Why no ISO 100 ??

"Mike Cawood, HND BIT" wrote in news:6d57tjFuujvU1
@mid.individual.net:

Some do, my Fujifilm S5700 goes down to ISO64.

Some cameras are low, but they don't tell you. The base ISO of the Canon
10D, for example, is actually about 64 even though it says "100". I have
shot my 10D, externally metered with a grey card to ISO 32, without
clipping white matte highlights (such as metering a grey card, covering it
with the color checker, and using the manual exposure given by the meter.
The white patch is just short of RAW clipping.

In controlled lighting, most DSLRs can be pulled a stop.

--


John P Sheehy

John Sheehy wrote:
Alan Browne wrote in
:

The reality is that most sensors in the "35mm" class have a natural
sensitivity somewhere between 100 and 200. (Which is why many of the
Sony based Nikon's all start at 200).

Well, I don't think they really warrant 200 per se, but for some stupid
reason, manufacturers like to have the base ISO as 50*(2^n), so since the
sensors can't do 100 with a repectable highlight headroom in the RAW data,
they make 200 the base "normal" ISO, but the camera might really be ISO 130
to 160 or 180, and they treat 160 as a weird ISO, when in fact it, or 140,
or whatever, should be the base ISO. They're not doing anyone any favors,
except the people who like to compare cameras at specific ISOs. Many
cameras could have more useful DR and SNR curves if the cameras started at
an ISO that digitized almost full-well capacity (with real gain, at the
critical points in the signal path).

Just stating the case. For example the D100/70/50, Maxxum 7D/5D and
Pentax (whatever the 6 mpix was) all had essentially the same chip.

On the Nikon's ISO began at 200; on the Maxxum's at 100 (Pentax too, IIRC).

Yet, the actual natural ISO was somewhere around 130 - 140 (judging from
the noise curves

The latest 1Ds Mk III has an "optional" ISO 50, which I decode as
meaning digital desensitizing (and clipped highlights if so).


--
-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm
-- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm
-- [SI] gallery & rulz: http://www.pbase.com/shootin
-- e-meil: Remove FreeLunch.

In article , John Sheehy
says...

The slappin' mirror has got to go the way of the dinosaurs. It is
primitive BS.

And the cameras should use the main sensor for metering (most DLSRs with
live preview don't do that). The only problem is that to have phase AF
you need a mirror.
--

Alfred Molon
------------------------------
Olympus 50X0, 8080, E3X0, E4X0, E5X0 and E3 forum at
http://tech.groups.yahoo.com/group/MyOlympus/
http://myolympus.org/ photo sharing site

Alan Browne wrote in
:

Yet, the actual natural ISO was somewhere around 130 - 140 (judging
from the noise curves

I don't know what you mean by "judging from the noise curves", as ISO has
nothing to do with noise.

Any standard that looks to equate ISO with noise should be questioned and
abandoned, IMO.

Each ISO setting in a digital camera, IMO, should have two numbers relating
to it; one, the ISO the camera meters for on an evenly illuminated surface
with 0 EC, and the other, the metered ISO which would result in a standard
amount of highlight headroom in the most sensitive color channel in white
light (somewhere from 2.5 - 3.5 stops, possibly). These figures can be
quite different from the number the camera states.

--


John P Sheehy

Alfred Molon wrote:
In article , John Sheehy
says...

The slappin' mirror has got to go the way of the dinosaurs. It is
primitive BS.

And the cameras should use the main sensor for metering (most DLSRs with
live preview don't do that). The only problem is that to have phase AF
you need a mirror.

Phase detection AF requires the image to be reflected onto the sensors,
not necessarily by a single large reflex mirror, as for the viewfinder.
In fact, I'm sure that Phase detection sensors have their own smaller
mirrors which are normally hidden behind the viewfinder reflex mirror.
As far as I'm aware, Phase detection AF has not been implemented in
anything other than SLR cameras.

It should be possible to have a non-slr stills camera with their smaller
mirrors flipping out of the way when the shutter is tripped and because
of their smaller mass shouldn't cause mirror-slap shake.
Something like a RF camera could have an AF mode that uses Phase
detection, but the handful of current RF camera makers (Leica, Cosina,
any-one else?) only have manual focus.

