[Kunzite]

Sorry, Barry, but it's you who must explain in which way the Moore's "Law" applies to EVFs, and show us some data to prove it. Your way is called shifting the burden of proof, and it's not acceptable.
You could start by specifying which characteristics are doubling every 2 years, so we'd know what to look for.

[falconeye]

"Moore's Law" (over 12 million hits on Google) is a term that identifies technology and industry trends recognised worldwide over decades.

You just lost any credibility.

Moore's Law is a topic which could fuel the discussions of an entire forum. You are both oversimplifying things.

Of course, there is no such law as Moore's. However, it is a heuristic rule which prooved true over stretched periods of time in the past. It is still somewhat applicable when it comes to feature sizes.

The most obvious breach of the rule is the 3 GHz barrier which processors reach but refuse to cross since 2002 (the Northwood Pentium 4).

It remains to be seen how this may ultimatively bring progress in the field of semiconductors to a halt, or not. The implications for our civilization and its future are huge. Personally, I believe everything will depend on our ability to scale one Exaflop down to fit into shoe box. We'll see ...

As for EVFs ... A high contrast and high resolution EVF has to be a larger chip than the current generation uses to be (they use minuscule chips). Already for its contrast abilities. It is like sensor size: if a chip which is coupled to an optical system is too small then there are problems which are expensive to solve.

So, future high end EVFs will use large chips which will both bring betetr resolution and better contrast. Current technology can already do this. But it won't be cheap and Moore's law won't help. However, progress in OLED technology will.

[IchabodCrane]

"Moore's Law" (over 12 million hits on Google) is a term that identifies technology and industry trends recognised worldwide over decades. If you are denying what is described by the Wikipedia article, the onus is on you to provide the evidence and analysis for your case.

If you are claiming that these established trends in electronics don't apply to electronic viewfinders, again the onus is on you to provide the evidence and analysis! (The year by year improvements in EVFs are compatible with Moore's Law, and obviously haven't stopped).

See "Why Moore's Law Applies to Photography". Perhaps you have a response to that article?

Sorry, we're not worshiping your idol. I realize that comes across as a bit flippant but there is an element of seriousness in what I wrote.

[Barry Pearson]

Moore's Law is a topic which could fuel the discussions of an entire forum. You are both oversimplifying things.

Of course, there is no such law as Moore's. However, it is a heuristic rule which prooved true over stretched periods of time in the past. It is still somewhat applicable when it comes to feature sizes.

The most obvious breach of the rule is the 3 GHz barrier which processors reach but refuse to cross since 2002 (the Northwood Pentium 4).

It remains to be seen how this may ultimatively bring progress in the field of semiconductors to a halt, or not. The implications for our civilization and its future are huge. Personally, I believe everything will depend on our ability to scale one Exaflop down to fit into shoe box. We'll see ...

As for EVFs ... A high contrast and high resolution EVF has to be a larger chip than the current generation uses to be (they use minuscule chips). Already for its contrast abilities. It is like sensor size: if a chip which is coupled to an optical system is too small then there are problems which are expensive to solve.

So, future high end EVFs will use large chips which will both bring betetr resolution and better contrast. Current technology can already do this. But it won't be cheap and Moore's law won't help. However, progress in OLED technology will.

(I apologise in advance for saying some things that are obvious to you).

Of course I'm simplifying things in the space available! (And I said earlier "If you mean "it is an observation, not a "physical law", true"). But my earlier analogies didn't rely on Moore's Law. They were observations from a career in the computer industry of the trends from mechanical to electronic subsystems. My observations are both that it happens, and that once it happens a whole lot of other things happen because of it. (Even when I was helping to design new systems for the future, I was often caught out by advances in the industry where I just thought "I never saw that coming!" It has reached the stage, about 48 years after I first programmed a computer, that instead of being surprised at such advances, I'm sometimes surprised when they don't happen).

Moore's Law gets stated as numbers (for example "double the processing power every 2 years" or some variant of that). But there are important implications behind that. It doesn't happen just because someone says "two years are up, now let's double the speed"! It happens because of (and enabled by) a lot of other advances: fabrication processes and technologies; materials sciences and technologies; ingenuity in exploiting existing technologies in new ways or in by-passing problems that can't currently be solved directly; etc. And some of those are not limited just to raw processing power, but may be transferable to other aspects of the final products.

