From: John Sheehy on
Lionel <usenet(a)> wrote in

> It does, because of the fill-factor problem. I've gone into this in
> more detail in my reply to another of your posts in this thread, but
> the bottom line is that the more pixels you put in a given area, the
> more of that area is being used for things other than collecting
> photons. In fact, as you increase the number of pixels in an area, you
> eventually reach the point where the whole area is support
> electronics, & there's no longer any room left for the actual
> photodiodes.

The Panasonic FZ50 collects as many photons at ISO 100 saturation, per unit
of sensor area, as the 1DmkII. This is a real-world fact, that shows that
your concern is pretty much a boogey-man story, in the range of current
pixel sizes. And, even when miniaturization of the sensel *does* lead to
photon loss per unit of area, it takes a huge difference in photon
collection to make a difference in shot noise. Shot noise is not
proportional to signal; it's proportional to its square root.


<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS(a)no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
From: John Sheehy on
Lionel <usenet(a)> wrote in

> Having said that, assuming that microlenses work similarly to lenses
> at the macro level, on the one hand, they should definitely help
> improve the signal to noise ratio at the level of the individual
> photodiode, but on the other hand, will fill the smaller well sooner,
> decreasing the effective dynamic range of each pixel.

Microlenses have zero effect on dynamic range. They only alter sensitivity
(and reduce aliasing strength somewhat, if insufficient AA filtering is

> And because the
> output signal will still be small, you'll get more amplifier noise
> when you read the signal from the pixel.

But not relative to absolute exposure.

> So my gut feeling is that as
> you decrease the pixel size, you'll start seeing an exponential factor
> appearing in the noise vs pixel-size curve.

Noise per pixel or per unit of sensor area?


<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS(a)no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
From: Roger N. Clark (change username to rnclark) on
John Sheehy wrote:
> Paul Rubin <> wrote in
> news:7xr6robovq.fsf(a)
>>John Sheehy <JPS(a)no.komm> writes:
>>>Your assistant is me, and I decide to replace the 16 containers with
>>>64 square containers, 4 of which fit in the same space as 1 of yours.
>>>I come back to you with a list of results from 64 smaller containers,
>>>instead of the 16 you asked for. The list is longer, and the total
>>>count is the same as it would be if I used your original containers,
>>>but have I created any *NOISE*?
>>Well, how to you measure how many raindrops fell in each container?
> This is just a thought experiment, not a real-world problem. Change it
> to marbles, and actually count them. Happy? The thought experiment is
> about shot noise. Read noise is another issue, but reality works in the
> favor of small pixels there, really.

In any digital counting, e.g. an A/D converter is always
+/- 1 count. If your count is 25 photons, it is 25+/- 1
plus the photons noise plus the read noise.
Counting small numbers plus read noise means smaller
pixels can NEVER do as well as large pixel when you
sum small pixels to equal the large pixels.

And we haven't even begun to discuss diffraction effects
at the edges of the pixels and microlenses. That makes the
situation worse still.


> You do understand that this is analogous to read noise, not shot noise,
> right? There really is no counting error in shot noise. Shot noise is
> not really noise; it is an accurate count of the *REAL* nature of light.
> We just don't like it in our images, because unlike the shot "noise" in
> our retinas, our perceptual system doesn't buffer it from us in our
> experience of media. Our brains only do noise reduction on what it
> thinks is direct experience of light.
> That's not a very likely scenario. Error is likely to be proportional to
> amount, at least in part, not a fixed number. But keep going...

No, digital counting is always +/- 1 count.

> Real World: My FZ50 has a read noise of 2.7 electrons at ISO 100, with a
> full-well capacity of 4800 electrons. 9 of its pixels, binned together,
> will have a read noise of 8.1 electrons, and a full count of 43,200
> photons, about the same as full-well capacity in my 20D, but the 20D has
> a read noise of about 26 electrons at ISO 100.

Where do you get these numbers? The read noise of 2.7 electrons
is lower than any other tested point and shoot or DSLR. Pretty
impressive if true. Remember that in a typical exposure, 4800 electrons
is saturation and an 18% gray card gets only about 864 electrons,
with a S/N ~ 29. The full well of 4800 at ISO 100 also pretty
impressive considering the full well is higher at lower
ISO (minimum iso = 80 => full well ~6000).
Do you have a reference?
> 43200/8.1 = 5233.33. The 20D has a full-well capacity of about 44000
> electrons. 44000/26 = 1692.31. The binned FZ50 pixels have over 3x the
> dynamic range of a 20D pixel at ISO 100 (where DR is the ratio of max
> signal to the level where 1:1 S/N lies); they cover about the same area,
> and collect about the same number of photons.

Show images that demonstrates this. Note too that cameras like the
20D at low ISO are limited by the A/D and its electronics, not true
sensor read noise. Now that 14-bit DSLRs are coming out, we'll
see the gap widening.

Then try the same thing at ISO 200, then 400, then 800.
You'll see the large pixel camera pull away quickly in image

From: John Sheehy on
Lionel <usenet(a)> wrote in

> Your read noise is going to be multplied by the number of photodiodes
> you're binning.

No. It is multiplied by the square root of that, while the signal is
multiplied by it, so the S/N increases by the square root.

Read noise is more of a problem when you have less and larger pixels.


<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS(a)no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
From: John Sheehy on
Lionel <usenet(a)> wrote in

> To include shot noise, you'd need to throw another bucket full of
> marbles into the air above your array, so they'd land randomly in your
> sensor buckets.

Actually, my thought experiment actually had shot noise, until Paul
declared that all the containers had the same number of drops (but that I
just had problems counting them).

>> Read noise is another issue, but reality works in the
>>favor of small pixels there, really.

> No, it definitely doesn't. The smaller the photodiode, the harder it
> is to accurately read the signal from it.

.... and the more pixels covering an area, the less important the accuracy
of any one pixel is. You're another one missing the forest (image/subject)
for the trees (pixels).


<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <JPS(a)no.komm>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><