Video would be another problem, as the mirrors for the AF sensors would
have to be out of the way the whole time.
I suppose you could side-step the need for moving mirrors with a
pellicle style mirror for the AF sensors, at least that way there is no
slap and the Phase detection AF sensors could also be used in stills mode.
The downside would be losing at least a stop of brightness going to the
imaging sensor.

Alfred Molon wrote:
[]
And the cameras should use the main sensor for metering (most DLSRs
with live preview don't do that).
[]


Why, when separate metering sensor can have a much greater dynamic range?

David

In article , David J
Taylor says...
Alfred Molon wrote:
[]
And the cameras should use the main sensor for metering (most DLSRs
with live preview don't do that).
[]


Why, when separate metering sensor can have a much greater dynamic range?

Because separate metering sensors are not that precise (there are too
few of them), forcing the camera to be very conservative in setting the
exposure, which often leads to underexposures.

Not sure why greater dynamic range than the sensor can capture would be
an advantage, since the sensor can't capture it anyway.
--

Alfred Molon
------------------------------
Olympus 50X0, 8080, E3X0, E4X0, E5X0 and E3 forum at
http://tech.groups.yahoo.com/group/MyOlympus/
http://myolympus.org/ photo sharing site

Alfred Molon wrote:
In article , David J
Taylor says...
Alfred Molon wrote:
[]
And the cameras should use the main sensor for metering (most DLSRs
with live preview don't do that).
[]


Why, when separate metering sensor can have a much greater dynamic
range?

Because separate metering sensors are not that precise (there are too
few of them), forcing the camera to be very conservative in setting
the exposure, which often leads to underexposures.

Not sure why greater dynamic range than the sensor can capture would
be an advantage, since the sensor can't capture it anyway.

Alfred,

I might argue that there's no "correct" exposure, and taking an average
over a few pixels (because of lack of area precision) is no worse than
taking pixel-sized values, but I take your point.

As to dynamic range, wouldn't you want a meter which had greater range
than the sensor? Otherwise at the high end you just get: "It's over the
limit" rather than it's 5% over the limit or it's 100% over the limit. I
was also thinking that larger pixels in the exposure sensor might lead to
greater precision (in the shadows) than inherently noisier (because they
are smaller) sensor pixels.

Cheers,
David

In article , David J
Taylor says...

As to dynamic range, wouldn't you want a meter which had greater range
than the sensor? Otherwise at the high end you just get: "It's over the
limit" rather than it's 5% over the limit or it's 100% over the limit. I
was also thinking that larger pixels in the exposure sensor might lead to
greater precision (in the shadows) than inherently noisier (because they
are smaller) sensor pixels.

Well, the main sensor of the camera has all the dynamic range you want,
because it can close the diaphragm if there is too much light and open
it more if there is not enough of it.

By the way, the Pentax K20D is an example of a camera which underexposes
because it only has few meter sensors, while the Sony A350 which uses
the live view sensor in the eyepiece to meter is more precise and
underexposes less.
--

Alfred Molon
------------------------------
Olympus 50X0, 8080, E3X0, E4X0, E5X0 and E3 forum at
http://tech.groups.yahoo.com/group/MyOlympus/
http://myolympus.org/ photo sharing site

Alfred Molon wrote:
David J Taylor says...
Alfred Molon wrote:

And the cameras should use the main sensor for metering (most DLSRs
with live preview don't do that).

Why, when separate metering sensor can have a much greater dynamic range?

Because separate metering sensors are not that precise (there are too
few of them), forcing the camera to be very conservative in setting the
exposure, which often leads to underexposures.

Not sure why greater dynamic range than the sensor can capture would be
an advantage, since the sensor can't capture it anyway.

Because the sensor CAN capture it given an appropriate exposure. The
metering sensor doesn't have the luxury of being able to adjust
shutter speed or diaphram to meter light levels from bright sunlight
to near darkness.

And no, adjusting the lens diaphram is not an option. It'd suck
power, wear down the lens, possibly burn out the diaphram motor,
and wouldn't even work for lenses that do not have camera-controlled
diaphrams.

--
Ray Fischer