Yet another d3 rumor

Tiny EVFs aren't a very good idea, but large live LCDs are great. This doesn't seem to be outside the realm of possibility.
 
This is one of those ideas that sounds marginally cool at first blush, but when you think about it, makes so little sense for so many reasons.

I have nothing against high-speed crop (emphasis being high-speed), but there's so little value to gain from making it "continuous."
 
Well, I'm no fan of EVFs (I doubt many DSLR users are), but I don't have anything against live view. The thing is, most of the arguments against EVFs hold for live view as well (delay, power usage, framerate in low-light, low resolution, etc.), and neither would solve the FOV problem when using the optical VF, so it's not a solution here.

Also, you haven't yet addressed the main issues I brought up. I really want to know why the heck prime lenses suddenly jumped in the conversation.
 
Your username was so close to the OP's that I thought he was the one replying to me. Sorry about that.

Still, I do think the issues of "zoom/crop overlap," along with uncontrollably variable framerate would make "continual" crop impractical in reality. Personally, I hope they keep some form of HSC, whether or not the D3 is 1.5x. I just don't see "continual" as compelling or intuitive to use.
 
Also, you haven't yet addressed the main issues I brought up. I
really want to know why the heck prime lenses suddenly jumped in
the conversation.
Hi Kenn, I think what the OP is trying to say is this:

By changing the crop size of the sensor you effectively change the focal length of the lens (a 300mm lens on full-frame becomes 450mm at 1.5x). If you could alter the crop size continuously you could, in effect, zoom with a prime lens.

I don't know how possible this is and I know absolutely nothing about any D3. I just thought I'd try to help clear this specific point up :)
 
Yep, I'm not the OP. No worries.
Well, I'm no fan of EVFs (I doubt many DSLR users are), but I don't
have anything against live view. The thing is, most of the
arguments against EVFs hold for live view as well (delay
I've never used any of the dSLR continuously driven sensor implementations, but this seems like something you could throw a combination of brute force and clever engineering at to make almost tolerable. It will be very expensive...
Absolutely. But then again, it will probably be very large..
framerate in low-light
I'm not sure why this is would be an issue.. Not a jab, I'm really clueless on this point.
low resolution
Ironically, as you zoom/crop the scale factor of the LCD resize would get larger. Cheating, sure, but one could argue that it works towards solving this problem without even upping the LCD res! Hey now!
and neither
would solve the FOV problem when using the optical VF, so it's not
a solution here.
I don't see how you could solve the problem in the optical domain when it's inherently not an optical mechanism. However, it seems to me like a possible solution. Is it the best solution? Probably not. Is it a good solution? Maybe. Butler Lampson: "the best is the enemy of the good."
 
Kenn, read the post in this thread called "Source AND speculation". That's where I mentioned the "prime" connection.

While I guess it's possible live view will be offered, it's not a feature that would appeal to pros.
Well, I'm no fan of EVFs (I doubt many DSLR users are), but I don't
have anything against live view. The thing is, most of the
arguments against EVFs hold for live view as well (delay, power
usage, framerate in low-light, low resolution, etc.), and neither
would solve the FOV problem when using the optical VF, so it's not
a solution here.

Also, you haven't yet addressed the main issues I brought up. I
really want to know why the heck prime lenses suddenly jumped in
the conversation.
 
This feature would be completely independent of the zoom function on zoom lenses, but has the most benefit for prime lenses. I use a 300mm for 90% of what I shoot, so now it becomes (assuming 1.1X sensor and 2X HSC crop) a 330mm-600mm zoom instead of being (choose 1) a 330mm or a 600mm.

Why not T-W rings concentric with the shutter buttons, just like point/shoots? Or, more likely, dual control dials like the fake D3 picture showed?

I have to admit I have no idea how the D2Xs mask works. You make it sound as if it would be hard to make it variable. I certainly don't believe there will be an electronic viewfinder.
 
Because continuous would be insanely difficult to engineer. The software needed to calculate the pixel set for any particular crop would probably drive the camera performance into the toilet. That, and it doesn't make any sense anyway. You already have that capability. It's called the crop tool on your editor of choice.
--

