>From - Sat Mar 02 00:57:24 2024
Received: by 10.236.187.39 with SMTP id x27mr1805856yhm.4.1336590797563;
Wed, 09 May 2012 12:13:17 -0700 (PDT)
X-BeenThere: tscm-l2006_at_googlegroups.com
Received: by 10.236.114.50 with SMTP id b38ls9831239yhh.1.gmail; Wed, 09 May
2012 12:13:12 -0700 (PDT)
Received: by 10.236.73.131 with SMTP id v3mr2756784yhd.6.1336590792884;
Wed, 09 May 2012 12:13:12 -0700 (PDT)
Received: by 10.236.73.131 with SMTP id v3mr2756782yhd.6.1336590792871;
Wed, 09 May 2012 12:13:12 -0700 (PDT)
Return-Path: <eric.s..._at_gmail.com>
Received: from mail-yw0-f46.google.com (mail-yw0-f46.google.com [209.85.213.46])
by gmr-mx.google.com with ESMTPS id y36si3689030yhg.2.2012.05.09.12.13.12
(version=TLSv1/SSLv3 cipher=OTHER);
Wed, 09 May 2012 12:13:12 -0700 (PDT)
Received-SPF: pass (google.com: domain of eric.s..._at_gmail.com designates 209.85.213.46 as permitted sender) client-ip=209.85.213.46;
Authentication-Results: gmr-mx.google.com; spf=pass (google.com: domain of eric.s..._at_gmail.com designates 209.85.213.46 as permitted sender) smtp.mail=eric.s..._at_gmail.com; dkim=pass head..._at_gmail.com
Received: by yhmm54 with SMTP id m54so625728yhm.19
for <tscm-..._at_googlegroups.com>; Wed, 09 May 2012 12:13:12 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
d=gmail.com; s=20120113;
h=mime-version:in-reply-to:references:date:message-id:subject:from:to
:content-type:content-transfer-encoding;
bh=WphDK+q9U4umt6fNITHLSCpFa4VzCk2d7eao5qe29o4=;
b=m5HxGn4BIOVkpUP3Yr1ZtUVdZhOcvD9kKhlf+u1Rxr0XrjKXEqTGCCz+5xs+5C8hl5
YWs4bg8DN9uryeH3Iu2xFYGhesplYGn9HX+6NqvrmJmapuTBjgIRDj6DQbodVLTVLkZX
k0CIZzl1nE+jzRFnR1a+DeNePWo2oeGykYDxszFXfhN7SWoxfEuuUbd2mdQjH7/1jWX/
mGkN4jTzSl7pryu7HvmdvDer0HPNR9dPZPEvm8pDEG6K9fhi338zwg/9DnKYWwr7JS8y
uYVXJdZfMJW6DMbJ8qZRVUkgXHfam33tnu/JyjzmZYefDCtC4qmtXK7Scu0tKD52fjNI
mGPA==
MIME-Version: 1.0
Received: by 10.60.7.196 with SMTP id l4mr1941373oea.8.1336590788929; Wed, 09
May 2012 12:13:08 -0700 (PDT)
Received: by 10.60.46.165 with HTTP; Wed, 9 May 2012 12:13:08 -0700 (PDT)
In-Reply-To: <4FA8A249.5040801_at_tscm.com>
References: <4FA8A249.5040801_at_tscm.com>
Date: Wed, 9 May 2012 21:13:08 +0200
Message-ID: <CA+HM+hH0jqSD5r1GMzWrygMMm_t-mmbW0vhhH-kvaKmF99gZNA_at_mail.gmail.com>
Subject: Re: [TSCM-L] {6130} Anatomy of the Human Eye, Evolving to Digital Cameras
From: Eric Schmiedl <eric.s..._at_gmail.com>
To: tscm-l2006_at_googlegroups.com
Content-Type: text/plain; charset=windows-1252
Content-Transfer-Encoding: quoted-printable
Regarding color sensitivity, keep in mind that camera sensors (CCDs at
least, and I believe CMOS as well) are much more sensitive to longer
wavelengths than shorter. If you white balance a camera indoors,
you'll see more noise than you will if you white balance the camera
outdoors. (and if you don't white balance, you'll get a yellow
picture)
Also, digital cameras by and large use Bayer pattern filters, which
means they have twice as many green-sensitive pixels as they do red-
and blue- sensitive pixels. This is because the human eye tends to get
most of the detail information in a scene from greenish colors. That
means extra resolution in red or blue would be wasted -- try it
yourself, bathe a room in red or blue and you'll see less detail.
