Fun with Bayer interpolation

[Rinus Borgsteede]

Hi Ron,

I do not think that the jacket patterns are a Bayer related problem. It looks like a typical Moire pattern I see in photoshop when my image information conflicts with my screen resolution. I see the same Moire patterns on the TV. I do have to admit that it is possible to get this from a digital file but it is an extrememe rarety. Improper bayer interpolation occurs more often in hair where color is mixed with fine lines and specul;ar non color (or over saturated colors) highlights.
Rinus
Sure like your FAQ webpage

 

[Erik Magnuson]

I am a bit curious about why you keep harping on this,

I don't introduce the subject. I only offer a dissenting point of view at times when I feel that you are overstating the potential benefits.

it's something that I've never pushed as a reason to be excited about
this technology. [...] it will really be cheaper to manufacture $100

Something is lost here. What is "it's"? You were the one to bring up the $100 camera:

http://www.dpreview.com/forums/read.asp?forum=1000&message=2326699

Saving $10 in production costs on a $100 camera is significant and this is their point

and

http://www.dpreview.com/forums/read.asp?forum=1000&message=2326739

For their low end sensor (1 MP), it's not that much bandwidth

and finally most recently

http://www.dpreview.com/forums/read.asp?forum=1000&message=2333587

I've asserted that for a cheap camera eliminating this step would give a cost savings.

What I've mainly done is dispute your contention in these posts that somehow bayer interpolation circuitry is expensive and that the larger bandwidth requirements of the Foveon chip will be cheap. I've offered several examples based on what current products do and cost. You just keep saying "it's obvious".

And you're obsessing about sub $100 cameras because...

You keep bringing them up.

I'm still waiting for you to tell us a serious issue.

• Only ISO 100 at full resolution. Higher ISO only at lower resolutions. (read the F10 spec sheet).

• Higher buffering, I/O and storage requirements

• Potentially slow workflow (e.g. the infamous 12s on a P4 for the high resolution images)

Any given photographer may consider these to be vital or trivial potential issues.

You can continue to believe that it's
possible to always guess right, along with all of the
inconsistencies that this implies. I really don't care what you
believe any more.

Eh? Exactly where have I ever said anything about interpolation always being right? I must have missed that post I made. If this is what you are reading then no wonder you think I'm a nut. The most extreme position I've taken is that for a given cost point interpolated sensors may produce "better" images in some situations.

This thread was about Bayer interpolation and what can happen when you do it.

I merely offered dissenting opinions on the cost of Bayer interpolation vs the alternatives where I've felt that your claims have been inadequately supported. Perhaps the problem is that I keep talking about current and announced cameras and parts while you are talking about a hypothetical future. My crystal ball is too cloudy for that.--ErikFree Windows JPEG comment editor http://home.cfl.rr.com/maderik/edjpgcom

 

[Erik Magnuson]

But in every
aspect of life there are always some who will resist change until
both sides are blue on the face.

Ugh. Change bad. Must fight change. I guess I need to start the "Bayer Interpolation Luddite Society." We'll hold secret meetings and wear those red-and-blue 3-d glasses so that we can interpolate all the time!

It could be that they don't feel
comfortable admiting past unawareness about problems associated
with the old... just a guess.

Well, from my side it looks like some people are so entranced with the new that they cannot see that there might be some problems there too.

--ErikFree Windows JPEG comment editor http://home.cfl.rr.com/maderik/edjpgcom

 

[Erik Magnuson]

I do not think that the jacket patterns are a Bayer related
problem. It looks like a typical Moire pattern I see in photoshop

If it were just a simple moire problem, you would expect it to change when you view the photo at different sizes and resolutions. But it doesn't. The yellow/blue stripes are pretty distinctive in bayer interpolation artifacts. This is the ugliest one I've ever seen.

--ErikFree Windows JPEG comment editor http://home.cfl.rr.com/maderik/edjpgcom

 

[Ron Parr]

You or Karl started complaining about Foveon's claims that their X3 approach is cheaper, so I responded to this. Somehow you guys got it in your heads that they were talking about SLRs when they said this. Now whenever there's a discussion about anything else, you drag out the cost issue - either to extend the misconception that their claims applied to SLRs or to splt hairs about costs for $100 cameras.

