Argument about pixels seeks comments

[Roland Karlsson]

I have over 30 CFA digital cameras. I have two X3 digital cameras,
that's hardly a case for CFA "envy" - LOL.

OK

The post is in response to the very annoying and inflamatory
statement by the Senior Technical Editor at EDN who "should" know
better.

Yes - the title plus some statements in the article is unnecessary.

It a simple fact that with identical b&w resolution chart
measurements, the X3 has higher color resolution. Nothing complicated
just simple measurement observations.

Its true that CFA has worse color resolution.

--
Roland

 

[Roland Karlsson]

The point being that a sensor which reveals all
the color resolution as well as it reveals black and white resolution
gives the viewer a more realistic expectation of accuracy in the
photo.

That contradicts the conventional description of how our visual
system works, e.g. humans do not have the same ability to perceive
detail in all colors. Thus far your point is a minority opinion.
That's why most reviewers do not accept it.

This is an interesting topic indeed! For two reasons.
1. Is this true? Can´t we see that an image is slightly color smeared?
2. Is the CFA image just a color smeared sharp image?

--
Roland

 

[Lin Evans]

Erik Magnuson wrote:
snip

The point being that a sensor which reveals all
the color resolution as well as it reveals black and white resolution
gives the viewer a more realistic expectation of accuracy in the
photo.

That contradicts the conventional description of how our visual
system works, e.g. humans do not have the same ability to perceive
detail in all colors. Thus far your point is a minority opinion.
That's why most reviewers do not accept it.

I doubt most reviewers have a clue about the human visual system other than very rudimentary knowledge. I think most reviewers don't accept it because they don't want to admit that there is something better than the "sacred cow" of conventional sensor technology. It's pretty easy for many to "see" the differences and regardless of how much difference the human visual acuity might have or not for specific colors, we have "enough" to see the red on blue lines in the X3 capture which is "missing" from the comparison CFA capture. So for whatever degree that's worth to the aesthetic of the print, the difference is palpable.

Best regards,

Lin

--
Erik

 

[ChromeLight]

The whole issue of this

thread is that by emphasizing 14MP they are inviting other
comparisons.

Actually, Sigma and Foveon have always wanted the definition that B&H lays out on their web site

Sigma SD14, 4.7 Megapixel, SLR, Digital Camera (Foveon X3 CMOS Sensor: 4.7M Red + 4.7M Green + 4.7M Blue = 14 Megapixel Total Resolution) (Camera Body)

The problem was that when the SD9 came out many 'media spaces' (web sites, catalogs, etc) only had room for one designation and that was MP count. They chose 3.4mp, which is misleading. 10.2 is misleading as well, and in fact Foveon wants neither. The B&H notation is the prefered way of presenting the camera and chip.

 

[Roland Karlsson]

I doubt most reviewers have a clue about the human visual system
other than very rudimentary knowledge.

Maybe - but many multimedia interested persons knows that MP3, JPEG, TV etc uses the fact that you can compress for the redundancy in our senses. It is quite likely that many know that Bayer CFA also uses this kind of redundancy.

I think most reviewers don't
accept it because they don't want to admit that there is something
better than the "sacred cow" of conventional sensor technology.

I am 100% sure you are wrong here. There is no sacred cow. It does simply not exist. Reviewers do not have any reason to favor CFA technology. Maybe favor brands - but not technology. They don´t care about the technology IMHO.

It's
pretty easy for many to "see" the differences and regardless of how
much difference the human visual acuity might have or not for
specific colors, we have "enough" to see the red on blue lines in the
X3 capture which is "missing" from the comparison CFA capture. So for
whatever degree that's worth to the aesthetic of the print, the
difference is palpable.

Nope - it is not easy to see the difference. The spread among different images is larger than the difference between the technologies. And to see the difference you need to do serious pixel peeping. You have to know what to look for. You have to be able to quantify the difference.

Reviewers seldom go by - "this feels good".

They measure resolution and noise. The noise is usually slightly larger than from good CFA cameras - or at least not significantly better. The resolution is usually in the order of a 10 Mpixel CFA camera for SD14.

And - given those facts - there are many cheaper CFA cameras out there - with more main stream mounts than the SA mount. So - the verdict is usually - not price worthy.

You (and also I) think there are some properties in the X3 images that might motivate paying the higher price and investing in lenses from the SA system. But - the properties are esoteric for most people - and is not enough to get good reviews.

--
Roland

 

[Erik Magnuson]

This is an interesting topic indeed! For two reasons.
1. Is this true?

This is conventional wisdom. It's also why many recommend sharpening just the Luminance channel.

Can´t we see that an image is slightly color smeared?