Different size pixels has also been done, by Fuji for their S5 Pro
camera and an accompanying (very rare) medium format digital back. But
those were for a different purpose -- the Fuji varied size (without
respect to color?), in order to create a chip that had more- and
less-sensitive pixels and which, combined together, could therefore
see a wider range from light to dark.
On Tue, May 8, 2012 at 6:34 AM, James M. Atkinson <jm..._at_tscm.com> wrote:
> =93The Photographic Eye,=94 written by Allan Weitz is an interesting arti=
cle
> about how the human eye and the modern camera have several similar
> attributes. What I found quite interesting in the article was how the hum=
an
> eye constructs the entire scene through narrow (.5 degree) snapshots of
> images. With some simple math, I was able to determine that each image th=
at
> we observe is actually 240-280 smaller images (120-140 degrees combined
> field of view), each involving 120 million panchromatic pixels, and 7
> million color pixels. This would reflect that to capture a single image t=
he
> eye collects 3.36 billion panchromatic pixels, and 196 million color pixe=
ls
> (and the 576 MP mentioned in the article may be a little incorrect). The
> poetic message in the rods versus cones comparison could be taken to mean
> that a black and white photograph conveys 17 times more powerful a messag=
e
> then one in color.
>
> The article by Weitz intrigued me further as I had recalled from my Anato=
my
> and Physiology courses that the cones do not respond evenly to color in t=
hat
> there are more red responsive cones then green, and more green then blue.
>
> After reading the Weitz article an issue then came to mind (or actually m=
ore
> of a hypothesis of mine) that modern digital cameras maintain an even
> balance of red, green, and blue pixels in their imaging sensors, when the=
y
> actually should mimic the cone densities of the human eye and be more bia=
sed
> to one color then another for optimal efficiency and fidelity. To this en=
d a
> modern digital camera should have over 70% of the pixels dedicated to red=
,
> 23% to green, and 7% to blue, or the inverse of these numbers depending o=
n
> how you look at it.
>
> The digital camera image sensor should also match the cone density patter=
ns
> of the human eye, which is a task that is actually far easier then it
> initial seems, as the sensors can be etched in virtually any color
> combination or density at the factory. My hypothesis is that by matching =
the
> cone density of the human eye to the position or patterns of certain pixe=
ls
> on the imaging sensor of a digital sensor that the camera can be made to
> =93see=94 more naturally, or more accurately the camera sensor and the ey=
e would
> see on a 1:1 basis. By collecting more pixels in the spectrum then the ey=
e
> can actually see, and not collecting pixels the eye would =93waste=94 the
> effective resolution of the digital camera would be dramatically increase=
d
> by doing nothing more then re-ordering the pixels, but not actually
> increasing their number.
>
> To look at this a different way, why should the camera sensor collect 7
> million blue pixels, 7 million green pixels, and 7 million red pixels? Th=
is
> symmetry of color pixels really is not needed, and an asymmetrical approa=
ch
> to sensor design may allow cameras to actually see far better then has
> previously been possible.
>
> To take this a step beyond merely using asymmetrical sensor design, as th=
e
> rods and cones of the human eye are both mechanically, electrically, and
> chemically different from one another and yet part of the same matrix the
> camera sensor could also be modified to increase the diameter of the
> monochromatic elements, and reduce the size of the color elements. This
> would allow the monochromatic elements within the same sensor to gain
> greater sensitivities and thus preserve image detail, while the color
> elements would fill in the gaps and mimic the performance of the human
> retina. Optionally, one sensor could be dedicated to color imagery, and a
> second to monochromatic, and perhaps additional sensors for alternate
> wavelengths similar to what is done with reconnaissance satellites.
>
> With silicon wafers now reaching 18-inches in diameter, and the new back =
lit
> digital camera sensors well exceeding 50 million pixels per square inch
> (i.e.: Kodak KAF-5K series) it will be possible for imaging sensors to re=
ach
> into the hundreds of millions of pixels within a few years time using onl=
y a
> single chip.
>
> While it may be costly to do so, it is currently within reason for a sing=
le
> silicon wafer to be etched into a single digital sensor involving 1.12+
> billion pixels; divided into whatever spectrum the designers may wish. Wh=
ile
> an 18-inch wafer (Taiwan Semiconductor) and support circuits may only fit
> onto the back of a 11x14 or larger camera the results would be breathtaki=
ng,
> and may well exceed the abilities of the human eye, or for that matter an=
y
> other creature on Earth.
>
> --
> James M. Atkinson
> President, Scientist and Sr. Engineer
> "Leonardo da Vinci of Bug Sweeps and Spy Hunting"
> Granite Island Group
> jm..._at_tscm.com
> http://www.tscm.com/
> (978) 546-3803
>
Received on Sat Mar 02 2024 - 00:57:24 CST