Regarding the "always guessing right" thing. If you think that Bayer interpolation doesn't degrade every day day images, then this implies that you think it always guesses right or close to right.

--Ron ParrFAQ: http://www.cs.duke.edu/~parr/photography/faq.htmlGallery: http://www.pbase.com/parr/

 

[Richard C]

Ron,

Thanks for the link to the PDF about Bayer Demosaicking (the picture link didn't work, but I can guess that it's a fine check pattern jacket - that seems to be the bane of digicams).

The process of demosaicking is fascinating - I'd often wondered how they avoided false colour fringes at edges. If you look closely at the subtractive method used to do that, you can see that there is an assumption that the G information is all of the luminance info, effectively you discard all the luminance info from the R & B channels at that stage, in order to eliminate colour fringes. That's undoubtedly a good deal, as colour fringing on edges would be very visually disturbing. Very early on in the history of CCD video they used a few 2 chip cameras, one monochrome and one with a mosaic filter. I can see now how they worked, having accurate luma for all pixels would allow the elimination of edge fringing without throwing out half the resolution. A 2 CCD unit would have the same resolution as a 3 CCD block and only slightly poorer colour.

It all seems to justify Foveon's claim to resolution equivalent to a Bayer chip with twice the number of pixel locations. Phil's samples point that way too. The other big advantage is that, with no spatial offset (at least in 2 dimensions) between the different colour sensors, another source of colour artifacting is eliminated.

Really, none of this should be a surprise, 3 CCD video cameras have comfortably 40% better resolving power than single chip units of the same pixel count (and my TRV900 has its pixel count specified as "470,000 X3 CCD")

There are many misconceptions about Bayer interpolation. I dispel
many of these in my FAQ (listed below). However, it's worth
mentioning a few here:

1. Bayer interpolation gets the luminance (the B&W part) of the
image right, and just interpolates the color. This is wrong.
Every pixel interpolates 2 out of 3 of the RGB values. Errors in
any one of these will cause errors in luminance (though errors in
green are worse). If you figure out how much of the luminance
signal a Bayer sensor is actually getting (by projecting into the
luminance dimension) you find that the sensor is capturing about
40% of the luminance information.

2. The types of patterns that cause trouble for Bayer sensors
don't occur in real life. Would that this were so! Here's an
amazing example I came across recently on pbase, with my thanks and
apologies to the original photographer (D30, I think):

http://www.pbase.com/image/1121096

Admittedly, this is a worst case kind if example, but it is real
and there's no reason to think that our images aren't riddled with
lots of smaller patches of errors like this in areas of fine detail
as well as lots of subtler errors that would only be noticeable in
comparison to the correct image.

If you want to see some examples of what's going on with Bayer
interpolation (and an education on interpolation methods), I
suggest you download this paper:

http://www4.ncsu.edu:8030/~rramana/Research/demosaicking4.pdf

Even if you don't follow the math, skip forward to the figures,
where you can see some wonderful examples of the types of errors
Bayer interpolation can make. Figure 18 is particularly striking.

--
Ron Parr
FAQ: http://www.cs.duke.edu/~parr/photography/faq.html
Gallery: http://www.pbase.com/parr/

--Richard C. South Australia

 

[Ron Parr]

pbase seems to be working again if you want to try the jacket link.

I agree completely about the video cameras. Here's a link to a page that has resolution tests for video cameras. The 340k x 3 TRV-900 does as well or better than 1+MP models.

http://www4.big.or.jp/~a_haru/index.html

To be fair, however, we should consider that some of the improvement could be from a better lens...

--Ron ParrFAQ: http://www.cs.duke.edu/~parr/photography/faq.htmlGallery: http://www.pbase.com/parr/

 

[Erik Magnuson]

Now whenever there's a discussion about anything else, you
drag out the cost issue - either to extend the misconception that
their claims applied to SLRs or to splt hairs about costs for $100
cameras.

All I ever said is that there is currently no basis to assert Foveon will be cheaper. When you say, "oh, they can just eliminate this part" without accounting for other parts that are likely to be required for a comparably performing camera, I think you are being unrealistic.

