Why no ISO100 w/o boost?

I wouldn't mind ISO50 or 25, even. Also, what does it mean to have "boost" to go from ISO 200 to ISO 100?

D300 looks killer -- I guess I'll be calling the camera shop tomorrow... ;-)
--
Thanks,
Mark

boost = they don't guarantee color accuracy

One thing I'd love to know about is how the DR compares to previous models. .Is it the same number of stops? increase?

And not to say the least, how's noise and banding.. can't wait for the hands-on reviews

--

If you find my gallery at http://www.digitaldias.com , you're welcome to use visitor/visitor for guest access

I'm wondering the same thing about the 100 iso. Why is it not without the boost?

Londongal said:

I'm wondering the same thing about the 100 iso. Why is it not without
the boost?

Click to expand...

Because its not the base ISO. All ISOs other than the base ISO are either over-amplifications or under-amplifications (if there is such a word!) of the actual signal the sensor & allied circuitry has been designed to output.

--
Click

MrClick said:

Londongal said:

I'm wondering the same thing about the 100 iso. Why is it not without
the boost?

Click to expand...

Because its not the base ISO. All ISOs other than the base ISO are
either over-amplifications or under-amplifications (if there is such
a word!)

Click to expand...

I'd go with attenuation.

MrClick #421805 said:

of the actual signal the sensor & allied circuitry has been
designed to output.

--
Click

Click to expand...

--
-Steve
===================

When everything is coming your way, you're in the wrong lane. ~ Larry the Cable Guy

I guess the ISO 100 performance would be one thing to look for in the forthcoming reviews.

--
Thanks,
Mark

MarkD2 said:

I wouldn't mind ISO50 or 25, even. Also, what does it mean to have
"boost" to go from ISO 200 to ISO 100?

D300 looks killer -- I guess I'll be calling the camera shop
tomorrow... ;-)
--
Thanks,
Mark

Click to expand...

The base ISO of a sensor is determined by how well it detects light and by the 'depth' of the electronic well in the photosensor. Once that well is full, the pixel is saturated and further reductions in ISO are only going to overexpose more pixels.

Let's say we have a sensor that is accurately exposed in sunny daylight at f16 with an exposure of 1/100 second. This would correspond to 100 ISO and while we could increase the exposure, all that would happen is that pixels would start to be oversaturated and clipping would occur so we wouldn't really gain anything. To reduce the base ISO, we would have to either reduce the sensitivity of the sensor so that fewer photons were converted into charge, or, increase the size of the electron well in each pixel. If we doubled the amount of charge that each pixel could hold, we could expose for twice as long before saturation occured and our base ISO would have halved. Lowering the base ISO by the first method is clearly a bad idea as we don't wannt less sensitive sensors but lowering it by the second method is limited by the behaviour of the silicon and the size of the light sensitive part of the pixel. CMOS imagers tend to be more restricted than some types of CCD as they have quite a lot of circuitry in each pixel that reduces the area that can be used to hold charge.

Pedro Dias said:

One thing I'd love to know about is how the DR compares to previous
models. .Is it the same number of stops? increase?

Click to expand...

Going from a 12 to 14 bit DAC would give two extra stops, but it's probably noise limited rather than DAC resolution limited.

Pedro Dias said:

And not to say the least, how's noise and banding.. can't wait for
the hands-on reviews

Click to expand...

The CMOS sensor should have lower noise (especially in the D3 with its larger sites) and banding should not be an issue since the DACs are on the sensor and are basically guaranteed to track.

Thank you for the info on 100 iso and boost. I'm always so appreciative of how much knowledge people share on these forums.

best wishes,
Londongal

Goona get me a D300

Londongal said:

I'm wondering the same thing about the 100 iso. Why is it not without
the boost?

Click to expand...

Having a base ISO of 200 will only have a minimal impact on most users that have been shooting at ISO 100 but it can go a long way towards quieting the incessant whining over high ISO noise characteristics by requiring less amplification at any given ISO over 200.

--
'Here, look at the monkey. Look at the silly monkey!'

Tom Young
http://www.pbase.com/tyoung/

This is an often misunderstood issue.

A base ISO 200 verses 100 is GOOD, not bad, in most fundamental ways. It means the sensor is more sensitive and, all other things equal, should deliver better high iso performance. It can develop a full exposure with half the light. That is a good thing.

I think Nikon put the ISO 100 "cut" in there to please numbers measurebators ("what? No ISO 100?), and as a convenience feature to allow fast lenses to be shot wide open without adding ND filters. And, of course, other recent models "support" this, for better or worse and it is a perceived benefit. That is more marketing than technological need.

Each time you double (or increase at any level) the output of the sensor, you add more noise. ISO 800 has to be doubled twice from base ISO 200, but THREE TIMES from base ISO1 100.

Having the ISO 100 "cut" saves you the trouble of mounting a 1 stop ND filter. That's all it does (most likely, but a good speculation).

A base ISO 400 would be even better, indicating an even more senstive sensor. You then have issues, of course, shooting fast lenses wide open in bright light, but everything is a double edged sword. A base 400 sensor only has to be doubled twice to get to ISO 1600.

In the film days, low ISO was "good" because low ISO film was finer grained. In digital terms, it would be like varying the number of pixels based on your ISO setting. People put up with low ISO slide film to get the fine grain detail (and other attributes). The low ISO rating was not, in itself, a benefit for anyone other than masochists, especially shooting handheld. We put up with it to get fine grain, not because we liked long shutter speeds (in most cases- there are exceptions like shooting waterfalls, I guess)

This is still true in a digital sense, to the extent that low res sensors are less noisy than high res sensors, but within a single body, for a given sensor, higher ISO for a given res/noise level is GOOD, not bad. If Ektachrome 200 had the same resolution and small grain size as Ektachrome 25, no one would ever have shot Ekt 25.

