OVF: Solution looking for problem

[Rob MacLennan]

Try performing manual focus with a EVF, especially in low light, and you'll answer all of your questions as to why a OVF is generally considered superior.
--
D620L -> D540 -> C750UZ -> E-500 -> E-510 -> E-3

 

[BorisK1]

Right, I know exactly what you're talking about. It's easy to see, too - when I rock my P&S side-to-side, the horizon on LCD rocks with a visible delay. Or if I wave a hand in front of it, the hand on the screen is half-a-wave behind.

I'd be surprised if the processing delay would be noticeable though. There should be minimal processing (no need to do things like demosaicing, for example). It's a memory-to-memory transfer, perhaps with a couple of arithmetic operations on each pixel to account for brightness and WB.

It's a fraction of what's needed to fully process 10M jpegs at a continuous rate of 3fps (demosaicing, sharpening, correcting for the lens geometry and vignetting, etc, etc, and then compressing the results and sending it to a flash card).

Perhaps if they use averaging to combat noise, things will get slower.

There's viewing lag and there's shutter lag.

With an SLR, viewing lag is at the speed of light. Shutter lag is
the time it takes at the time the shutter is actuated for the mirror
to swing up and the shutter to open.

With an EVF, viewing lag is at the speed at which the sensor
processes the image and sends it to the EVF. Shutter lag is reduced
because the mirror no longer exists.

OLED displays are capable of 0.01ms, so smearing and other artifacts
seen in older LCDs is not evident. They are also brighter and have
higher contrast. The display is no longer part of the time factor,
but the sensor to display pathway can still slow down.

This is especially evident in low light, when the image requires more
processing to be usable (hence reports that at low light the EVF on
the m4/3 prototype slows down and gets jerky... this is not because
of the EVF, per se, but rather the low light processing engine).

 

[PBasepeterb]

Try performing manual focus with a EVF, especially in low light, and
you'll answer all of your questions as to why a OVF is generally
considered superior.
--

Of course using an EVF in near total
dark makes a shot quite possible.

You can't "gain up" an OVF.

OVFs are only "generally considered superior"
by those who generally consider them superior.

--
Keep your lens clean and your mind open.

http://www.pbase.com/peterb/

 

[BorisK1]

For that situation prefocus is the norm, right? So the problem lies
in the fact that OVF provides you with information at the speed of
light while the EVF is slowed down by the sensor to EVF pathway,
which involves not just the physical pathway through circuits (which
could conceivable be virtually the speed of light) but the processing
pathway for translating the information read by the sensor into an
image on the EVF. The EVF's response time to this information is not
an issue any more (even LCDs can get down to 3-5ms, but the newer
OLED technology goes below 1ms), but the information reaching that
EVF is the bottleneck.

Some cameras address that by constantly taking a stream of images and keeping the last few in memory. When you press a shutter button, the camera saves the image that that it shot earlier - just earlier enough to compensate for the display lag. So what you saw on the screen is what gets saved.

 

[BorisK1]

Try performing manual focus with a EVF, especially in low light, and
you'll answer all of your questions as to why a OVF is generally
considered superior.

You forgot to add "without using the 10x zoom, because that's cheating"

 

[Robert999]

I don't understand point 4. Why would the viewfinder on an EVIL be
blacked out?
Maybe I just don't understand the situation you are describing here.

I think the poster above was referring to having the EVF also used as the post-shot review monitor, in which case it would be busy reviewing the last shot taken, and could not be used immediately to frame the next shot. This is what would happen if you only had one LCD panel on the camera, and it was used for framing and review.

Of course, the better way is to have an EVF with an eyecup for framing, which NEVER reviews, and then a main LCD screen on the back of the camera which has an adjustable review function, just like any DSLR. So if you want to frame continuously, just use the EVF with a single eye, or if you want to chimp, well then pull back and look at the larger LCD screen.

--
Robert
Olympus E-3 & E-500, 7-14mm, 14-54mm, 40-150mm, 30mm
Olympus 1030SW P&S
Gandolfi 'Universal' Large Format, 127mm Schneider
Yashica Lynx-1000

 

[olyflyer]

Try performing manual focus with a EVF, especially in low light, and
you'll answer all of your questions as to why a OVF is generally
considered superior.

I just did with the SP570. It is really a pain in the back part of my body to use that camera in manual mode. Even composing an image in low light is a PITA, as composition is made in the dark and then during AF the image becomes brighter and after that dim again. OK, I used AF, but I can't see how MF would make the focusing better. By far I prefer the dim VF of the E-500. Don't even mention a good VF like the E-3.

