Canon EOS-A1 with Hybrid EVF? [CR1]

[jrista]

all those woul would "love to get a hybrid viewfinder" .. .could you elaborate in a little more detail, how exactly you would like to have that implemented?

e.g.
* optical rangefinder with EVF overlay (Fuji X-Pro1 style) with or without optical parallax correction .. if yes, to what focal lengths and focus distances?

or
* fixed, semitransparent pellicle mirror (Sony SLT, Canon EOS RT) style?

or ... what else?

Willing to accept and swallow all the "unwanted effects"?
 
Let's face it, time is ripe for EVF. All that's needed is just one more little push ... a higher res [why not 4k?]  OLED display with no discernible lag [lets say less than 3ms - or any other number you consider good enough] and no blackouts between shot. POf course combined with 100% mechanics free, vibration-free global shutter (electronic).
 

Anybody unwilling to accept EVFs will be relegated to using OVFs as long as those will be available in old-style "niche" DSLRs.  Similar to analogue film cameras.

An EVF needs more than 4k to be good enough. An EVF with a viewfinder size the same as a 5D III and the maximum eye relief possible in a mirrorless body would need 12,000 PPI (I did the calculations for that here on CR a number of months ago) in order for the pixels to be invisible to the naked eye for viewers with the highest visual acuity (which is around 20/10, according to my ophthalmologist, who has given me 20/10 eyesights with contacts.) That results in pixels that are so small, they filter out some red light wavelengths at the longer end of the spectrum.

The best EVF's today have around 2600 PPI, so we are a LONG way off from achieving the necessary pixel pitch required to render pixels invisible. That, to me, is the tipping point. When I look through an EVF and see image, rather than a matrix of pixels, then I'll finally be willing to make the move. But making functioning pixels that are only 700nm in size? I dunno, reading about Sony's 2600 PPI EVF and the technological advancements behind it, it sounded like just that was a pretty amazing feat.

So, I'll happily take some kind of hybrid OVF/EVF. In absolutely NO way would I consider a rangefinder style viewfinder with parallax error to be acceptable. I expect such a hybrid to offer me the exact same advantage I have now with an OVF, so TTL for sure. I see no reason why Canon couldn't develop some kind of Color Transmissive LCD, similar to their current two-bit Transmissive LCD, that would overlay the pentaprism. In EVF mode, the mirror would lock up, providing a black backdrop, allowing the full benefit of a color T-LCD that should look/work just like any normal LCD. No reason it couldn't be OLED from an actual pixel technology standpoint, Samsung has already demonstrated transmissive OLED displays. Whatever the technology, though, that's what I would expect. True OVF when the mirror is not locked up, and a full color EVF when it is.

[AvTvM]

Sony TruFinder EVF used in A99, NEX-7 and A7/R to my knowlegde is 0.5" OLED  @ 1024x768 (XGA) RGB.
If I made no mistake using Pythagoras, that calculates as 2560 ppi  [if c=0.5" and a:b = 4:3 then -> a=0.4", b= 0.3" -> 1024/0.4=2560 or 768/0.3=2560] 
Viewed from a distance of about 1 inch (if at all).
Should be plenty resolution for any human eye, no matter what eyesight, or not?

[jrista]

Sony TruFinder EVF used in A99, NEX-7 and A7/R to my knowlegde is 0.5" OLED  @ 1024x768 (XGA) RGB.
If I made no mistake using Pythagoras, that calculates as 2560 ppi  [if c=0.5" and a:b = 4:3 then -> a=0.4", b= 0.3" -> 1024/0.4=2560 or 768/0.3=2560] 
Viewed from a distance of about 1 inch (if at all).
Should be plenty resolution for any human eye, no matter what eyesight, or not?