Regarding the "always guessing right" thing. If you think that
Bayer interpolation doesn't degrade every day day images, then this
implies that you think it always guesses right or close to right.

There is a huge difference between "always" and "close to". This is the type of hyperbole that I've been objecting to. If you think that these are the same, then I guess we have too little in common to ever engage in reasoned discourse.--ErikFree Windows JPEG comment editor http://home.cfl.rr.com/maderik/edjpgcom

 

[Ron Parr]

There is a huge difference between "always" and "close to". This
is the type of hyperbole that I've been objecting to. If you think
that these are the same, then I guess we have too little in common
to ever engage in reasoned discourse.

I can't say that I agree with the antecedent of the last sentence, but I'm afraid that the consequent may be true.

--Ron ParrFAQ: http://www.cs.duke.edu/~parr/photography/faq.htmlGallery: http://www.pbase.com/parr/

 

[Michael w]

Not true. And not only might, we know there WILL be problems with the new as well - it's inevitable. But I clearly see the differences between "fundamental" problems vs. "implementation" problems. Remember the new Foveons chips are still the first generation of their kind.

Well, from my side it looks like some people are so entranced with
the new that they cannot see that there might be some problems
there too.

--Michael

 

[Karl Guttag]

Steven,

I think you are greaty mis-characterizing what some of us are saying. We want better images. We care also about cost. We are just trying to point out that it is just not clear at this point how significant X3 will be.

IF X3 produces the same signal to noise, has good color separation, has at or near the same resolution and costs less it will be a winner. I think everyone would agree that Bayer is not an ideal solution, but neither is it SO horrible that it is critical limiting factor in keeping people from getting good images.

As it stands right now, we don't really know about the noise at high ISO (I think in the long haul this is where X3 will live or die). Right now the 6MP Bayers are ahead in terms of getting to market than the 3.4 MP X3 (which are probably "comparable" in resolution). The frame rate is lower with the X3. The memory and storage buffer requirements are higher with the X3. Based on my 24 years of semiconductor experience, the X3 almost certainly cost more to produce TODAY (due to being run by a lower volume producer, National, and due to being a new and different process). It is just not a clear win today that changes everything.

There are a lot of factors that affect both image quality and cost and MOST of these factors have nothing to do with X3 or not, this is particularly true if you are comparing a 3.4MP X3 to a 6MP bayer. The Lens quality, exposure control, auto-focus system, frame rate, burst buffer length, and a lot of other factors are more important to more people.

I have been shooting with a D30 since it was first available. I have lost a lot of shots to the AF system, but in 15,000 shots I can't remember loosing one to a Bayer problem. Even a 3MP Bayer'ed D30 can "see" the differences in high and low quality lenses (particularly when shot wide open).

Karl

It amazes me the number of people that do not want a better image
sensor to be available. while the Bayer pattern is good it is
easily fooled and causes a serious performance hit on sharpness and
contrast. It forces photos to be sharpened because the output of
even the best interpolations are soft. Look at the unsharpened
outputs of cameras like the D30 and 1D. They are SOFT. This s a
direct
result of the Bayer pattern. This is really true with black on red or
black on blue.

The X3 (While not 3 times better) will make any image better. They
will
be sharper without the need to sharpen. I hate the artifacts left
behind
by sharpening.

Don't get me wrong, my G1 always surprises me and me 1D blows it out
of the water. But they would both be better with a X3 based even
sensor at
a lower resolution.

The perfect camera:
EOS 1D, 6-8 MPixel X3 sensor, 1.3-> 1.0 multiplier, ISO to 3200.

In time...

--Karl

 

[Rinus Borgsteede]

I do not think that the jacket patterns are a Bayer related
problem. It looks like a typical Moire pattern I see in photoshop

If it were just a simple moire problem, you would expect it to
change when you view the photo at different sizes and resolutions.
But it doesn't. The yellow/blue stripes are pretty distinctive in
bayer interpolation artifacts. This is the ugliest one I've ever
seen.

If a photo is interpolated and saved to a resolution that causes Moire with the screen resolution, it becomes a permanent part of the image. At that point it will always show the Moire.

