Digital Camera Sensor Performane Summary updated Feb 3

[A complimentary Cc of this posting was sent to
ejmartin
<ejm_60657.RemoveThis@yahoo.com>], who wrote in article <4a5a3cfc-4155-4322-b4da-6538e76d92e5.RemoveThis@e23g2000prf.googlegroups.com>:
> > Of course, this measure is just what Rodger's left pinky wanted.
> > However, note that until some reasonable investigations of psychology
> > of vision are performed/found, we are confined to guesswork anyway.
> > So I, personally, do not have a large problem with this particular
> > part of what Roger added.

> Well, I would have said that two a priori independent quantities
> should be reported as independent quantities rather than multiplied
> together. But that's just me.

If possible, yes. But for some purposes a 2-dimensional space of
parameters is not feasible; 1-dimensional metric is a compromize (like
mapping all the values to a monetary scale with some coefficients).

> > Did not you see my posts about relation of dynamic range and
> > resolution?
>
> Um, no. I left this place for a long time because the SNR got too
> low.

Well, I more or less summed it up in what followed. Let me know if
you want the older discussions (might have high Noise/Signal ratios,
as typical for this newsgroup).

> Well, yes, all valid information but probably not helpful. The art of
> data interpretation is to take the mass of data and distill from it a
> small number of parameters that distill the essence of their meaning.
> Just listing all the data in a repackaged form is not particularly
> useful.

Hmm, note that I was in no way listing ALL the data. In particular, I
kept no information about signal strength alone, and no information
about the noise alone. Only the ratio is kept.

> particular there are going to be degeneracies -- is a camera with good
> SNR and poor resolution better or worse than one with poor SNR and
> good resolution?

Since SNR and resolution are coupled, you can immediate see the answer
using "my" curves. Plot the SNR vs spacial frequency curves; if one
is on top of the other, the answer is clear.

In general, this might even be reduced to one number: take the target
resolution, and consider the spacial frequency about 70% of the
Nyquist (where - usually! - the most visually important information is
contained). Calculate the S/N ratio on THIS spacial frequency.

> That depends on individual preferences regarding
> noise and resolution; I know some people who just abhor noise, to the
> point that they smear fine scale detail to pot in order to get rid of
> it, and others who do little or no NR in order to maximize detail.

Note that postprocessing ("smearing") rarely changes S/N ratio at a
particular spacial frequency. Hence the metric I use will cater to
both camps.

Of course, there are cases when there is some flexibility with the
target resolution. THEN studying the psychology of vision is not
avoidable.

Yours,
Ilya

On Sat, 16 Feb 2008 05:53:12 +0000 (UTC), Ilya Zakharevich
<nospam-abuse RemoveThis @ilyaz.org> wrote in <fp5to8$qcf$1@agate.berkeley.edu>:

>Well, I more or less summed it up in what followed. Let me know if
>you want the older discussions (might have high Noise/Signal ratios,
>as typical for this newsgroup).

I'm guessing you actually meant low signal:noise ratio.

--
Best regards,
John Navas
Panasonic DMC-FZ8 (and several others)