Velocity Reviews - Computer Hardware Reviews

Velocity Reviews > Newsgroups > Computing > Digital Photography > Beyond Megapixels...

Reply
Thread Tools

Beyond Megapixels...

 
 
Paul H.
Guest
Posts: n/a
 
      04-29-2004

"David J. Littleboy" <(E-Mail Removed)> wrote in message
news:c6prm2$drg$(E-Mail Removed)...
<snip>
> But this is, as has been said before, incorrect: Foveon insists that they
> don't need an antialiasing filter, which is simply a lie. (The mathematics
> of sampling require that the signal be low-pass filtered before sampling:
> this is an unavoidable law of nature: it's not even physics: it's
> mathematics; not even God can revoke it.) As a result, the apparent
> sharpness comes at a cost of incorrect imaging, Moire, and other problems.
>


Not to quibble, David, but I don't think God, if He/She/It exists, routinely
revokes the laws of nature, either.

Besides, though I'm not defending Foveon in any way, I really think most
camera makers are too aggressive with anti-aliasing filtering: After all,
most of us don't routine photograph zebras or stair bannisters at 100 yards
and the blurriness caused by extreme anti-aliasing is often more annoying
than the occasional aliasing artefact might be. A switchable anti-aliasing
filter would be nice, if it could be implemented (I'm thinking of the Canon
G5's internal ND filter).

----------------------------------------------------------------------------
---------------------------
"All science is either Physics, or stamp collecting."

--Lord Kelvin.



 
Reply With Quote
 
 
 
 
Roland Karlsson
Guest
Posts: n/a
 
      04-29-2004
Don Stauffer <(E-Mail Removed)> wrote in news:(E-Mail Removed):

> The silicon CCD chip (or even CMOS, for that matter) is limited in its
> photoresponse by the basic EO properties of the material. For faster
> (more sensitive) chip we must use some different material. Silicon is a
> wonder material for its fabrication qualities, so there is little push
> to develop other semiconductor materials. Even with the increasing
> popularity of digital cameras, the image chips represent a small
> fraction of the total silicon ICs sold. In order to develop a new
> material to the point where fabrication ease and expense is near today's
> silicon would take REALLY big bucks. And, for that matter, the EO
> properties of silicon are already pretty good, so it is going to be hard
> to find a better material.
>
> We have been spoiled by photographic film. It comes closer to magic
> than semiconductors. I am serious about that. Last time I really
> looked into photo-physics of film, the film experts really were not
> fully able to explain the photo-chemical process and how it works and is
> so sensitive. Admittedly that was about twenty years ago, and they may
> have really discovered how it works by now, but at least two decades ago
> there was still a lot of art and black magic in making photo emulsions.


It is rather the oposite. Digital excels at high sensitivity, but film
gives it a good match at lower ISO. A DSLR at ISO 1600 is much better
than film at ISO 1600.


/Roland
 
Reply With Quote
 
 
 
 
Roland Karlsson
Guest
Posts: n/a
 
      04-29-2004
Fuzzy Logic <(E-Mail Removed)> wrote in
news:Xns94D996C43D171bobarcabca@198.161.157.145:

> My contention on this issue is that the average photographer does not
> need more megapixels, they need better megapixels. I would much, much
> rather own a 5 megapixel digicam with good resolution at high ISO
> sensitivities and tough shooting situations than an 8 megapixel camera
> with great resolution only in ideal situations.


This is correct.

I took some pictures in bright sunlight over the sea today. It would
be just fantastic if I could catch the tones in the reflections in
the water at the same time as the rest of the picture. More dynamic
range is the most important quest right now!

> Thankfully, some
> manufacturers have moved beyond pushing megapixels.


Yes - there have been lots of very good cameras with
few megapixels - some DSLR from Nikon comes to my mind.

> Cameras that
> utilize Foveon's X3 sensor produce smaller images, but they are much
> sharper, as red, blue and green color channels are captured in every
> photosite, as opposed to the more standard use of Bayer interpolation.


This sounds good in theory, but the Foveon implementation
is lacking.