Cheers,
Eric
 
Yep, I'm not the OP. No worries.
Yeah, too fast on the draw here :)
Kenn Hwang wrote:
I've never used any of the dSLR continuously driven sensor
implementations, but this seems like something you could throw a
combination of brute force and clever engineering at to make almost
tolerable. It will be very expensive...
Expensive, yes. But this is a pro flagship cam, so I'd let that slip. However, from a user perspective, it'd be incredibly confusing and complex too, which is mainly why I think it would be a no-go.
framerate in low-light
I'm not sure why this is would be an issue.. Not a jab, I'm really
clueless on this point.
Analogy is full-open f/stop in an optical viewfinder. In order to let you see your subject in an EVF/Liveview, digicams slow the framerate to allow more light in (so each frame shows you 1/5sec instead of 1/30sec) for the display. If you have a P&S cam, you can see the display lags a lot more in low light. This would be unacceptable in a pro camera in non-studio settings (and in a studio setting, you wouldn't need HSC or a "continuous" crop setting).
low resolution
Ironically, as you zoom/crop the scale factor of the LCD resize
would get larger. Cheating, sure, but one could argue that it works
towards solving this problem without even upping the LCD res! Hey
now!
This isn't really the same issue I'm addressing - at any crop the current resolution of EVFs/LCDs is still inadequate to reliably manual focus or confirm AF. It's already incredibly difficult with smaller optical VFs like the D70! And again, showing how it may work with "extreme zoom" at max HSC demonstrates a specialized and expensive feature is only useful at the margins - not very practical imo!
and neither
would solve the FOV problem when using the optical VF, so it's not
a solution here.
I don't see how you could solve the problem in the optical domain
when it's inherently not an optical mechanism. However, it seems to
me like a possible solution. Is it the best solution? Probably not.
Is it a good solution? Maybe. Butler Lampson: "the best is the
enemy of the good."
HSC essentially masks the sensor. The D2x's solution for matching the VF view to the masked sensor is by electroncially masking the VF with a blacked-out LCD frame. If you don't use an EVF, the only real solutions you have to display a "continuous HSC" mode is either to have an insanely silly number of separate LCD masks for every crop range between 1.1x and 2.0x (and a 2.0x crop on a FF view would be TINY, to the point of near-uselessness), or an optical solution that uses a mechanism to "zoom" viewfinder to correspond to the crop. P&S cameras still do this (though the view is of course not TTL), and olld rangefinders had pop-up magnifying lenses for use with different lenses. Again, expensive, impractical, and antithetical to the exacting precision and clarity Nikon/Canon demand for their pro flagship cameras.

Again, I'm not saying it can't be done on a technical basis. Just that it introduces so many inconsistencies and singular quibbles that it's realistically not a practical solution.
 
Kenn, read the post in this thread called "Source AND speculation".
That's where I mentioned the "prime" connection.
But it still makes no sense to have brought up primes unless you're ready to say that this feature will only work with primes. Unless you're saying that now, you still haven't addressed the issue with regard to zooms.

Electronic zoom works on P&S cams because its function is tied directly to the zoom controls. Having independent zoom and continuous HSC controls sounds like a nightmare. You zoom all the way in, but have to go to a different set of controls to crop in more? Or worse, you try zooming all the way out but you're cropped in, requiring that you change controls to crop out too? And everything in between - you're at some intermediate zoom and crop, but have to be mindful of your resolution/speed? Then you have to fudge with two separate controls working at odds, just to recompose to the same frame you currently want? All the while your max FPS is flucutating because you're continually cropping and uncropping by benefit of the "continuous" function? It just doesn't sound like it's worth it.

Finally, to answer your other question, the D2Xs uses an LCD mask for its HSC. It's workable because the mask is set at a single value (2.0x I think). Each LCD mask has to be discretely cut into the substrate glass, and if your HSC is truly "variable," then using the same mask option introduces a host of problems: 1) you'd have a huge maze of leads and frame cutouts all around the VF to accomodate the continuous crop values, 2) even then, you still wouldn't have true 100% frame coverage at any given crop (which Nikon wouldn't give up on a flagship), 3) a 2.0x mask over a 1.0x/1.1x VF would be just tiny. There are just so many reasons to be incredulous, but as you can see, I do enjoy the conversation :)
 
Hmmm,

I often lurk and read the comments, but this sounds to good not to put in my two cents. Looking at the source, I think it's plausible to believe that something is coming. I myself would love to see a full frame sensor, however, I am more interested in a camera that give me less noise at higher ISO settings in addition to a wider ISO range.

What I would really like is a Nikkor 24 -120 f2.8 VR lens from Nikon. Now that would be something!

--
Nikon D2X, D70, F100,
Contax 645 w/MegaVision S3 Digital Back
 
Also, you haven't yet addressed the main issues I brought up. I
really want to know why the heck prime lenses suddenly jumped in
the conversation.
Hi Kenn, I think what the OP is trying to say is this:

By changing the crop size of the sensor you effectively change the
focal length of the lens (a 300mm lens on full-frame becomes 450mm
at 1.5x). If you could alter the crop size continuously you could,
in effect, zoom with a prime lens.

I don't know how possible this is and I know absolutely nothing
about any D3. I just thought I'd try to help clear this specific
point up :)
Of course it's not changing the crop factor that changes the effective focal length of the lens, it's the pixel pitch so unless the pixels can get bigger or smaller on demand this idea is pretty much dead in the water!
 
Hello. I have just stumbled on this thread and it makes sense to me.

Firstly, I bought a D2x simply because of the x2 crop. It means that smaller, lighter lenses can be used and the crop switched in a second.

Secondly, it's true that even today's D2x sensor can out-resolve some zoom lenses.

When a London dealer told me that a FF camera was on the way (a friend of his is involved in pre-production testing) my first thought was of the resolving power of an average lens - how could it keep up with a really big-pixel sensor?

The hint that snappers should hang on to their primes makes perfect sense for a power FF or near-FF sensor.

And there's nothing to stop it having a D2x-style selective crop. We have all got use to the extra magnification of DX sensors. Smaller files would also speed up file saving times.

I would love the idea of a camera that could switch from low noise/mega-resolving power (I have Canon 5D files to hand and they are amazing) to x2 crop convinience in 3 steps. It's just the viewfinder problem of a tiny x2 crop that is a hurdle.
 

Keyboard shortcuts

Back
Top