Visual acuity of a person with 20/10 vision is 1/87th of a degree (less than 1 arcminute). There is no way that 2560ppi at less than 1" is enough. Not even remotely close. I derived the formula and did the calculations here on the forum once before, I'll see if I can find that post. For 20/20 vision (1/60th of a degree, or 1 arcminute) at just under half an inch would need at least 5000ppi to barely reach that point where pixels are just on the edge of not visible. I figure 5700ppi would result in an EVF screen that, at less than 1" eye relief, would just look like an image, rather than a bunch of pixels that make up an image. For me, though, 5700ppi isn't enough, I can easily see the pixels at 0.5" (and I know there are some mirrorless cameras where the viewfinder eye relief is less than an inch, maybe even as little as just under a half to 1/3 inch for some models), which is where 12,000ppi for 20/10 vision comes into play.) With 20/10 vision (contacts or my glasses in) I can easily see iPhone Retina display pixels at several inches (and I can see my Lumia 920 pixels as well, which have a slightly smaller pitch than even the iPhone Retina).

Since more and more often people with corrective lenses have better vision (unless their uncorrected vision is particularly bad), achieving the necessary resolution for 20/10 vision is only fair and appropriate. I think the basic formula I derived for required PPI at any given distance is:

1/(P * D)

Where P is a constant (for 20/10, I think it is 0.0002, 20/20 is 0.00022 I believe), and D is the distance. All units in inches. So at 1" eye relief, you would need 1px/(0.0002" * 1") = 5000ppi. At just under half an inch, you would need 1px/(0.0002" * 0.4") = 12,500ppi.

[mkabi]

What if in the future, they produce that prototype camera that is able to shoot 24, 30, 60 and/or 120fps, and its 100MP with Native ISOs of 1600?

That camera would require sensor cells about 16times of the current ones. At the same time you have about 4 times the cells. I.e. each side is sqrt(16)*sqrt(4)=8 times as long, we're talking about a sensor the size of a large format film/plates and lenses large enough to cover it. To have a real light collecting advantage said lenses would have to be faster then the current ones, while DOF gets thin as a razorblade. They used f/64 for a reason.  There might be practical problems with that.
NB: The QE of current silicon doesn't allow for much higher sensitivity as we're already close to counting single photons. Unless we're getting more photons we're de facto stuck.

So you're saying its a pipe dream? Like the flying car...
But what about that guy that figured out how to slow it down to the point of taking pictures of photons... was it 1,000,000 fps?

You do realize that "taking pictures of photons" isn't really what the femtocamera does, right? It actually takes millions of frames through a microscopic slit, basically time-slicing the movement of millions of separate pulses of light. It can't actually record the motion of light waves in a single  "frame"...it has to take hundreds of millions to billions of frames, which are then processed via very specialized software, in order to make an actual video clip. The device is rather large, bulky, power hungry, and is unlikely to ever look anything like a normal camera. I suspect it will eventually be shrunk in size, and probably packaged into scientific grade devices where such high speed "photography" could be immensely useful for scientific purposes. But there isn't going to be a million fps digital camera capable of actually taking anything that even remotely resembles a "picture" one millionth of a second long on the market any time soon... 

What would I do with a million fps camera? Nothing.
But I'm talking about the possibility that some day there will be a camera both stills and motion based that we won't care that it does both.

The needs of motion based photography are different than the needs of stills based photography. You can always have a better stills camera if you focus on making a better stills camera. Just because you can capture a thousand frames per second at high resolution doesn't mean that any one of those "frames" will be acceptable as a still photograph.

Sure, we can have both, but you can have better of each if you dedicate research to both independently.

I beg to differ...
Have you heard of the saying that it takes 10,000 pictures to get 1 good shot?

I'm sure that they are independently researched, but they come together so well...

How many of us complain that our cell phones can play music, surf the net and host a bunch of applications?
How many of us complain that a phone should be a phone and all the others should be separate?

Why is it that you hate video so much? Do you not watch any movies, home made and otherwise?
People here in this forum say that if they want video, they would get a camcorder or a video camera....
So if you go on vacation, you would rather carry 2 different bodies and shoot stills with one and video with another?

Also, how many camcorders allow us to change lenses? As I said before, video is in every way like stills... I want wide angle, medium and close ups, even macro shots. I want shallow depth of field, creamy bokeh in my video. Can a regular camcorder provide that for me?