Does that make sense?

--
Regards,
Neil

Hi Neil,

That makes sense to me, absolutely; particularly when you frame (pun not really intended, but what the heck!) the analogy around grain and Ektachrome.

thanks!
Londongal

In a sense it is identical, I think, although I'm not a camera or film engineer.

Big film crystals captured more photons than small crystals, at the expense of resoluton. At least that is the end result.

FF is a sneaky way to reduce the megapixel count. There is no (serious, and viable) market for a $2K 5mpx camera. But a 10mpx FF camera is really a 5mpx DX camera, with more space. Think about it

--
Regards,
Neil

High base ISO is not in itself a good thing and without knowing other aspects of the camera's performance doesn't tell you much about how good sensitivity or noise control is.

It is trivial to raise base ISO by just reducing the ability of each pixel to hold charge. You wouldn't have improved the sensitivity at all but you would have increased base ISO because the pixels would saturate sooner. Lowering base ISO by increasing well size is a good thing as it allows for greater dynamic range and signal/noise ratio. The ideal sensor would detect 100% of incident light but also have a base ISO as low as possible through the use of a massive charge storage capability within each pixel.

ISO in itself is not always a good way of looking at the performance of a sensor which is why for scientific cameras you will see other figures quoted:

QE - quantum efficiency - the percentage of photons hitting the photodiode that are converted to charge.

Fill ratio - the proportion of each pixel that is occupied by the photodiode and is light sensitive

Well capacity - the amount of charge (number of electrons) that can be stored by each photodiode.

Read noise - noise introduced per pixel when the image is read from the chip.

Thermal noise - noise introduced per pixel during exposure - scales linearly with exposure time and approximately doubles for every 6-7 degree C rise in temperature.

Obviously it introduces more complexity but these figures allow a more accurate assessment of a sensor's performance than a simple ISO figure.

Obviously base ISO as a yardstick oversimplifies things, but here I am simplifying things to the extent that ALL OTHER THINGS EQUAL, high base ISO is better. That assumes, of course that all things are equal and the ISO 100 sensor has the same noise as the ISO 200 sensor at their respective base ISO levels. I think that is in the context of the OP's "complaint".

Doesn't a higher base ISO sensor store charge more efficiently than a lower base ISO sensor? (All other things equal of course). Your statement seemed contradictory to me.

--
Regards,
Neil

Nevermind... I misread this:

NRothschild said:

NRothschild said:

Well capacity - the amount of charge (number of electrons) that can be
stored by each photodiode.

Click to expand...

Click to expand...

And was thinking about "rate of charge" rather than QUANTITY. I understand what you are saying.

Still, all things equal, I think many people here misunderstand the basic concept of what base ISO means to them, in real world shooting.

How exactly does "Lo1" (ISO 100 on D300) work? Is the sensor itself somehow "desensitized"? I have never shot a DSLR that had this feature.

--
Regards,
Neil

Technical talk to one side. What we are assuming is the base ISO 200 of the D300 will produce the same image quality as the D2x does at its base ISO of 100? Only with the advantage/disadvantage of being an F/stop faster depending on how you look at it? Correct or am I missing something here?

Simon
--
http://www.sbphotography.org.uk

I am assuming that. Can't speak for others. That is why I stressed "all else equal" above. I have confidence that Nikon did not go backwards on basic image quality. To me, basic image quality includes noise at base ISO. That's a religious and pilosophical statement at this point.

I could also argue that we should, for the time being, assume the D300 is at least as good at the base ISO, until proven otherwise with actual images and user reports. And I will

Not that I am any expert, but here we are all free to act like experts

Neil

X-Sync said:

Technical talk to one side. What we are assuming is the base ISO 200
of the D300 will produce the same image quality as the D2x does at
its base ISO of 100? Only with the advantage/disadvantage of being an
F/stop faster depending on how you look at it? Correct or am I
missing something here?

Simon
--
http://www.sbphotography.org.uk

Click to expand...

--
Regards,
Neil

NRothschild said:

I think Nikon put the ISO 100 "cut" in there to please numbers measurebators

Click to expand...

Those of us who shoot landscapes or other types of things where you want controlled blurs like low end ISOs. It isn't about measurbation. It is about not having to add ND filters. We want 100 or 50 in order to blur things.

Having the base at 200 is likely a nod toward all the High ISO noise furor and a desire to have a very clean high ISO camera.

--
Ed C.

NRothschild said:

Obviously base ISO as a yardstick oversimplifies things, but here I
am simplifying things to the extent that ALL OTHER THINGS EQUAL, high
base ISO is better. That assumes, of course that all things are
equal and the ISO 100 sensor has the same noise as the ISO 200 sensor
at their respective base ISO levels. I think that is in the context
of the OP's "complaint".

Doesn't a higher base ISO sensor store charge more efficiently than a
lower base ISO sensor? (All other things equal of course). Your
statement seemed contradictory to me.

Click to expand...

Well, if QE and fill ratio remain the same between our two sensors then actually a LOWER base ISO is better because it means that particular sensor has larger capacity electron wells and is therefore capable of greater dynamic range and S/N ratio.

If well capacity is the same in each case then a HIGHER base ISO is better because it means that QE and/or fill ratio is higher so the chip is more sensitive.

Unfortunately, because the base ISO is related to these three factors, it becomes very complex - you could have two sensors with the same base ISO but one is twice as sensitive and with twice the electron well depth. The two improvements would ironically cancel each other out so you wouldn't know that the sensor was actually better.