The penetration of EVFs into different areas of photography depends on the demands of those areas. For example, when I am shooting a plane or bird in flight, or a racing bike or car, or jetski freestyle, perhaps at 7 fps, do I need "better resolution and better contrast"? I think things are happening too fast for me to exploit those. I could probably manage with a silhouette! I assume (without ever having used an EVF for those subjects) that lag will be one problem, and any auto-focus limitations will certainly be a problem. (Just as it is with my K-5IIs at the moment).

But in my studio, a different set of demands applies. I need to be able to judge subtleties of tones and colours and expressions, and it would be useful to have clipping indicators in case the model has moved away from the place around which the lights are arranged. I don't know how much lag would be tolerable, but other EVF limitations such as the "tearing" of the field-sequential LCD used in the new Panasonic GX7 may be acceptable. After all, I have used my K-5IIs tethered in the studio. I was shooting vertically downward with the camera near the ceiling), and (in effect) using a laptop as a sort of electronic viewing aid. Even with those massive delays, I still got some useful shots.

I am aware that the processing performance curves are no longer following Moore's Law in the same way, (they lag behind), and there are many who think that it no longer predicts the future. Then I read speculation that quantum computing will put things back on track! I have no idea who is right, or even if the language they are using is the same. But what I am observing is that EVFs in cameras being announced by major manufacturers are not close to stagnating. I expect to see far more advances in EVFs than in OVFs over the next few years, and that is not because OVFs are so far ahead and don't need advances. Unless someone knows of a fundamental limitation that will forever exclude EVFs from important areas of photography, I'll assume they will gradually take over from OVFs and cause the latter to become niche subsystems. We'll see over the next few years.

[IchabodCrane]

subtleties of tones and colours and expressions

If that is your objective then nothing can be better than an OVF. As stated before, no practical EVF can render the dynamic range that a sensor can record. Will it someday? I don't venture to guess but if one does I doubt it will be linked to the number of transistors in a processor doubling every two years (now believed to be every three years). Also, no EVF can "gain up" to match a brightly sunlit scene which means there will be an additional lag that no processor can ever address -- the lag in our eyes adjusting from sunlight to whatever an EVF can pump out.

What's wrong with instant review blinkies and histograms for finding clipped highlights?

[Pål Jensen]

I didn't know Moore's law was relevant for optics....

[Winder]

Do we need to start a Moore's Law thread?

[Kunzite]

Pal: Of course it is, the speed of light doubles every 2 years, so does the number of elements in lenses

Barry, there are many computerized things which don't listen to Moore's "Law". Your experience applies in a much more limited context than you want us to believe...
I would not want a silhouette when shooting birds, by the way (in addition to the lag). And I'm not sure about judging "subtleties of tones and colours and expressions" on an uncalibrated miniature LCD...

The areas where the EVFs needs most progress are precisely the ones where OVFs don't need any: lag, resolution (assuming high quality FF viewfinder), dynamic range.

[johnmflores]

Moore's Law gets stated as numbers (for example "double the processing power every 2 years" or some variant of that). But there are important implications behind that. It doesn't happen just because someone says "two years are up, now let's double the speed"! It happens because of (and enabled by) a lot of other advances: fabrication processes and technologies; materials sciences and technologies; ingenuity in exploiting existing technologies in new ways or in by-passing problems that can't currently be solved directly; etc. And some of those are not limited just to raw processing power, but may be transferable to other aspects of the final products.

...

I am aware that the processing performance curves are no longer following Moore's Law in the same way, (they lag behind), and there are many who think that it no longer predicts the future. Then I read speculation that quantum computing will put things back on track! I have no idea who is right, or even if the language they are using is the same. But what I am observing is that EVFs in cameras being announced by major manufacturers are not close to stagnating. I expect to see far more advances in EVFs than in OVFs over the next few years, and that is not because OVFs are so far ahead and don't need advances. Unless someone knows of a fundamental limitation that will forever exclude EVFs from important areas of photography, I'll assume they will gradually take over from OVFs and cause the latter to become niche subsystems. We'll see over the next few years.

I'm with Barry on this one. Advances in display technology are being fueled by the integration of data into our everyday lives—smartphones, tablets, wearables, embedded systems, etc... There's a lot of smart people with a lot of money behind them being tasked with making improved displays a competitive advantage. These improvements (performance and price) will trickle down to our little world of photography eventually, and at some point in time most manufacturers will realize that they can improve their profit margins by replacing a very good OVF with an as good/nearly as good/better EVF that costs them less to produce.