They omit an anti alias filter. I know that som prefere
the "sharper" pictures you then get. But ... at some stage
in our evolving skill in looking at digital pictures most
of us will see that something is wrong. The sharpness is
faked. It might look good - just as a any artificial image
might look good - but it is faked, and most of us will eventually
be able to see that. All textured areas (grass, greenery, rock,
water, skin, ...) has a very unnatural look. It is not all about
fences and zebras - it is all over the picture. And - it cannot
be removed - as aliasing produces artefacts that might be larger
than the pixels.

Moreover - the filtering technique is not good. The color
sensitivity of the three layers is weird. Converting leads to
lots of noise and also faulty colors.

Moreover - you lose dynamic range by stacking the three layers
on top of each other.

Lastly - they are lying. The call it a 10 MPixel camera. They
have themselves redefined the word pixel to fit that description.

> Fujifilm is also taking things up a notch by adding a set of
> photosites just for the purpose of improving dynamic range with their
> SuperCCD IV SR sensors.


This is an interesting idea. Unfortunately - it is in the same
cameras that tilt the sensor 45 degrees. This tilting doubles
the needed Megapixels in the files without really any advantage
IMHO.

> Read the rest here:
>
> http://www.thetechlounge.com/article..._megapixels_pa
> rt_1


A typical B-grade article. Written to catch interest, but without
any real knowledge by the writer.



/Roland
 
Reply With Quote
 
Bart van der Wolf
Guest
Posts: n/a
 
      04-29-2004

"George Preddy" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed) m...
> "David J. Littleboy" <(E-Mail Removed)> wrote in message

news:<c6prm2$drg$(E-Mail Removed)>...
SNIP
> > But this is, as has been said before, incorrect: Foveon insists
> > that they don't need an antialiasing filter, which is simply a lie.
> > (The mathematics of sampling require that the signal be low-
> > pass filtered before sampling:

>
> A blur filter is needed with Bayer because of color sampling.


Wrong. ALL sampling needs low-pass fitering to reduce aliasing.

SNIP
> Foveon obviosuly has 3X as many locations to store charges than Bayer,


Wrong again, they are smaller because three have to share the same space as
a single potential well.

Bart

 
Reply With Quote
 
Philip Homburg
Guest
Posts: n/a
 
      04-29-2004
In article <(E-Mail Removed)>,
Paul H. <(E-Mail Removed)> wrote:
>Besides, though I'm not defending Foveon in any way, I really think most
>camera makers are too aggressive with anti-aliasing filtering:


There are plenty of pictures from dSLRs that show aliasing effects. I think
the D70 is quite bad in that respect, but the 1Ds can have problems too.
Of course, with a very high resolution sensor (say 20 Mpixel or more) an
anti-alias filter is less important than for a 5 Mpixel sensor. 10 Mpixel
might be a good trade-off. Anti-alias filters will reduce sharpness. For
that reason, using a sensor a higher resolution than you would need in
theory might be a good idea.



--
Everyone I've met who had any experience with the phenomenon have confirmed my
opinion that if a Ph.D. in computer science knows anything at all about
computers, it's probably pretty much an accident. -- J.D. Baldwin, in asr
 
Reply With Quote
 
Bart van der Wolf
Guest
Posts: n/a
 
      04-29-2004

"Philip Homburg" <(E-Mail Removed)> wrote in message
news:bb96v4c2a3cintmudi84fa6537@inews_id.stereo.hq .phicoh.net...
SNIP
> Anti-alias filters will reduce sharpness.


But only for the highest spatial frequencies. And the effect is more a
reduction of contrast than a blur as we know it. A lot of that contrast can
be restored by postprocessing, and now with less risk of false color
artifacts.

> For that reason, using a sensor a higher resolution than you would
> need in theory might be a good idea.


Unfortunately, as the Nikon D70 and the Kodak 14N show, aliasing artifacts
are bigger than a single pixel, so there is no really practical limit to
super sampling.

