Go Back   Steve's Digicams Forums > Digital Cameras (Point and Shoot) > Panasonic / Leica

Reply
 
Thread Tools Search this Thread
Old Oct 28, 2005, 2:30 PM   #21
Senior Member
 
squirl033's Avatar
 
Join Date: Dec 2004
Posts: 6,720
Default

by upsampling your 2MP file and resharpening it, you are effectively creating another 4MP file, which is bound to look much like the image originally shot at 4MP. here's a much more accurate and appropriate way to test your theory. shoot a photo of a detailed subjectusing your camera's full 4MP resolution. then reset the camera to 2MP resolution and take the same photo, same light, same angle,everything.now look at the two images on your screen, using the zoom function in your image viewer to make the 2MP file the same visual size as the 4 (don't shrink the 4MP file), without resampling. you will find that the 4MP image is clearer and more detailed. if it isn't, then you may have a valid concern about your FZ15's sensor. otherwise, don't worry about it.


squirl033 is offline   Reply With Quote
Old Oct 28, 2005, 3:55 PM   #22
Senior Member
 
Join Date: Nov 2004
Posts: 511
Default

Squirl, Thats the way I saw it, but he seemed to be having so much fun his way.:lol:
Tazzie is offline   Reply With Quote
Old Oct 28, 2005, 5:06 PM   #23
Member
 
Join Date: Oct 2005
Posts: 36
Default

Hi Everybody!

Although this is my first post, I have been visiting these forums for over two
years. This thread caught my interest because it is very relevant to a project
I am currently involved in. That's why I absolutely had to have my say in this
thread.

The fact is that Rubinsky is half right and half wrong. His camera really does
have a maximum resolution of only 2 megapixels. But, because of the technical
aspect of Bayer pattern color filter arrays (CFA), a four megapixel image is
necessary to maintain ALL of the detail. This is explained in greater detail
below.

To understand this, let's first consider the true meaning of the word "pixel".
The word "pixel" comes from the two words "picture element". A pixel is the
building "block" of a digital picture. As we all know by now, each pixel is
made up of three color components: R, G, and B.

The big confusion in megapixels
is created when manufacturers call a color-blind photodiode a pixel. A photodiode
is NOT an entire pixel, it's true effect is that of a single color component of
a pixel. A photodiode cannot detect color, only brightness. Therefore, a color
filtering dye must be placed over the photodiode so that it will only see ONE
of the primary colors. Three of these photodiodes are necessary to produce a
full color pixel. So, in reality, an 8 megapixel camera is more correctly called
an 8 megaphotodiode camera.

Each photodiode is, however, treated as though it were a full color pixel by
filling in the two missing color component values. This is done by examining the
nearest photodiodes that are of the same color as the missing components. For
example, if finding the missing red and blue values for a green photodiode, look
at the nearest red and blue photodiodes and use their values. This is known as
interpolation and it is basically used to try to determine the missing values
that were most likely to have been in a certain location.

Virtually every digital camera (including Rubinsky's) uses a Bayer pattern CFA.
In every Bayer pattern CFA, there are just as many green "pixels" as there are
red and blue "pixels" combined! That means that an 8 megapixel camera, for example,
truly has only 4 million green "pixels", 2 million red "pixels" and 2 million
blue "pixels". Bayer pattern CFAs were designed to contain more green "pixels"
to take advantage of the fact that the human eye (and brain) perceives most of an
image's detail from the green wavelengths of the spectrum. So, in simple terms,
the Bayer pattern CFA provides more green "pixels" on the sensor to give the
image it's detail and the red and blue "pixels" are there to provide color to
the image.

Now, because the green component of an 8 mega"pixel" image was captured with
only 4 million photodiodes, and because green provides the detail for the image,
virtually all digital cameras have only half the advertised resolution; and that
is true only under ideal conditions.

Now you may be asking, "If the red and blue components only have 1/2 the resolution
of the green component, then why do they appear to be the same resolution as the
green component?". The simple answer to this question is edge detection. Because
the green component has the most detail, it's edges are carefully copied to the
red and blue components. This process not only has the effect of increased
resolution; it also helps to hide the artifacts caused by sampling the three color
components for a single pixel from differing locations on the xy axis of the sensor.

If you don't believe this to be true, you can prove to yourself that it is true.
Simply use you Bayer CFA based digital camera to take a picture of a completely
red colored object on a black background. Because the green channel will contain
no edge detail, no edges can be copied from it. That means that the red channel
must be left untouched and therefore shows it's original edge detail. Hence, you
will notice an increased softness and pixelation. There are many irreparable
artifacts introduced by copying edges from one component to another; but they
are too complicated to go into at this time.

This ultimately means that with ANY Bayer CFA based digital camera, if you take
a picture of a completely red or blue object on a black background, the actual
resolution will be 1/4 of the advertised resolution. This also means that if you
were to take a picture of a completely gray scene, the green component would be
able to provide completely accurate edge data to the red and blue components
with very little or no adverse effects; thus providing a completely accurate
1/2 advertised pixel count. A gray scene is an example of the most ideal condition
for a Bayer CFA. A completely red or blue object on a black background is an
example of the worst possible condition for Bayer CFA. So, when using a Bayer
CFA based digital camera, the best possible resolution that can be expected
is 1/2 of the advertised resolution and the worst that can be expected is 1/4
the advertised resolution.

Now you are probably asking "So, if an 8 mega"pixel" image really only has a
maximum resolution of 4 megapixels, then why do I need an 8 megapixel image to
maintain ALL of the detail?". This is because the green "pixels" on a Bayer
pattern CFA based sensor arranged in a diagonal fashion.

I hope this can be understood easily enough. It quite a complicated subject,
so I apologize if I haven't made everything clear.