OVFs will eventually become a niche feature, much like rangefinder viewing systems are now.

My $.02, non-refundable.

Is the K-03 here yet?

[konraDarnok]

I didn't know Moore's law was relevant for optics....

It's relevant to semi-conductors in general, as they all use similar manufacturing techniques. There's thing called Haitz's law specifically dealing with the light output of LEDs for example.


I don't know if it's the very nature and storied history of photography or what, but it does seem to have a lot of Luddites. EV vs. OV is a lot like the combative debates of film vs. digital from 10 years ago,

The optical viewfinder's days are numbered, there is no doubt in my mind. Both Sony and Olympus seem to have completely abandoned them.

[LamyTax]

It's relevant to semi-conductors in general, as they all use similar manufacturing techniques. There's thing called Haitz's law specifically dealing with the light output of LEDs for example.


I don't know if it's the very nature and storied history of photography or what, but it does seem to have a lot of Luddites. EV vs. OV is a lot like the combative debates of film vs. digital from 10 years ago,

The optical viewfinder's days are numbered, there is no doubt in my mind. Both Sony and Olympus seem to have completely abandoned them.

The days of the OVF are only numbered if you see major adoption of EVF. You currently don't.
It'll eventually get there, but until you see Canon/Nikon put EVF into their high-end and midrange you'll have to wait for the statement to be true .

[mamethot]

I didn't know Moore's law was relevant for optics....

It's the PentaxForums corollary law.

Optics quality doubles every 18 months, Pentax is doomed every next day.

[konraDarnok]

The days of the OVF are only numbered if you see major adoption of EVF. You currently don't.
It'll eventually get there, but until you see Canon/Nikon put EVF into their high-end and midrange you'll have to wait for the statement to be true .

I'd disagree with that assessment. Phones are EVFs. Most cameras are, and have been for quite some time. The SLR is a mature technology, and I think it's tracking the "obsolescence before plateau" phenomenon. SLRs right now are as good as they're ever going to get. Sure, there's going to be some more marginal improvements, but the craze of buying is over. For the vast majority of people, the camera they bought 2 years ago is good enough. More megapixels or more dynamic range is not going to get them into the stores to buy a new one.

Canon and Nikon have the most to lose. So they will be the last to adopt, kinda like how Pentax as the last to abandon screw mounts. Undoubtedly high end SLRs will continue to be made, but then so are phonograph turntables.

[Aristophanes]

So, future high end EVFs will use large chips which will both bring better resolution and better contrast.

And larger chips means larger form factor. And already a Sony EVF on their A-mount is quite large. The external EVF on the RX-1 isn't exactly small either.

So you get into trade-offs both technical and economic. Same as for sensor size.

[Aristophanes]

I'd disagree with that assessment. Phones are EVFs. Most cameras are, and have been for quite some time. The SLR is a mature technology, and I think it's tracking the "obsolescence before plateau" phenomenon. SLRs right now are as good as they're ever going to get. Sure, there's going to be some more marginal improvements, but the craze of buying is over. For the vast majority of people, the camera they bought 2 years ago is good enough. More megapixels or more dynamic range is not going to get them into the stores to buy a new one.

Canon and Nikon have the most to lose. So they will be the last to adopt, kinda like how Pentax as the last to abandon screw mounts. Undoubtedly high end SLRs will continue to be made, but then so are phonograph turntables.

I agree, but that may strengthen the argument to continue the OVF. Familiarity is a very strong economic motivator in personal purchases. This is likely why the uptake on EVF's has been poor.

There are all sorts of areas of economic activity where older tech still holds the edge simply because it does a good enough job at a good enough price with no new learning curve or adaptability.

Eyeglasses are a prime example. I remember when we were ALL going to have contact lenses because they'd only get better and they'd be the only choice anyone would want.

Nope.

Then laser surgery.

Nope.

I do see the DSLR market plateauing and even contracting modestly. I actually see this happening in FF earlier precisely because of the very high prices. I see a fractured market where some people will never give up their OVF at any price point and others will be OK with an EVF at the same price point and many will opt for no VF at all.

The wholesale replacement concept is not what we see in most markets and product categories. Not everything is a buggy whip (which, BTW, are still being made).