Point in case is that this image you are looking at is a fixed file and has this pattern in any screen resolution. Reproducing it will still show the pattern. As a photo engraver, I was always concerned with the angle of the halftone screen in reproduction and if the angle could not eliminate the Moire, we changed the dpi of the screen. Once you interpolate to the wrong ppi, the moire is fixed.

This image is not the original file, it is a rather small jpeg that obviously is a reduced version of the original. There may not have been any Moire at all. Nice try to use this as a sample of bayer pattern. Very convincing to other people but experts.

I do not want to understate the problems with bayer patterning because they are a problem. Foveon will solve that.

My next camera will again be the bayer pattern CCD and after all these years, I've never seen a bayer pattern that could not be fixed easely.
Rinus

 

[Ron Parr]

Of course you're right that the picture could have sufered some indignities in postprocessing, but the picture is 2160x1440, which is exactly full size for a D30 shot.

--Ron ParrFAQ: http://www.cs.duke.edu/~parr/photography/faq.htmlGallery: http://www.pbase.com/parr/

 

[Steven Noyes]

Steven,

I think you are greaty mis-characterizing what some of us are
saying. We want better images. We care also about cost. We are
just trying to point out that it is just not clear at this point
how significant X3 will be.

From the samples that have been posted on Phil's site, I think that it
is clear that it is very significant.

IF X3 produces the same signal to noise, has good color separation,
has at or near the same resolution and costs less it will be a
winner. I think everyone would agree that Bayer is not an ideal
solution, but neither is it SO horrible that it is critical
limiting factor in keeping people from getting good images.

As it stands right now, we don't really know about the noise at
high ISO (I think in the long haul this is where X3 will live or
die).

Personally, I don't think CCD or CMOS do that well. I decided to
limit the 1D to 400 ISO. Anything above is just to grainy. The D30
images look like I would limit to 200.

Right now the 6MP Bayers are ahead in terms of getting to
market than the 3.4 MP X3 (which are probably "comparable" in
resolution). The frame rate is lower with the X3. The memory
and storage buffer requirements are higher with the X3.

??? This is only true for RAW. Because it has 3 times the information.
For JPEG it is actually about 1/2 for the same detail.

Based on
my 24 years of semiconductor experience, the X3 almost certainly
cost more to produce TODAY (due to being run by a lower volume
producer, National, and due to being a new and different process).
It is just not a clear win today that changes everything.

I agree that it is not a clear winer at this point in time but it has
so much more room to improve on than technologies that have been used
for 3 decades.

There are a lot of factors that affect both image quality and cost
and MOST of these factors have nothing to do with X3 or not, this
is particularly true if you are comparing a 3.4MP X3 to a 6MP
bayer. The Lens quality, exposure control, auto-focus system,
frame rate, burst buffer length, and a lot of other factors are
more important to more people.

I have been shooting with a D30 since it was first available. I
have lost a lot of shots to the AF system, but in 15,000 shots I
can't remember loosing one to a Bayer problem.

Take a look at the resolution chart you posted. If you think that this is
wonderful, this is why you never have an issue with Bayer interpolation.
The resolution chart is full of color and edge errors. Get a chart that
is Blue on Black or Blue on Red. Better yet, create a RAW data file that
simulates this and do the conversion. It is really ugly.

Steven

 

[Tau]

Steven,

I think you are greaty mis-characterizing what some of us are
saying. We want better images. We care also about cost. We are
just trying to point out that it is just not clear at this point
how significant X3 will be.

From the samples that have been posted on Phil's site, I think that it
is clear that it is very significant.

IF X3 produces the same signal to noise, has good color separation,
has at or near the same resolution and costs less it will be a
winner. I think everyone would agree that Bayer is not an ideal
solution, but neither is it SO horrible that it is critical
limiting factor in keeping people from getting good images.

As it stands right now, we don't really know about the noise at
high ISO (I think in the long haul this is where X3 will live or
die).

Personally, I don't think CCD or CMOS do that well. I decided to
limit the 1D to 400 ISO. Anything above is just to grainy. The D30
images look like I would limit to 200.

Right now the 6MP Bayers are ahead in terms of getting to
market than the 3.4 MP X3 (which are probably "comparable" in
resolution). The frame rate is lower with the X3. The memory
and storage buffer requirements are higher with the X3.

