[Kunzite]

I'll join.
I said in the past and I'll repeat: perhaps Pentax - a name tied to the history of the SLR camera - is the proper brand to make another evolutionary step. The DSLR evolved, that's what I'd like to see. And Ricoh Imaging has some interesting patents on this subject... but would they use them?
(Dear Ricoh Imaging, if you're reading this: I'd support such a thing with my wallet)

[goldenarrow]

It isn't the components themselves that's expensive. They're nearly without cost. It's the assembly people. People are expensive.

People assemble and adjust the many moving, mechanical parts in SLR's. The manufacturing advantage of MILC is to replace expensive people assembling and adjusting analog components with inexpensive machines producing digital subassemblies, which can be manufactured without human intervention. As the oft-cited Roger Cicala Sony A7r teardown article revealed, a Sony MILC is simple and elegant internally. The sensor mount / shutter / lens plate do not need to be adjusted or calibrated. Read the article.

"The completely disassembled Sony A7R consists of about a dozen major pieces, held together with 29 screws of just three different sizes. A typical DSLR has around 120 screws of 11 different sizes. . . . This is rather amazing. How much easier it will be to fix this camera when it breaks? How much simpler it must be to perform all the calibration that must be done during assembly? And how much simpler it must be to assemble the A7R in the first place. In other words, how much cheaper it must be to make this camera, than to make a DSLR"

It doesn't really matter what traditionalists want. The manufacturers will simply abandon and replace us because they must.

That is a very eye-opening article. I can see manufacturers drooling over this. As an end-user amateur, it does not immediately become apparent why I should care. Until they stop making K-mount lenses!! Then it would become a problem for me for sure. But, like a Zombified movie character, the K-mount lives on even though it is dead....


---------- Post added 07-15-17 at 01:14 PM ----------

I'll join.
I said in the past and I'll repeat: perhaps Pentax - a name tied to the history of the SLR camera - is the proper brand to make another evolutionary step. The DSLR evolved, that's what I'd like to see. And Ricoh Imaging has some interesting patents on this subject... but would they use them?
(Dear Ricoh Imaging, if you're reading this: I'd support such a thing with my wallet)

Dumb question. Is it possible to make a mirrorless Medium format camera that natively uses K-mount lenses??? Or is the sensor so big that the image circle will not work with K-mount image circle???

[Kunzite]

Even the "cropped" 44x33mm medium format sensor used in the 645D/Z is too large for K-mount lenses designed to cover 24x36mm.
There is a possible solution, a rear converter which would expand the image circle.

[goldenarrow]

Even the "cropped" 44x33mm medium format sensor used in the 645D/Z is too large for K-mount lenses designed to cover 24x36mm.
There is a possible solution, a rear converter which would expand the image circle.

Something you said gave me a silly idea for Pentax to think about. Come out with a Full Frame mirrorless camera but make sure it comes with a K-mount adapter with ALL the electronic pins and a chip to make sure K-mount lenses get recognized by post-processing software in the EXIF info. Software is now a major component of the creative (artistic) process and past "adapters" were dumb physical adapters that automatically made the native lens unrecognizable by software that I use (DxO Optics Pro). Why not make a "smart-adapter" that allows the software to see the native lens that was attached (and all of its features) and the benefits are not lost. This allows K-mount lenses to be used effectively and "M-mount" (mirrorless mount lenses) to be produced in parallel without causing too much turmoil amongst the user base. I am sure someone else has mentioned this idea but the "smart-adapter" is key to making it work. If not Full Frame, then do it for APS-C instead.

See, there is a way to make this happen! (if it were THAT easy.....)

[Kunzite]

The mirrorless + adapter solution only works for those wanting a mirrorless. I'm not against the idea, if they continue to develop the K-mount/DSLR line... I wonder if they can develop K-mount and "M-mount" lenses simultaneously, at a reasonable fast pace.
IMO, they'll first have to do some essential work on the D FA lens line.