Maybe I did something wrong, but anyway, the OVF is by far superior an EVF in my opinion.
--
http://www.olyflyer.blogspot.com/

 

[PerL]

Its in principle the same thing. A real view is much more satisfying. I hate EVF:s.

I've often wondered what the point of SLR is in a digital world. The
whole mirror system was, I assume, originaly intended to allow you to
see through the lens so that you will see what will end up hitting
the film or whatever image surface. It was a neat solution to a
particular problem. It allows you to see what the film sees.

Since digital can both show you the view through the lens and capture
the image, there's no point having the old mirror system. In a
digital world, it's a solution to a problem that no longer exists.
Sure, it's nice to have the optical path from the eye to the lens and
on to the subject, but it's not necessary. It doesn't 'solve'
anything. To me it's more of a hang-up or habit than a necessity.

I bought a dSLR simply because I wanted interchangable lenses. I
thought having a mirror system was dumb, but I had to put up with it

• no choice.

I thought an electronic viewfinder was an obvious thing. Especially
handy if you can tilt it as you can use interesting camera angles.
The only reason why live view has been awkward to impliment is that
the bloody mirror is in the way. To say that live view was a solution
looking for a problem was, to me, absurd. I was thinking the exact
opposite: that the mirror box was now a solution looking for a
problem. And, anyone who doesn't appreciate live view doesn't
understand real world photography, where interesting angles open up
great possibilities. (Above a crowd, held above the surface of a
river, etc...)

So, to me, the non-mirror EVIL system makes sense.

And the whole optical view finder and box of mirrors set-up is now
the solution that is looking for a problem.

And I hope we'll see some interesting video gear with interchangable
lenses. I hope Sony are thinking about that.

--
http://bonusphotography.wordpress.com/

 

[deep7]

It did away with mirrors etc. and seemed to be the answer to my dreams. Then I tried one and the electronic viewfinder killed it for me. All I could see was a grid of dots! Really distracting, which was a shame because that camera had a lot of strengths.

One thing it missed was small size. The lens was not that fast and it had no mirror box, yet a 12-60 on an E420 would, I think, be smaller. I'll wait to see how this new, smaller standard actually pans out.

I also wonder what will happen to the "tele-centric" lens design with the new, short-reach lenses but that's a topic for another thread!

Meanwhile, there is one real advantage of doing away with the mirror - it would allow a near silent camera for weddings. I used to use a Leica and the sound advantage was very significant.
--
Don.

A Land Rover, a camera ... I'm happy!

 

[strawman_John]

The flipping mirror is a solution to a problem that no longer exists.

No the problem exists, its just the optical system offers a different solution to the EVF.

The mirror system was intended to use a single source of light for
two distinct purposes: 1) to expose an image surface and, 2) to
display the image to the user.

An electronic sensor can both capture the image, and display it to
the user. And it can do these things pretty much simultaneously. So
the problem which the mirror was invented to solve no longer exists.
That is the key point I was trying to make.

The original reason was it was the only way to get the actual image viewed by the lens that is correct, but things have moved on. The original reason for the EVF was because they could not get a decent optical viewfinder matched with the small sensors. With a dSLR the sensor got bigger. Should we kill the EVF as its original purpose has gone?

With the modern AF system on an SLR, not only can the sensor make an assessment of whether the lens is focused or not, it can often also calculate how far out of focus the lens is. This gives a big advantage in the speed of focus over contrast focus. also contrast focus has to read the picture, i.e photo decode whole portions of the sensor (take a picture, move take a picture) etc suddenly you can start to see why SLR focusing is better than video type techniques. Plus you need to stop image recording to take a picture, its, you guess, another delay.

Lag, resolution and dynamic range can be addressed with increased
buffer size and improved sensor and display quality. But let’s not
lose sight of the fact that the viewfinder is there to help you frame
the shot.

Sorry, no lag cannot be addressed by a bigger buffer, in fact a bigger buffer will make that worse
.

I think that to some extent we all conceive and even compose shots
without the camera. We look at the scene with our naked eyes,
consider composition, light, shade, etc., and then just use the
camera to frame and capture the image. There may be exceptions such
as long-lens sniping and birding. But do many of you really stick
your eye on the optical view finder and wave the camera about trying
to compose a shot? Or sit there waiting for something to happen? I
use the viewfinder to frame the shot. But most of the work – whether
it be composing or waiting for a magic moment – is done with the
naked eye.

It depends, if its people or sport or wildlife, i.e. the decisive moment bit, I use the viewfinder. I want to see what the camera see's, and most importantly I want it to take a photo when I decide it should. For landscapes or still life I frequently use live view and manual focus. also with focal lengths removed from @ 35mm on a crop dSLR I use the viewfinder to comprehend and compose for the different focal length.