??? This is only true for RAW. Because it has 3 times the
information.

What else does the SD9 shoot than RAW?

For JPEG it is actually about 1/2 for the same detail.

If it did shoot in JPEG, I doubt the size would be 1/2 for the same detail.

The X3 technology sees more real detail, but pictures are cleaner from artifacts. That would be about the same or maybe 90% of the size?

Jack.

Based on
my 24 years of semiconductor experience, the X3 almost certainly
cost more to produce TODAY (due to being run by a lower volume
producer, National, and due to being a new and different process).
It is just not a clear win today that changes everything.

I agree that it is not a clear winer at this point in time but it has
so much more room to improve on than technologies that have been used
for 3 decades.

There are a lot of factors that affect both image quality and cost
and MOST of these factors have nothing to do with X3 or not, this
is particularly true if you are comparing a 3.4MP X3 to a 6MP
bayer. The Lens quality, exposure control, auto-focus system,
frame rate, burst buffer length, and a lot of other factors are
more important to more people.

I have been shooting with a D30 since it was first available. I
have lost a lot of shots to the AF system, but in 15,000 shots I
can't remember loosing one to a Bayer problem.

Take a look at the resolution chart you posted. If you think that
this is
wonderful, this is why you never have an issue with Bayer
interpolation.
The resolution chart is full of color and edge errors. Get a chart
that
is Blue on Black or Blue on Red. Better yet, create a RAW data
file that
simulates this and do the conversion. It is really ugly.

Steven

--The significant problems we face can not be solved at the same level of thinking we were at when we created them.(A.E.)

 

[budhines]

Steven you are correct, but as my grandmother used to say, you are preaching to deaf ears. Bud

Steven,

I think you are greaty mis-characterizing what some of us are
saying. We want better images. We care also about cost. We are
just trying to point out that it is just not clear at this point
how significant X3 will be.

From the samples that have been posted on Phil's site, I think that it
is clear that it is very significant.

IF X3 produces the same signal to noise, has good color separation,
has at or near the same resolution and costs less it will be a
winner. I think everyone would agree that Bayer is not an ideal
solution, but neither is it SO horrible that it is critical
limiting factor in keeping people from getting good images.

As it stands right now, we don't really know about the noise at
high ISO (I think in the long haul this is where X3 will live or
die).

Personally, I don't think CCD or CMOS do that well. I decided to
limit the 1D to 400 ISO. Anything above is just to grainy. The D30
images look like I would limit to 200.

Right now the 6MP Bayers are ahead in terms of getting to
market than the 3.4 MP X3 (which are probably "comparable" in
resolution). The frame rate is lower with the X3. The memory
and storage buffer requirements are higher with the X3.

??? This is only true for RAW. Because it has 3 times the
information.
For JPEG it is actually about 1/2 for the same detail.

Based on
my 24 years of semiconductor experience, the X3 almost certainly
cost more to produce TODAY (due to being run by a lower volume
producer, National, and due to being a new and different process).
It is just not a clear win today that changes everything.

I agree that it is not a clear winer at this point in time but it has
so much more room to improve on than technologies that have been used
for 3 decades.

There are a lot of factors that affect both image quality and cost
and MOST of these factors have nothing to do with X3 or not, this
is particularly true if you are comparing a 3.4MP X3 to a 6MP
bayer. The Lens quality, exposure control, auto-focus system,
frame rate, burst buffer length, and a lot of other factors are
more important to more people.

I have been shooting with a D30 since it was first available. I
have lost a lot of shots to the AF system, but in 15,000 shots I
can't remember loosing one to a Bayer problem.

Take a look at the resolution chart you posted. If you think that
this is
wonderful, this is why you never have an issue with Bayer
interpolation.
The resolution chart is full of color and edge errors. Get a chart
that
is Blue on Black or Blue on Red. Better yet, create a RAW data
file that
simulates this and do the conversion. It is really ugly.

Steven

 

[David SL]

I've never asserted that it's a performance problem. I've
asserted that for a cheap camera eliminating this step would give a
cost savings. This should be obvious to you.

No, it's not obvious to me:

1. The same "cost savings" is available for Bayer CCD cameras and
it's not being done even on $100 retail cameras (Well, the Dj-1000
and some of the Agfas did it.) So I don't see that this is as big
a savings as you make it out to be.