They say that the AA filter is giving them a softer image, so give me a shot that you are proud of... and tell me why the AA filter got in the way of your shot.

Well, first off, I disagree that it takes 10,000 photos to get 1 good one. I would say the ratio is more along the lines of 100 to 1 for an average photographer, and considerably less than that for a skilled photographer. That argument generally implies that the photographer doesn't matter, and that if you capture enough frames, a good photo will happen by chance. Possibly, but that is most certainly NOT what having a good stills camera is all about.

The opposite, actually. A good stills photography camera is all about making it easier to get a good shot EVERY time, not once every ten thousand shots! I could probably rest my argument on that alone...but...

Cell phones are not artists tools (not by design.) By design, a cell phone, or as they are more appropriately called these days smartphones and phablets, are multipurpose devices. Of course we don't complain that they can do a lot of things...that is what they are designed to do. They aren't meant to be the best still photography camera money can buy, nor will they ever be. They are meant to be general purpose devices that allow you to augment their functionality in a wide variety of ways via apps so you can COMPUTE ON THE RUN. Sorry, but I don't see any analogy between a smartphone and a camera. One is designed for computing, the other is designed for art.

Regarding me "hating" video, don't put words in my mouth! I really dislike that! I have in no way said I hate video! On the contrary, I would love to make videos of the nature I photograph. It is just impractical to try and do both stills photography and cinematography at the same time, and do both well. As I said before, the two have different core goals, and therefor need equipment tuned for different purposes. I mean, I CAN take photographs with my Lumia 920, and they are pretty nice...but they are always lacking, even when I put in all the effort I have into making them as good as possible. Sure, my Lumia 920 can take video as well, but movies aren't about video...movies are about storytelling, and using the facets of optics to portray that story in a beautiful way. I WANT to do video...but to achieve the kind of storytelling through motion picture that I want, I need to dedicate time and equipment to it. To that end, if/when I get into it, I'll buy a camera actually designed for cinematography.

As for vacation...I wouldn't bring along high end equipment. I'd bring along something like an EOS-M instead. Vacations aren't professional photography or cinematography outings, they are vacations...time to spend with the family, sightsee but not necessarily spend hours photographing each sight, etc. I could take family videos and decent photography with that one tiny little camera, and still have the option to change lenses for different needs. But the needs of a vacation are radically different from the needs I have for my photography, or the needs I would have if I truly wanted to put together a quality movie of the natural world I photograph. I would indeed lug along both separate stills and cinema cameras if I intended to do both, however I'd keep such trips separate from "vacations."

To be quite frank, I feel, as a photographer, that my explicit needs HAVE been glossed over by Canon in some ways. I feel as though if they dedicated some more effort to making a better stills camera, they would, and it would be on the market faster. The simple fact of the matter is I KNOW they are sharing resources, developing some stills technology but even more cinematography technology, and that as a stills photographer, I'm not getting as much from Canon as I used to. Your lofty vision of video cameras that can be ideal for both stills photography and cinematography, assuming it is possible, is a VERY LONG, LONG way off. Decade or two at least. Sorry. but I'm not willing to wait that long for the mere possibility that something like a 10,000fps "camera" that can produce video or stills, might eventually appear.

Sorry, but there could be immediate value to Canon investing a few more resources into stills specific photography research, the results of which could be seen soon, rather than decades down the road. I don't hate video, but I would rather it not be the primary focus of research for stills photography devices. I would rather we not have this endless march towards hybrid devices. I think Canon could produce both better stills cameras AND better cinema cameras by dedicating research teams to both, keeping the two separate. The Cinema EOS line, while not a bad start, still lags far behind the competition, just like their stills cameras are. Companies like Red are already pushing 20 stops of dynamic range...Cinema EOS is still stuck at less than 12. They have to deal with extra crap to blend the two technologies together, and I think that is sapping energy from both as individual technologies.

Lets agree to disagree, because I don't want to take every line and create a counter argument to it.