Bart

 
Reply With Quote
 
David J. Littleboy
Guest
Posts: n/a
 
      04-29-2004

"Paul H." <(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
>
> "David J. Littleboy" <(E-Mail Removed)> wrote in message
> news:c6prm2$drg$(E-Mail Removed)...
> <snip>
> > But this is, as has been said before, incorrect: Foveon insists that

they
> > don't need an antialiasing filter, which is simply a lie. (The

mathematics
> > of sampling require that the signal be low-pass filtered before

sampling:
> > this is an unavoidable law of nature: it's not even physics: it's
> > mathematics; not even God can revoke it.) As a result, the apparent
> > sharpness comes at a cost of incorrect imaging, Moire, and other

problems.
>
> Not to quibble, David, but I don't think God, if He/She/It exists,

routinely
> revokes the laws of nature, either.
>
> Besides, though I'm not defending Foveon in any way, I really think most
> camera makers are too aggressive with anti-aliasing filtering: After all,
> most of us don't routine photograph zebras or stair bannisters at 100

yards
> and the blurriness caused by extreme anti-aliasing is often more annoying
> than the occasional aliasing artefact might be. A switchable

anti-aliasing
> filter would be nice, if it could be implemented (I'm thinking of the

Canon
> G5's internal ND filter).


Actually, most dSLRs' AA filters are insufficient. The Nikon D100 is the
only dSLR with a fully adequate AA filter, and they get ranked on for soft
images. So they built the D70 with the most inadequate AA filter (other than
the Kodak dSLRs) out there, and it (occasiohnally) does horrible things with
patterns with fine detail. But makes very sharp images otherwise.

With regards to the G5. Put your camera on a tripod, set it to ISO 50,
manual focus at infinity, aperture priority mode at f/5.6 or f/8, and take
some landscape or cityscape shots with lots of detail. You will find that
the images are extremely sharp.

By the way, note that there's a basic tradeoff between sharpness and
accuracy: if you insist on pixel-per-pixel sharpness, then the accuracy of
placement of detail is lower relative to the size of the detail that you are
resolving. So the same apparent sharpness with more pixels and more AA
filtering is far more accurate than that apparent sharpness with fewer
pixels and less AA.

David J. Littleboy
Tokyo, Japan


 
Reply With Quote
 
JPS@no.komm
Guest
Posts: n/a
 
      04-30-2004
In message <(E-Mail Removed)> ,
http://www.velocityreviews.com/forums/(E-Mail Removed) (George Preddy) wrote:

>A blur filter is needed with Bayer because of color sampling. It
>prevents rainbowing at the expense of accuracy. Foveon obviously
>doesn't need one since color is not sampled then guessed, but rather
>directly measured at every photosite, same for luminance.


Wrong. Foveon samples, and all sampling requires an AA filter, unless
the optics do the job with softness, or the display is so hi-res that
individual pixels are much smaller than what can be seen. Bayer merely
requires a slightly stronger AA filter than a stacked-RGB sensor.
--

<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
John P Sheehy <(E-Mail Removed)>
><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><

 
Reply With Quote
 
Dave Haynie
Guest
Posts: n/a
 
      05-02-2004
On 28 Apr 2004 18:51:06 -0700, (E-Mail Removed) (hfs2) wrote:

>I wonder.....
>
>Will you ever 'pick your film'? Different sensors
>for different situations?


Possible, maybe. But before then, I expect you'll have realtime DSP
living between the raw sensor data and the stored JPG, especially in
consumer cameras. Imagine, for example, 12 to 16 bit image sensors,
enough extra data to effectively simulate the color curves of your
favorite films. You want Veliva, you dial up the Veliva plug-in; you
want Tri-X or a tin-type, you dial up that "virtual film". Folks with
RAW cameras could certainly do some of this today, at least where the
selected "virtual film" isn't dramatically messing with your exposure.


>Something like: screw off lens, pop out
>fuji sensor, snap in Sharp sensor. I know there's more to it
>than that, but not that much.