2. The RAW file output for the F10 Foveon sensor would be about 3.7
x 12 bits / 8 / 2 = ~ 2.8MB. A normal mode JPEG from a current 2
MP camera is 400 kb. That's a 7-1 difference (or pick your own
numbers). To achieve comparable shot-to-shot time, this camera is
going to have to have a D30 class I/O subsystem. You seem to think
that's a trivial cost OR that it's not needed and a 4-7x write time
will be acceptable. I disagree here as well.

Erik, this is exactly what I'd been telling Ron but he has a different pov that I think is based in not having designed these sorts of systems. He's basing much of his view on analysis on the computational issues of the processing software (his area of expertise) and neglecting the issues with handling all that raw data by the in camera electronics I/O. (Which an engineer would know more about.) I am curious what is your profession ? I happen to be an electrical engineer having concentrated on computer engineering and digital signal processing. I wonder if you speak from the same background as I, that would shed some light as to why Ron sees things differently.

So, is the problem that you don't understand why a manufacturer
would want to put one less chip or a cheaper chip in a sub $100
product?

No, I think that the chips they will have to add to handle the
additional buffering and I/O will more than balance out the cost of
the interpolation processing chips.

I think that there's an implicit assumption that after production
stabilizes the cost per chip should be no more than a Bayer pattern
chip for equivalent resolution.

There is both and IF and a WHEN component to this assumption.

1. There are some real issues with Bayer interpolation that many
people currently accept without really thinking about it.

Most people in the sub$100 class could care less about these
issues. They just don't show up in 4x6 prints or screen resolutions
enough to matter. (I predict Foveon sensors will be preferred by
landscape and studio photographers.)

2. Attempting to address these issues is a good idea.

Yes, but what if you are just trading one set of issues for another?

3. Foveon's claims to be capturing 3X as much data as a Bayer
interpolated sensor are not ridiculous and have serious potential
to translate into visible improvements in pictures.

Of course the x3 has potential. I personally hope it's all it's
cracked up to be. But:

1. I don't believe the Bayer CCD issues are as significant as you
make them out.

2. I think there are some possibly significant issues with X3 that
may limit it's competitiveness. (Low ISO and larger file sizes.)

The only real information we have is from Foveon who, like anyone
selling something, is only presenting the positive side of their
product. What are the downsides? Can we live with them? Can they be
fixed? Will the potential be realized in a useful form? If yes,
when?
--
Erik
Free Windows JPEG comment editor
http://home.cfl.rr.com/maderik/edjpgcom

--DSL

 

[David SL]

Hi Ron,
I do not think that the jacket patterns are a Bayer related
problem. It looks like a typical Moire pattern I see in photoshop
when my image information conflicts with my screen resolution. I
see the same Moire patterns on the TV.

Rinus,

Actually, what defines the moire patterns caused by bayer pattern filters is the finite sampling of continuous colors in the scene, this quantization by the filters over the photosensors leads to moire when the nyquist frequency of the filter patterns is reached. The moire that you see when you view various images at particular sizes and resolutions on a monitor in photoshop or on patterned detail in images on tv is the same root cause.

The image that you see both on a monitor and a tv is refreshed at a particular frequency. On monitors its variable depending on resolution and on tv's its usually fixed at 30frame interlaced (60hz) (for a standard NTSC tv not accounting for hi res or hdtv models) that is half the screen lines of a standard tv are drawn in 1/30s and the alternate lines are drawn on the other 1/30s thanks to the persistence of human vision this has the dual advantage of reducing eye strain and not requiring a highier refresh rate for a continuous full screen draw. Just how the drawn lines vary in color is explained next.

Each point on the screen is illuminated through a mesh located on the inner surface of the CRT, the electromagnetic guns at the rear of the CRT concentrate the electron beams on a pattern of photosensitive dots or lines(as in trinitron tv's) of R, B and G sensitive phosphors through the mesh of holes that have a repeated spacial pattern very similar to a bayer pattern filter.(the pattern not the color filtering aspect, the colors come from the phosphor dots or lines themselves when they are struck by a beam)

If you were to look through the mesh on the innerside of a CRT it would look something like this (non trinitron):



Note the grey triangles mark the position of the triad (It roughly defines where the hole in the shadow mask is located for each group of 3 phosphors) You'll also notice that CRT's are not corrected for the enhanced sensitivity of the human eye to green.(phosphor dots sensitive to each color are about equal in number) This is one source of color inaccuracy between reality and video of reality displayed on CRT's.