[Pål Jensen]

"The completely disassembled Sony A7R consists of about a dozen major pieces, held together with 29 screws of just three different sizes. A typical DSLR has around 120 screws of 11 different sizes. . . . This is rather amazing. How much easier it will be to fix this camera when it breaks? How much simpler it must be to perform all the calibration that must be done during assembly? And how much simpler it must be to assemble the A7R in the first place. In other words, how much cheaper it must be to make this camera, than to make a DSLR"

It doesn't really matter what traditionalists want. The manufacturers will simply abandon and replace us because they must.

This proves that this Sony is more elegantly put together than the Nikon. The mirrorless and DSLR have the same form factor and there's no reason why they couldn't be put together in the same way. The difference is the mirror and the AF module. The mirorrless and the DSLR demand exacty the same precision in the distance between lens mount and sensor, the difference is that the DSLR have a mirror placed in the "box" between. my guess is that the added "complexity" amount to about $10 in production cost. Remember a DVD/Blu-ray player cost $15 in production cost (admittedly in China) including the "complex" transport/laser assembly.

[photoptimist]

The Sony MILC only seems simple until you look inside the chips. There's literally over a billion parts inside them. Sure, assembly labor might be "expensive" but so is building and running a multi-billion dollar semiconductor fabrication plant that goes obsolete in 3 to 5 years.

People who obsess over the complexity or simplicity of product have missed the overall trend in product design and development which is toward ever more complexity but at ever lower costs (look under the hood of the average car and see how much it's changed over time or look at all the car's accessories like power windows that used be only found on very expensive luxury cars).

Sure, a simpler camera might be cheaper to make. But if a more complicated camera offers additional features sought by some customers (e.g., an OVF that shows the actual DR of the scene instead of a headache-inducing EVF with clipped highlights and shadows) and at an affordable price, customers especially the high-end ones will gladly pay for that complexity (and they do).

After decades of development, the SLR mechanism is dirt cheap and worth every penny even if it involves some extra parts and extra labor.

---

Last edited by photoptimist; 07-15-2017 at 01:10 PM.

[Kunzite]

It's largely an elegant design, but it's also partially cost cutting. The A7R is one of the Sony cameras with a flimsy mount - there are third party reinforced mounts for it.

[monochrome]

Missing the point. People - the most expensive and least dependable manufacturing process tool - don't have to adjust and calibrate a mirrorless camera. The same undeniable economic force that has driven humans out of auto manufacturing - robots- is driving them out of cameras. Manufacturers will do anything to reduce labor cost.

Volvo (Geely) has announced after 2019 ALL THEIR CARS will be fully electric or hybrid. Not because they're responding to demand or they want to be a green company. Because eliminating / reducing the traditional gasoline engine / transmission assemblies makes vehicle assembly less complex and less dependent on expensive humans.

in 20 years there won't be mass-consumption cars made with traditional gasoline propulsion systems at all, and Geely wants to exercise first-mover advantage. It has nothing to do with energy or regulation or taxes. It has to do with manufacturing process.
-

[Zygonyx]

Basically when reading monochrome, customers whishes or expectations don't count.
Environnemental considerations neither....
Let me just doubt you got what post-industrial era meant....

[Pål Jensen]

You overestimate the cost of calibrating a camera. I'm pretty sure calibration can be done by robots as well.

That car analogy doesn't hold. It is driven by emission cut demands made by governments. A hybrid car is more complex and expensive than a non-hybrid counterpart. Electric cars are still too expensive, but that will change.

[mecrox]

The Sony MILC only seems simple until you look inside the chips. There's literally over a billion parts inside them. Sure, assembly labor might be "expensive" but so is building and running a multi-billion dollar semiconductor fabrication plant that goes obsolete in 3 to 5 years.

People who obsess over the complexity or simplicity of product have missed the overall trend in product design and development which is toward ever more complexity but at ever lower costs (look under the hood of the average car and see how much it's changed over time or look at all the car's accessories like power windows that used be only found on very expensive luxury cars).