-Ted
dimagez1 is offline   Reply With Quote
Old Oct 28, 2005, 6:19 PM   #24
Senior Member
 
Join Date: Nov 2004
Posts: 511
Default

Wow that was well said, Ted. Alrighty then, if ya wanna put it that way. Its just that somehow you seem to have taken a good deal of the fun out of it... now that we can better understand his point. So then are we as well off or better served by running at standard instead of highest quality as he is trying to say?
Tazzie is offline   Reply With Quote
Old Oct 28, 2005, 6:59 PM   #25
Senior Member
 
squirl033's Avatar
 
Join Date: Dec 2004
Posts: 6,720
Default

it's all relative. if a 4MP camera can produce what is generally accepted as a 4MP image using - in effect - half it's photodiode count, then a 2MP camera is only using 1MP. it's all in the nomenclature; we've gotten in the habit of considering an image of a given size (in this case, 2304x1760)to be a "4MP" photo, and a file size of 1600x1200 is considered a "2MP" image, even if the sensor/CVA arrangement only yields half that many "real" pixels. either way, the "4MP" camera delivers twice the resolution of a "2MP" one.

the real issue isn't absolute pixels (or photodiodes), it's whether a "nominal" 2MP image, when upsampled and sharpened, can look as good as a "nominal" 4MP unprocessed image. or, using Dimagez1's count,whether a 1MP pic can match a 2MP for detail and clarity?
squirl033 is offline   Reply With Quote
Old Oct 28, 2005, 7:45 PM   #26
Senior Member
 
Join Date: Oct 2003
Location: Indian Rocks Beach, FL
Posts: 4,036
Default

I took the same shots at 2 & 4 Mp with my FZ10. A small part is cropped out and blown up for viewing with no other processing. The 4Mp image clearly has more detail. I doubt they changed sensors between the two cameras.

2Mp:


4Mp:


Full shot crop was taken from:


slipe is offline   Reply With Quote
Old Oct 28, 2005, 8:03 PM   #27
Member
 
Join Date: Oct 2005
Posts: 36
Default

You'll definitely want to use the full advertised resolution setting,
because as I said, it is necessary to maintain ALL of the detail.
Let me explain that point in greater detail. First, take a look at the
illustration below.

RGRGRG
GBGBGB
RGRGRG
GBGBGB

That shows the typical arrangement of a Bayer pattern CFA.
Notice how the green photodiodes exist in all columns and all
rows; unlike the red and blue photodiodes. Take any column or
row away (to reduce the size of the image) and you will lose
green photodiodes. Remember, the green photodiodes are
providing most of the image's detail, so you want to avoid losing
them.

Some people may think that taking a 4 megapixel digital camera
image and reducing it to 1 megapixel would cause a massive
amount of detail loss because "it is getting rid of 3/4 of all the
pixels". In reality, doing so only reduces the real pixel count by
1/2 at the very most. Like I pointed out before, if it's a picture
of a completely red or blue object on a black background, then
reducing the pixel count to 1/4 would actually induce no detail
loss at all!

So what's my point? It is that Rubinsky's claims of reducing the
pixel count to 1/2 while keeping the image "practically identical"
to the full size image is very nearly completely true. Doing this
doesn't cause anywhere near as much detail loss as might be
expected; especially when done with a high quality resampling
algorithm. However, I still can't say that anyone should be doing
this because it will cause some detail loss. In-camera resizing
may not be of sufficient quality; potentially causing more problems.
dimagez1 is offline   Reply With Quote
Old Oct 28, 2005, 8:14 PM   #28
Member
 
Join Date: Oct 2005
Posts: 36
Default

I should also point out the aspect of JPEG compression. In a larger
image, compression artifacts will be smaller in relation to size of the
entire image. So, if you are using JPEG mode, then reducing the pixel
count to half before compression will ultimately cause compression
artifacts to appear larger after the image has been resized back up
to full size.
dimagez1 is offline   Reply With Quote
Old Oct 28, 2005, 8:20 PM   #29
Senior Member
 
squirl033's Avatar
 
Join Date: Dec 2004
Posts: 6,720
Default

dimagez1 wrote:
Quote:
So what's my point? It is that Rubinsky's claims of reducing the
pixel count to 1/2 while keeping the image "practically identical"
to the full size image is very nearly completely true.
except that Rubinsky didn't claim that. he claimed that he couldn't see the difference between an unertouched 4MP image and animage taken at 2MPthat he upsized and resampled to 4MP and sharpened. he was comparing two "4MP" images.

if you take a 4MP photo at 2304x1760 resolution,resize to 1600x1200, and then simply compare that reduced image toa 4MP photo at the same viewing size, it's true you won't see a lot of difference. but if you shoot a photo at 1600x1200, and then "zoom" the image (without resampling)up to the same visual size as the original, full-size4MP file, you will see quite a difference difference io the image quality.
squirl033 is offline   Reply With Quote
Old Oct 28, 2005, 8:42 PM   #30
Member
 
Join Date: Oct 2005
Posts: 36
Default

Actually, he did say that. It was in his first sentence.

The fact is that virtually all digital cameras have only half the
pixel resolution that they are advertised to have. Simply adding
in missing color values by using neighboring photodiodes does
not magically increase the actual resolution to the full advertised
amount. Therefore, theoretically, resizing the image to have half
of the advertised pixel count should maintain all detail. However,
a regular resample algorithm isn't optimized to deal with Bayer
CFA sampled images; it will not maintain ALL of the detail but the
end effect will be very close to that of an optimized resample
algorithm.


dimagez1 is offline   Reply With Quote
 
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

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 Off
Pingbacks are Off
Refbacks are Off



All times are GMT -5. The time now is 4:35 AM.