A standard for, basically, sensor modules (pop off the camera back,
unplug the Canon module, plug in the Sony module) is possible, but not
easy. For one, the modules would have to be smart. Today, the camera's
software is written specifically for that sensor. That level of
intelligence would have to travel with the sensor, not the camera, and
you'd need a standard API to still support things like RAW formats,
some of the film emulations I've suggested, etc. So it's not just the
CCD/CMOS there, but something else (which could be integrated on a
CMOS chips, but would require an extra part or more for a CCD).

And then there's the practical aspect: who would cooperate? Camera
companies can't usually agree on the same lens mount, much less a more
advanced plug-in sensor standard. They also are, currently, the
beneficiaries of the "cameras change like computers" phenomena, which
is not making the camera companies sad, you can be sure. They love the
idea that you might upgrade every couple years, rather than every
decade or two. The main reason you'd upgrade is for a new sensor, so
it's going to be hard to eliminate that reason. One company might try
to build their market share by offering that as a competitive
advantage (eg, probably not Canon or Sony, they don't need to convince
Minolta or Sigma users, or any of the other dozen or so companies who
aren't Canon, Sony, Kodak, Nikon, or Olympus, to switch -- let them
fight over the remaining 25% or less of the digital camera market.

So this sort of thing, maybe from a company like Pentax or KM?
Probably also not from one of the companies (Canon, Sony, Kodak)
making their own sensors -- why make it easy for 47th Street Photo to
bundle your body with the cheap knock-off sensor?


Dave Haynie | Chief Toady, Frog Pond Media Consulting
(E-Mail Removed)| Take Back Freedom! Bush no more in 2004!
"Deathbed Vigil" now on DVD! See http://www.frogpondmedia.com
 
Reply With Quote
 
Dave Haynie
Guest
Posts: n/a
 
      05-02-2004
On Thu, 29 Apr 2004 18:50:48 +0200, "Bart van der Wolf"
<(E-Mail Removed)> wrote:

>
>"George Preddy" <(E-Mail Removed)> wrote in message
>news:(E-Mail Removed) om...
>> "David J. Littleboy" <(E-Mail Removed)> wrote in message

>news:<c6prm2$drg$(E-Mail Removed)>...
>SNIP
>> > But this is, as has been said before, incorrect: Foveon insists
>> > that they don't need an antialiasing filter, which is simply a lie.
>> > (The mathematics of sampling require that the signal be low-
>> > pass filtered before sampling:


>> A blur filter is needed with Bayer because of color sampling.


>Wrong. ALL sampling needs low-pass fitering to reduce aliasing.


Of course it does. But George doesn't even begin to understand the
issues he talks about, that much is clear. I doubt he could explain
alising or imaging in digital sampling systems to save his life.
"Foveon isn't sampling", George? I'm sure the image just magically
grows in the X3 sensor, tended by faries and elves. Do us all a favor
and learn what Analog to Digital Sampling actually is, before
pretending to have half a clue (and coming up short).

>SNIP
>> Foveon obviosuly has 3X as many locations to store charges than Bayer,


>Wrong again, they are smaller because three have to share the same space as
>a single potential well.


Which is also one reason they have to be kept at such low resolutions,
versus CCDs or conventioal CMOS chips. Foveon, for example, makes a
16MPixel CMOS sensor (used by Hasselblad, at least last I checked),
but not an X3 type. Not a sensor for the professional market, for many
of the reasons stated.
Dave Haynie | Chief Toady, Frog Pond Media Consulting
(E-Mail Removed)| Take Back Freedom! Bush no more in 2004!
"Deathbed Vigil" now on DVD! See http://www.frogpondmedia.com
 
Reply With Quote
 
 
 
Reply

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Wap and beyond Frank Wireless Networking 1 06-18-2005 03:00 AM
How to get beyond a shared internet connection =?Utf-8?B?bWFyYnk=?= Wireless Networking 0 01-04-2005 03:33 AM
Re: Why is non-routeable 10.100.10.x routing far beyond my network? Peter Cisco 0 11-15-2004 08:31 AM
Any way to increase logging level beyond debugging? Forrest Cisco 0 03-05-2004 10:05 PM
VHDL: practical questions: beyond just hobbies Neil Zanella VHDL 0 11-16-2003 06:46 AM



Advertisments