So, really – what problem is there that an optical viewfinder solves
that cannot be solved by some other means? If we wanted to design a
camera from scratch, would we have an optical viewfinder? If we had
to design an autofocus system from scratch, would a box full of
flipping mirror be the best solution?

It solves the speed of focusing issue. and at the moment it also provides an intimacy that EVF's lack. Perhaps the latter can be fixed, the first item, I am not certain. today the EVF equipped camera is too slow and un-involving.

And for the record I use a dSLR, its optical and its live view system for composing etc, plus I also use a film manual everything rangefinder. I choose my mode to suit my subject. Killing the optical viewfinder would hinder me, sometimes the EVF would be fine.

How does the range finder solve the problem, by allowing you to pre-set the camera, focusing distance etc all set, all it needs is me to press the button. It's perhaps the way of the past, but it works.

John-

 

[m43]

SLR does not offer a 'real view', far from it: you look through a series of optical contraptions, light is reflected from a mirror, bend in a pentaprism, interrupted with delay by a mechanical and noisy mirror up / down mechanism, you see a smaller picture framed by black margins typically with series of numbers (various parameters), and you hold SLR close to your eyes (assuming you look through the pentaprism). Compare this with a tilting, large LCD hold at waist level and observing with your eyes (both open) real world and a picture on LCD panel. Some prefer SLR, some prefer other methods, but none of them equates to opening a window and looking at real world with your own eyes.

I've often wondered what the point of SLR is in a digital world. The
whole mirror system was, I assume, originaly intended to allow you to
see through the lens so that you will see what will end up hitting
the film or whatever image surface. It was a neat solution to a
particular problem. It allows you to see what the film sees.

Since digital can both show you the view through the lens and capture
the image, there's no point having the old mirror system. In a
digital world, it's a solution to a problem that no longer exists.
Sure, it's nice to have the optical path from the eye to the lens and
on to the subject, but it's not necessary. It doesn't 'solve'
anything. To me it's more of a hang-up or habit than a necessity.

I bought a dSLR simply because I wanted interchangable lenses. I
thought having a mirror system was dumb, but I had to put up with it

• no choice.

I thought an electronic viewfinder was an obvious thing. Especially
handy if you can tilt it as you can use interesting camera angles.
The only reason why live view has been awkward to impliment is that
the bloody mirror is in the way. To say that live view was a solution
looking for a problem was, to me, absurd. I was thinking the exact
opposite: that the mirror box was now a solution looking for a
problem. And, anyone who doesn't appreciate live view doesn't
understand real world photography, where interesting angles open up
great possibilities. (Above a crowd, held above the surface of a
river, etc...)

So, to me, the non-mirror EVIL system makes sense.

And the whole optical view finder and box of mirrors set-up is now
the solution that is looking for a problem.

And I hope we'll see some interesting video gear with interchangable
lenses. I hope Sony are thinking about that.

--
http://bonusphotography.wordpress.com/

 

[echolsphoto]

or if you want to chimp, well then pull back and look at

the larger LCD screen.

Chimp ???

 

[kevin wilkins]

Try performing manual focus with a EVF, especially in low light, and
you'll answer all of your questions as to why a OVF is generally
considered superior.

i know this all to well, manual focus on moving subjects (in my case a band) in a live performance and the OVF wins hands down, your eyes adjust to light faster, can see in darker light and if the OVF is set up correctly the focus will be correct if its in focus through the OVF.

ive not used an EVF in these conditions for manual focusing but just trying to compose a picture when the camera had to keep up with the changing light conditions by altering the viewfinder image must be a PITA.

--
if you can imagine the picture, then do all you can to make it

 

[fldspringer]

SLR does not offer a 'real view', far from it: you look through a
series of optical contraptions, light is reflected from a mirror,
bend in a pentaprism, interrupted with delay by a mechanical and
noisy mirror up / down mechanism, you see a smaller picture framed by
black margins typically with series of numbers (various parameters),
and you hold SLR close to your eyes (assuming you look through the
pentaprism). Compare this with a tilting, large LCD hold at waist
level and observing with your eyes (both open) real world and a
picture on LCD panel. Some prefer SLR, some prefer other methods, but
none of them equates to opening a window and looking at real world
with your own eyes.

--
The speed of light is 3x10^8 m/s in air.

The speed in glass is about 70% of that figure.

If you calculate the extra distance the light must travel because of reflected paths and the slowing down caused by passing through glass, the total time (measured in seconds) is preceeded by so many zeros it is absolutely no consequence with our ability to detect it.

