Nikon's ISO 200

Alan Browne wrote:
Me wrote:
Alan Browne wrote:
Me wrote:
Paul Furman wrote:

Dynamic range is what concerns me more and I don't know how much
impact there is on that.
The data I've seen suggests that with a D300, dynamic range
(measured at a standard signal to noise ratio) is very slightly
better at "Lo 1.0" than ISO 200, and it's very slightly worse with a
D3.

Could you clarify:

Does that mean "ISO 200 + Lo. 1.0"?

Lo 1.0 is one stop below ISO200.
(increments can also be set)

Still not clear to me.

1. Can you even set ISO 100?

2. To do get this "one stop below ISO200" do you set ISO 200 _AND_ Lo 1.0?

No, just set the dial down another stop - it's not an extra setting.
"ISO 100" could be displayed, but Nikon chose to display it as "lo"
(plus increment of stop) below true ISO, as well as "Hi" (plus increment
of a stop) at above true high ISO.
For all practical purposes it is ISO 100. For pixel-peepers, perhaps
not. Perhaps Nikon just chose to pre-empt whining from pedants.

Me wrote:
Alan Browne wrote:
Me wrote:
Alan Browne wrote:
Me wrote:
Paul Furman wrote:

Dynamic range is what concerns me more and I don't know how much
impact there is on that.
The data I've seen suggests that with a D300, dynamic range
(measured at a standard signal to noise ratio) is very slightly
better at "Lo 1.0" than ISO 200, and it's very slightly worse with
a D3.

Could you clarify:

Does that mean "ISO 200 + Lo. 1.0"?

Lo 1.0 is one stop below ISO200.
(increments can also be set)

Still not clear to me.

1. Can you even set ISO 100?

2. To do get this "one stop below ISO200" do you set ISO 200 _AND_ Lo
1.0?

No, just set the dial down another stop - it's not an extra setting.
"ISO 100" could be displayed, but Nikon chose to display it as "lo"
(plus increment of stop) below true ISO, as well as "Hi" (plus increment
of a stop) at above true high ISO.
For all practical purposes it is ISO 100. For pixel-peepers, perhaps
not. Perhaps Nikon just chose to pre-empt whining from pedants.

Got it. Thx.

Nothing pre-empts whining from pedants.

Me wrote:
Paul Furman wrote:

Dynamic range is what concerns me more and I don't know how much
impact there is on that.
The data I've seen suggests that with a D300, dynamic range (measured at
a standard signal to noise ratio) is very slightly better at "Lo 1.0"
than ISO 200, and it's very slightly worse with a D3. That's probably
due to the fact that at ISO 200, "shot" noise with APS-c is slightly
over 2x that with 35mm, read noise more or less the same, but shot noise
decreases slightly as ISO reduces a further half stop to "native" ISO.
Despite "common knowledge" to the contrary, there's actually very little
difference in usable dynamic range between current APS-c and 35mm Nikons
at base ISO (200), except for the D3x, which has extremely low read
noise reducing down to ISO 100. I haven't seen any data for the D3s yet.

What do you mean by 'measured at a standard signal to noise ratio'? Is
that accomplished by underexposing the Lo 1 shot?

Presumably they wouldn't give it a special name if there wasn't some
trade-off. It's effectively like overexposing so there ought to be more
blown highlights & less noise.

--
Paul Furman
www.edgehill.net
www.baynatives.com

all google groups messages filtered due to spam

Paul Furman wrote:
Me wrote:
Paul Furman wrote:

Dynamic range is what concerns me more and I don't know how much
impact there is on that.
The data I've seen suggests that with a D300, dynamic range (measured
at a standard signal to noise ratio) is very slightly better at "Lo
1.0" than ISO 200, and it's very slightly worse with a D3. That's
probably due to the fact that at ISO 200, "shot" noise with APS-c is
slightly over 2x that with 35mm, read noise more or less the same, but
shot noise decreases slightly as ISO reduces a further half stop to
"native" ISO.
Despite "common knowledge" to the contrary, there's actually very
little difference in usable dynamic range between current APS-c and
35mm Nikons at base ISO (200), except for the D3x, which has extremely
low read noise reducing down to ISO 100. I haven't seen any data for
the D3s yet.

What do you mean by 'measured at a standard signal to noise ratio'? Is
that accomplished by underexposing the Lo 1 shot?

Presumably they wouldn't give it a special name if there wasn't some
trade-off. It's effectively like overexposing so there ought to be more
blown highlights & less noise.

If you expose a frame so that highlights aren't lost, then you can
recover detail from the shadows to more than is visible in a print,
either selectively by area, or selectively by "levels" etc to the whole
frame, or both.
The more photographic stops you raise the shadows, the closer you come
to the point where detail in the darkest areas is lost to noise.
So assume a standard print size, then measure how many stops adjustment
can be applied to shadows until a standard s/n ratio applies. Then
you've got some basis for comparison of dynamic range between raw files.
It's not perfect though, as it doesn't tell you how the noise looks at
the limit. Random fine noise might be perfectly acceptable, but blotchy
or pattern noise, banding etc may not be.

Me wrote:

"ISO 100" could be displayed, but Nikon chose to display it as "lo"
(plus increment of stop) below true ISO, as well as "Hi" (plus increment
of a stop) at above true high ISO.
For all practical purposes it is ISO 100.

Except that I understand that what should just fill a pixel
well at ISO 100 might well overfill it at lo +1.0. Hence
'lo' and not ISO 100.

-Wolfgang

Wolfgang Weisselberg wrote:
Me wrote:

"ISO 100" could be displayed, but Nikon chose to display it as "lo"
(plus increment of stop) below true ISO, as well as "Hi" (plus increment
of a stop) at above true high ISO.
For all practical purposes it is ISO 100.

Except that I understand that what should just fill a pixel
well at ISO 100 might well overfill it at lo +1.0. Hence
'lo' and not ISO 100.

-Wolfgang

Yes, but that loss of dynamic range at the high end is offset by reduced
shot-noise, at least down to "real" ISO 150 or so, so in the end there's
not much practical difference. Perhaps a bit more care should be taken
not to overexpose highlights - so if that means under-exposing a little,
then adjusting again in PP, you may as well just use ISO 200 in the
first place.
ND filters seem to be the only answer if you want long exposures at
apertures large enough not to cause significant diffraction losses.
Has anyone tried the Singh-Ray "vari" ND filters? Are they good? At
US$390 or so I'd hope so.

Me wrote:
Wolfgang Weisselberg wrote:
Me wrote:

"ISO 100" could be displayed, but Nikon chose to display it as "lo"
(plus increment of stop) below true ISO, as well as "Hi" (plus increment
of a stop) at above true high ISO.
For all practical purposes it is ISO 100.

Except that I understand that what should just fill a pixel
well at ISO 100 might well overfill it at lo +1.0. Hence
'lo' and not ISO 100.

Yes, but that loss of dynamic range at the high end is offset by reduced
shot-noise, at least down to "real" ISO 150 or so, so in the end there's
not much practical difference.

is it?

After all, you are throwing away rather large parts of your
digital numbers coming from the sensor: your sensor clips before
they can reach the maximum value.

Perhaps a bit more care should be taken
not to overexpose highlights - so if that means under-exposing a little,
then adjusting again in PP, you may as well just use ISO 200 in the
first place.

And hence 'lo' and not ISO 100.

ND filters seem to be the only answer if you want long exposures at
apertures large enough not to cause significant diffraction losses.

Waiting for the dark hours of the night works just as well. :-)

And of course you can shoot long series and blend them into one
long exposure --- ask the astrophotographers, they do it all
the time.

-Wolfgang