Here's another image with a 3d look at the phosphors on the screen, the mesh and the beams shining through it:



The electron beam varies in intensity and shines through holes of the mesh (called a shadow mask) as the beams scan the screen and combine with different intensities for the other two beams illuminating their respective phosphor dot in the group on the screen causes the phosphors to phosphores..(give off light of its characteristic frequency) the mixing of the light creates a particular color which you see on the other side of the screen as a bright area.

Thus, you should now see how both monitors and tv signals themselves can introduce moire patterns when the sampling interval of displayed image detail is less than the nyquist frequency of the phosphor pattern on the inside of your CRT (tv and monitor) this is why you see the moire. Just like bayer pattern filters in digicams, different technology same problem. In fact, you'll notice that you get rid of the moire when the spacial frequency in the viewed image on screen is changed in size, for example when you zoom in you are increasing the size of the image and its high frequency detail so that the phosphor groups on the inside of the screen can correctly sample it, and thus the moire goes away, it comes back and actually "dances" if you zoom out detail. The same thing happens with video on tv when the guy with the wild zebra patterned tie moves around and the moire dances as well.

It's really cool stuff.

If you'd like to read more on how tv's work, click here:
http://www.howstuffworks.com/tv.htm

For a detailed technical break down of color and imaging displays:
http://www.cse.fau.edu/~maria/COURSES/COP4930-GS/colornotes.htm

Enjoy!

Regards,
--DSL

 

[Erik Magnuson]

I am curious what is your profession ?

I am a software engineer. Most of my work has related to real-time data acquistion and analysis. I've seen very few CPU limited systems, none of them based on modern microprocessors. The I/O or even the memory bus is nearly always the performance bottleneck.

Which is why I was so puzzled when the SD-9 specs said "RAW only". The absolute quality zealots were aghast that anyone might want to degrade their precious data into an 8-bit lossy format. But there are times when performance and convenience are important as well. (After all, how many of us use 8x10 view cameras?) And all current cameras can manage to interpolate-whitebalance-compress-write faster than they can write plain raw data (i.e., they seem to be I/O limited as well.) So I started digging into the specs and published information too see if I could figure out why.--ErikFree Windows JPEG comment editor http://home.cfl.rr.com/maderik/edjpgcom

 

[Erik Magnuson]

Personally, I don't think CCD or CMOS do that well. I decided to
limit the 1D to 400 ISO. Anything above is just to grainy. The D30
images look like I would limit to 200.

Well, you may be disappointed with the 1st generation of Foveon technology. Go to http://www.foveon.net/docs/F7_Datasheet.pdf . Look at the part where it says "ISO: 100 Full Resolution mode". Think on what the "variable pixel size" implies for resolution.

Then go to http://www.sigma-photo.com/Html/news/news_sd9_fs.htm

Read the part where it says "LOW mode (756x504 pixels) x3 allows consecutive shooting" and the highest ISO listed is 400 (and it does NOT say which modes this applies to.)

It may be a while before the X3 can match your 1D at higher ISO. it certainly will not match the frame rate at any reasonable resolution.

??? This is only true for RAW. Because it has 3 times the
information.
For JPEG it is actually about 1/2 for the same detail.

But the SD9 does NOT support JPEG in camera! Again, look at the Sigma page. You will have to make significant resolution tradeoffs to get lower storage sizes or frame rates.

I agree that it is not a clear winer at this point in time but it has
so much more room to improve on than technologies that have been used
for 3 decades.

And this is based on what? Just the fact that it's newer? Note that I'm not saying that's its impossible or even unlikely that X3 will take over from mosiac CCDs. But there may also be a few tough problems that will prevent its universal adoption. We just don't know yet.
--ErikFree Windows JPEG comment editor http://home.cfl.rr.com/maderik/edjpgcom