Sure, a simpler camera might be cheaper to make. But if a more complicated camera offers additional features sought by some customers (e.g., an OVF that shows the actual DR of the scene instead of a headache-inducing EVF with clipped highlights and shadows) and at an affordable price, customers especially the high-end ones will gladly pay for that complexity (and they do).

After decades of development, the SLR mechanism is dirt cheap and worth every penny even if it involves some extra parts and extra labor.

What if people don't care? I was in a local camera store this afternoon before seeing Baby Driver (great film). They don't sell DSLRs, only mirrorless. I asked why and they told me no one comes in wanting to buy a DSLR these days. Maybe not true in other areas, but wow. It had never occurred to me that a camera store would ever do this.

[Reed]

I worked as service engineer for one of the biggest asian computer companies for few years. Durring the "boom" of notebooks sells, the evolution was not made in the insides, if you seen one, you've seen them all-electronics manufactuere processes etc are almost unchanged. The real change is in design of chassis and methods of making the stuff to hold intact. Till end of 00's notebooks ware almost entirely hold by very precise screwed plates,rigs, and frames,lot of metal parts inside,decorative plastic outside. Second generation was made with "snap-in" interlocking of chassis plastic plates tech and much less screws type used(max 2-3). Last and mostly current generation is just glued together, with few or no screws used(for hinges).Its just stamped plastic(consumer) or metal(premium), hold around the edges with hot glue or adhesive tapes. Is it bad? No. Is it good for repairs etc? Not realy, stuff is made in "throw this to dumpster and then get a new one philosophy.

[RGlasel]

Missing the point. People - the most expensive and least dependable manufacturing process tool - don't have to adjust and calibrate a mirrorless camera. The same undeniable economic force that has driven humans out of auto manufacturing - robots- is driving them out of cameras. Manufacturers will do anything to reduce labor cost.

You are badly off track. You will find plenty of human beings working in automobile factories, what has changed is they aren't doing the work of robots anymore. Human beings don't hold welding rods, they don't tighten fasteners and they don't carry components from one station to another. A human being monitors the movement of parts and checks that the automated steps were successfully completed. Robots are useless for anticipating unscheduled problems. The other major saving in labour costs has come from third party assembly of sub-components, done by just in time suppliers with non-unionized factories. Show me an assembly line where the only human worker is a janitor and I'll show you an assembly line that has never been restarted after a failure that wasn't in the original design. In particular, automobile manufacturers haven't taken complexity out through the use of automation, instead they have been throwing new features into the automotive shopping cart like a stoner in the junk food aisle.

To your original point, DSLRs persist in the marketplace because human beings want to buy DSLRs, not because DSLR manufacturers haven't figured out how to build MILCs for less money. What do we see in the marketplace? MILCs that are as expensive as DSLRs and even less development of new MILC models than DSLRs. At least with APS-C and FF, there are two major DSLR manufacturers, as opposed to a single major player in MILC for those sensor formats. None of this points to the inevitable displacement of DSLRs by MILCs.

And one more thing: If any industry could be 100% automated, it should be the mass production of clothing. Instead, garment manufacturers have moved to lowest cost labour markets instead of investing in robotics.

[Rondec]

To me, I think the question the camera companies are asking is clearly not "What is the next evolution of the camera?" or even "How can we get our users to make better images?" The question they are asking (this includes Ricoh) is "How can we make more money?" or at the least continue an adequate revenue stream. If mirrorless cameras are cheaper to build than SLRs and can command similar prices, then the writing is on the wall for the SLR cameras. If the time comes that SLRs are all priced fifty percent or even twenty percent more than mirrorless cameras with similar tech, than who would buy an SLR unless they just can't live with an EVF?

That said, people keep predicting this event and the reality has been that SLRs and mirrorless ILCs are priced almost exactly the same for similar specifications.