The scene in the optical viewfinder is as it actually happens. It is in real time.

-
Greg

http://www.spanielsport.com/

 

[Stujoe]

For me, that is what a non-mirror camera means. Until they cure those 2 things, I will stick with the old 'outdated' mirror box.

--
Stu (N80, N50, D50, E510, TZ4)
http://www.flickr.com/photos/stujoe/

.

 

[m43]

Yes, obviously SLR shows the picture in real time, no argument here. But for a split second, when a picture actually happen, it shows nothing. In SLRs this is technology limited, and probably already reached the limits of mirror speed, reliability and quietness. As opposed to electronic based solutions which can be improved (and continuously are).

When using any camera you never look at reality, but at a representation of it, it is never 'real view'. Some things are compromised, some a little, some a lot, some lost completely. No camera offers experience equivalent to opening a window and looking outside with your own eyes. Which camera design is 'closer' to real view is disputable, it depends on your style of work and your preferences. Some of the best ever photos were taken using view cameras where image is composed upside down - how is that for 'real view'.

SLR does not offer a 'real view', far from it: you look through a
series of optical contraptions, light is reflected from a mirror,
bend in a pentaprism, interrupted with delay by a mechanical and
noisy mirror up / down mechanism, you see a smaller picture framed by
black margins typically with series of numbers (various parameters),
and you hold SLR close to your eyes (assuming you look through the
pentaprism). Compare this with a tilting, large LCD hold at waist
level and observing with your eyes (both open) real world and a
picture on LCD panel. Some prefer SLR, some prefer other methods, but
none of them equates to opening a window and looking at real world
with your own eyes.

--
The speed of light is 3x10^8 m/s in air.

The speed in glass is about 70% of that figure.

If you calculate the extra distance the light must travel because of
reflected paths and the slowing down caused by passing through glass,
the total time (measured in seconds) is preceeded by so many zeros it
is absolutely no consequence with our ability to detect it.

The scene in the optical viewfinder is as it actually happens. It is
in real time.

-
Greg

http://www.spanielsport.com/

 

[Lee Jay]

Since digital can both show you the view through the lens and capture
the image, there's no point having the old mirror system. In a
digital world, it's a solution to a problem that no longer exists.
Sure, it's nice to have the optical path from the eye to the lens and
on to the subject, but it's not necessary. It doesn't 'solve'
anything.

It solves the following:

View is instant - no lag.
It uses no power.
It doesn't heat the sensor because it leaves the sensor off most of the time.
Resolution is limited by your eyes, not the device (VF magnifiers work too).
Allows a separate optical path for phase-detection auto-focus.
Works in very, very low light (much lower than an EVF with tons of gain).

If none of those mean anything to you then fine, but don't claim they don't exist.

--
Lee Jay
(see profile for equipment)

 

[Lee Jay]

When tracking a moving target, especially for a fast-moving target in
good light, the phase-detection system is actually at a disadvantage.
Whenever the image of the car falls between the AF sensors, the
camera loses its target lock and has to reacquire it. CD-AF system,
on the other hand, can use every pixel of the main sensor - there are
no coverage holes.

Now, initially acquiring the lock can be faster for the
phase-detection system.

It seems you don't understand the advantages of the PD system. I'll explain.

The CD system needs to "hunt", moving focus back and forth to figure out which way to go, and then how far to go in that direction. Dozens of samples have to be taken before focus can be achieved.

The PD system can take a single sample, and know how much, and in which direction precise focus will be located. Only one sample is needed and, bam, you have it. That makes it inherently faster, especially while tracking. It can keep track of these samples and determine a velocity and acceleration of the object, and use that information to predict where focus must be during the exposure. That means the image can be in-focus at the time you hit the shutter button, the object can move out-of-focus during the shutter lag, and the system can place the lens at the right point so that the object will be back in-focus at time of exposure. This allows tracking of very high-speed objects with very shallow depth-of-field, and it will still produce many in-focus image like this one, which was covering the depth-of-field every 40ms or so:



--
Lee Jay
(see profile for equipment)

 

[JeffreyW]

I can add:

I am red/green colorblind, and the optical viewfinder is MUCH easier for me to see. My former EVF (FZ50) was "muddy" to me, and it's one of the reasons I moved up to an SLR. I can't describe it exactly, but perhaps it's the better contrast that allows me to see more clearly via OVF.

 

[glitched]

The best solution would to be have a hybrid OVF/EVF camera. Don't ask me if it's possible or not, but maybe there would be a way to push the mirror out of the way if you didn't want to use it. Or maybe it could operate in OVF when you're doing nothing to save power, automatically changing to EVF when you half-click so you can see exactly what you'll get.