the same translation? They've not decided whether to release a FF camera, but they have decided it will not be mirrorless.
"Up until last year, there was a possibility to create a mirror-less full-frame camera. Looking at all the tradeoffs now, I personally think "with mirror" is the priority we are thinking of for full-frame cameras. It is important to create products with new value propositions, however it perhaps won't be accepted in the market if that doesn't meet usability and needs of the customer. I think mirror-less is more appropriate for medium format cameras than full-frame from a mirror-shock reduction standpoint."
https://www.pentaxforums.com/forums/pentax-news-rumors/209035-pentax-plan-201...ml#post2214364

The implication is that we will see APS-C and MF MILC before FF.

[jon404]

Audio -- you're right, I didn't read it correctly. Fuzzily remembered "Up until last year, there was a possibility to create a mirror-less full-frame camera..." but didn't remember the rest of it. Oops.

My Christmas fantasy actually is that Ricoh buys Kodak Imaging and sticks a giant 60 MB sensor in the 645... and my REAL fantasy is that they sell a digital sensor-insert that would work somehow on my 645N! End of daydream. But then, there are eleven days left in Christmas, right?

[Class A]

Compare like for like, and the MILCs tend to be more expensive. The Olympus E-M5 is $999. The Pentax K-30 is +$200 less expensive.

One must not confuse the cost at which something is produced with the price it is sold at.

I believe there is no doubt that MILCS are cheaper to produce than DSLRs (IchabodCrane mentioned a few aspects). A DSLRs' precision mechanics and its calibration requirements (e.g., for 100% viewfinders) are comparatively cost intensive. I reckon soon EVFs will cost next to nothing in comparison.

[goldenarrow]

One must not confuse the cost at which something is produced with the price it is sold at.

I believe there is no doubt that MILCS are cheaper to produce than DSLRs (IchabodCrane mentioned a few aspects). A DSLRs' precision mechanics and its calibration requirements (e.g., for 100% viewfinders) are comparatively cost intensive. I reckon soon EVFs will cost next to nothing in comparison.

I will have to agree with Class A that the trend for costs is going in the direction of MILC development. Materials are going to keep going up, up, up. So is labor, especially Chinese labor. No wonder Pentax moved lens production to Vietnam (all that hand assembly is expensive). I would not be surprised if large global players followed suit, including going into Cambodia, at least for assembly work. Anything that requires precision manufacturing and alignment will be under pressure. The way around that is to build as much as possible electronically/digitally and pump out as many units as possible (digital can compensate for assembly variances). EVF's, sensors, imaging engines, "digital" stabilization (like in the Pentax Optio series), etc.

Now, what might be really cool would be a MILC that could have a standard backplane (a-la personal computer) that could be upgraded over many years. Maybe that muddies the waters too much, since MILC's are supposed to be throw-away devices, or so we have been conditioned. Enter Ricoh, with its modular camera system merged with Pentax APS-C, K-mount MILC (wish I could patent the idea and make a few million$$ for myself!!!)........

[Kunzite]

My Christmas fantasy actually is that Ricoh buys Kodak Imaging and sticks a giant 60 MB sensor in the 645...

It wouldn't help them, since Kodak already sold their sensor division. All they would "gain" is some bankrupt company eating all their money.
However, there are alternatives...

Indeed, the trend is to make cameras consumer products, worsening them and making them cheaper. I can see in few years people would "upgrade" their point&shoot-like MILC to another almost identical, because the Facebook integration would be better or something.

[lytrytyr]

I wonder how true, in practical terms, the cost advantages of MILCs really are.

K01 at USD 300 versus K30 at USD 600?

[lytrytyr]

As for size. I can speak to that! Have found that there is a HUGE difference between the shirt-pocket pocketability of my little Olympus XZ-1 -- and EVERYTHING else. Once a camera can't fit in my shirt pocket, it's in another world. And -- to me -- there is no difference in that world between a K-01, a K-5, an old film Nikon... or my 645N. None fit in my pocket; they all have to be worn on a strap around my neck or shoulder.

The intermediate size is what fits in a bridge camera waist-pack,
like the K-x or K01, with a lens up to around 89mm in length.

[audiobomber]

K01 at USD 300 versus K30 at USD 600?

You would need to add a hot shoe VF to make the K-01 comparable. The Sony EVF for NEX-5N, the most comparable body, is $300.

[lytrytyr]

You would need to add a hot shoe VF to make the K-01 comparable.

Most users manage perfectly well without.
That's where the price advantage of a MILC kicks in.

I've had P&S cameras with optical viewfinders,
but only used them for about one shot in a hundred.

[IchabodCrane]

Most users manage perfectly well without.
That's where the price advantage of a MILC kicks in.

I've had P&S cameras with optical viewfinders,
but only used them for about one shot in a hundred.

And there are loads of people here who have stated they have low interest in a K-01 because it doesn't have a viewfinder. Also, comparing an OVF on a P&S camera to a modern EVF (like on a NEX-6 or 7) is like comparing a Yugo to a Lexus.