Why no digicam with large sensor...

[Petteri Sulonen]

No theoretical reason, but for now a very practical one. DSLR
sensors are not designed to show a "preview". Generally they have
no dedicated circuitry for refreshing the image continuously while
focusing/viewing. This means they could never be used with an EVF
or "preview" LCD and require some other mechanism for framing,
focusing and exposure.

Actually, many current DSLR sensors can be sampled continuously.

This would increase noise because the sensor is getting hotter.
Probably a few more issues with tihs approach. Having a second
sensor maybe better.

Not by much. The power drain of a sensor is pretty negligible. The signal processing is the bit that generates heat. OTOH, the double-sensor solution would be a great deal more complex and expensive -- in particular, you'd have to have the secondary sensor the same size as the primary one, if you wanted the same field of view on the image. Trying to fit two APS-C sized sensors in one DSLR-sized body would be a quite a squeeze.

[snip]

Petteri
--
[ http://www.prime-junta.tk ]

 

[Michael Long]

Not by much. The power drain of a sensor is pretty negligible. The
signal processing is the bit that generates heat. OTOH, the
double-sensor solution would be a great deal more complex and
expensive -- in particular, you'd have to have the secondary sensor
the same size as the primary one, if you wanted the same field of
view on the image. Trying to fit two APS-C sized sensors in one
DSLR-sized body would be a quite a squeeze.

Ummm.... the mirror/pentaprism/optics assembly of a DSLR manages to deliver the current FOV to an eyepiece about 12mm in diameter.

Two full-size sensors isn't the only way to go about it.

Much like the detachable prism assembly of the Nikon F3, you could probably build a camera where you could swap the optical prism with an EVF version.

 

[Petteri Sulonen]

Not by much. The power drain of a sensor is pretty negligible. The
signal processing is the bit that generates heat. OTOH, the
double-sensor solution would be a great deal more complex and
expensive -- in particular, you'd have to have the secondary sensor
the same size as the primary one, if you wanted the same field of
view on the image. Trying to fit two APS-C sized sensors in one
DSLR-sized body would be a quite a squeeze.

Ummm.... the mirror/pentaprism/optics assembly of a DSLR manages to
deliver the current FOV to an eyepiece about 12mm in diameter.

But it's not in focus at the plane of the eyepiece: it's focused on the ground glass, which is the same size as the sensor.

I'm sure it could be done, though, with some additional optics, especially if optical quality isn't of paramount importance. In any case, it would add a quite a bit of complexity to the setup -- you couldn't just replace the mirror with a sensor, f'r instance.

Two full-size sensors isn't the only way to go about it.

Much like the detachable prism assembly of the Nikon F3, you could
probably build a camera where you could swap the optical prism with
an EVF version.

And what would be the advantage of the EVF over the OVF in that case?

Petteri
--
[ http://www.prime-junta.tk ]

 

[Michael Long]

But it's not in focus at the plane of the eyepiece: it's focused on
the ground glass, which is the same size as the sensor.
I'm sure it could be done, though, with some additional optics,
especially if optical quality isn't of paramount importance. In any
case, it would add a quite a bit of complexity to the setup -- you
couldn't just replace the mirror with a sensor, f'r instance.

"Quite a bit" may be overstating the case a bit. I think, for example, the exposure meters are also located in the prism assembly. An EVF ccd could be used to eliminate that complexity.

And what would be the advantage of the EVF over the OVF in that case?

Well, theoretically you could also feed the results back to the LCD in the body.

As to other advantages, heck, ask the guys who're wanting one. I'm simply responding to the two-full-size-sensors-required statement.

 

[Petteri Sulonen]

But it's not in focus at the plane of the eyepiece: it's focused on
the ground glass, which is the same size as the sensor.
I'm sure it could be done, though, with some additional optics,
especially if optical quality isn't of paramount importance. In any
case, it would add a quite a bit of complexity to the setup -- you
couldn't just replace the mirror with a sensor, f'r instance.

"Quite a bit" may be overstating the case a bit. I think, for
example, the exposure meters are also located in the prism
assembly. An EVF ccd could be used to eliminate that complexity.

The trouble is that you need to focus the image on the EVF, while there's no requirement for that with the exposure meters. I'd guess you'd need at least four elements (two doublets) to do the job adequately.

And what would be the advantage of the EVF over the OVF in that case?

Well, theoretically you could also feed the results back to the LCD
in the body.

Which would need additional electronics -- an interface with the body that feeds to the LCD.

As to other advantages, heck, ask the guys who're wanting one. I'm
simply responding to the two-full-size-sensors-required statement.

You're right, actually, two full-sized sensors isn't the only solution (and probably not even the best solution). However, you would need a significant amount of optics to get the image focused on a smaller secondary sensor, which would take up at least as much space and not be that cheap to make, either.

Petteri
--
[ http://www.prime-junta.tk ]

 

[*isteve]

Yes you can - but a 4MP LBCAST sensor can only operate at 8fps, not enough for even the most turgid video output to an EVF.

Sure there are ways round it - most only scan a fraction of the rows for instance - but all CCDs and APS sensors (CMOS, LBCAST whatever) still need to generate a charge to trigger the read operation which requires a power supply. Doing this continuously 30X a second does use power and generate heat. Not as much as an LCD for sure, but the heat is generated in the sensor where you least want it.

Like you, I would love a mini APS sensor device I can slip in a pocket, but the *istD with the 50mm F1.7 is pretty small. Eliminating the mirror box and pentaprism would save a bit of space, but then you need to add a video processor and an EVF.

I think there is still quite a bit of room left to "shrink" APS cameras even more if you could just dispense with the lens mount. After all the *istD body is not much bigger than a Canon G5 and smaller than my Coolpix 990.
Steve

No theoretical reason, but for now a very practical one. DSLR
sensors are not designed to show a "preview". Generally they have
no dedicated circuitry for refreshing the image continuously while
focusing/viewing. This means they could never be used with an EVF
or "preview" LCD and require some other mechanism for framing,
focusing and exposure.

Actually, many current DSLR sensors can be sampled continuously. At
least the Canon and Foveon CMOS sensors can do this. I don't know
about the Sony CCD's used on Nikons, though. It would be quite
feasible to make a DSLR that would allow you to flip up the mirror
and use contrast-detection AF and live preview, just like a
digicam. Trouble is, it would cost money to develop these features
in the firmware, and the manufacturers probably figure that the
features probabaly wouldn't sell enough cameras to make them
worthwhile. I think we'll see a DSLR that works like this one day
pretty soon, though.

That mechanism is already available in autofocus SLRs, so there was
no need to include the additional circuitry. This means simpler
chip designs with more real estate available for the light sensing
part of the CCD, less thermal noise and (because they are only
switched on for a fraction of a second) much lower power
consumption. An APS chip used for preview would porobably use a lot
more power than a 2/3" chip.

Other than the bit about power consumption, this only applies to
some CCD's, not all DSLR sensors.

So, what you ask would be possible with a new chip, or with a fixed
lens SLR (many of these existed, like most Olympus SLRs for many
years) but to what advantage? An SLR with a Sigma 18-125 lens would
give you roughly the same thing. Look at the Pentax *istD and you
reallise it would not even have to be bigger either. The *istD is
no bigger overall that many bridge cameras and smaller than the
Sony 828.

Yah, but I want something that fits in a jacket pocket and is more
discreet than an SLR. IOW, I'd really like a compact digicam with
an APS-C sensor. A bright prime (say, f/2) between 28 and 50 mm
equivalent would be ideal, but I'd settle for, say, a 28-70 f/4
zoom.

Petteri
--
[ http://www.prime-junta.tk ]