Here is an experiment to try:
http://forums.dpreview.com/forums/read.asp?forum=1027&message=6088907

Another way to do the same experiment is to compare 1:1 and 2:1 and 2:2 chroma sampling in JPEG. At what size can you spot the differences? (Hint: just like CFA, red/blue borders show the most.)

2. Is the CFA image just a color smeared sharp image?

Well the blur is not perfect. There are reconstruction artifacts added in.

--
Erik

 

[Roland Karlsson]

Actually, Sigma and Foveon have always wanted the definition that B&H
lays out on their web site

Sigma SD14, 4.7 Megapixel, SLR, Digital Camera (Foveon X3 CMOS
Sensor: 4.7M Red + 4.7M Green + 4.7M Blue = 14 Megapixel Total
Resolution) (Camera Body)

The problem was that when the SD9 came out many 'media spaces' (web
sites, catalogs, etc) only had room for one designation and that was
MP count. They chose 3.4mp, which is misleading. 10.2 is misleading
as well, and in fact Foveon wants neither. The B&H notation is the
prefered way of presenting the camera and chip.

Thats a good definition - somewhat lengthy maybe.

What Foveon wants I don't know. I can only guess. And I guess that they are more than satisfied with the number 14 Mpixels.

--
Roland

 

[Roland Karlsson]

1. Is this true?

This is conventional wisdom. It's also why many recommend sharpening
just the Luminance channel.

There is another issue. Even if it is true. How do color smeared images react to image manipulations, e.g. sharpening and increasing saturation?

2. Is the CFA image just a color smeared sharp image?

Well the blur is not perfect. There are reconstruction artifacts
added in.

Yepp - and this is the main issue for digital photography and number of pixels. Four, six, eight, ten Mpixels might just be fine - if the pixels are good. But - if the pixels are sub par quality - e.g. by noise or artifacts - then you really want more pixels - or better pixels.

NOTE - X3 images are not artifact free.

--
Roland

 

[ChromeLight]

Roland,

I'm shocked you've never been to the following Foveon site.
http://www.x3f.info/

The URL is in SPP!!!!!!!!!!! It's been there for years! Foveon is quite clear they want the longer definition that's used by B&H.

Actually, Sigma and Foveon have always wanted the definition that B&H
lays out on their web site

Sigma SD14, 4.7 Megapixel, SLR, Digital Camera (Foveon X3 CMOS
Sensor: 4.7M Red + 4.7M Green + 4.7M Blue = 14 Megapixel Total
Resolution) (Camera Body)

The problem was that when the SD9 came out many 'media spaces' (web
sites, catalogs, etc) only had room for one designation and that was
MP count. They chose 3.4mp, which is misleading. 10.2 is misleading
as well, and in fact Foveon wants neither. The B&H notation is the
prefered way of presenting the camera and chip.

Thats a good definition - somewhat lengthy maybe.

What Foveon wants I don't know. I can only guess. And I guess that
they are more than satisfied with the number 14 Mpixels.

--
Roland

 

[Erik Magnuson]

we have "enough" to see the red on blue lines in the
X3 capture which is "missing" from the comparison CFA capture.

In similar situations we also usually have enough to see the "missing" data from lower luminance resolution of the SD cameras. One of the reasons why the purported "superiority" almost never shows up in side-by-side shots is that when looking at both images we see both limitations. We may find areas where the red/blue or red/black is limited in CFA images, but we also see areas where the higher spatial resolution of the CFA sensor has resolved detail that is missing or rendered incorrectly in the X3 image.

Which set of limitations is more important how often is the heart of point I was making in the "weighting" post.

--
Erik

 

[Erik Magnuson]

How do color smeared
images react to image manipulations, e.g. sharpening and increasing
saturation?

Don't forget the differences in demosaicing approaches. And camera AA filters. Too many variables to generalize.

then you really want more pixels - or better pixels.

Only if you need to make larger prints to be viewed at closer than normal viewing distance. (It's also somewhat subject dependent as well.)

--
Erik

 

[SandyF]

Roland,

I'm shocked you've never been to the following Foveon site.
http://www.x3f.info/

The URL is in SPP!!!!!!!!!!! It's been there for years! Foveon is
quite clear they want the longer definition that's used by B&H.

thanks ChromeLight!

I've seen most of the included hotlinks before, but not all. Thanks very much for the link!
Best regards Sandy
[email protected]
http://www.pbase.com/sandyfleischman
http://www.flickr.com/photos/sandyfleischmann

 

[Roland Karlsson]

I'm shocked you've never been to the following Foveon site.
http://www.x3f.info/

I have never heard of the x3f.info site.
It is owned by Foveon, but why is it hidden on this weird domain?
This text is old and for SPP2. What link do you find in SPP3?

The URL is in SPP!!!!!!!!!!! It's been there for years!

I don´t own a Sigma camera and therefore have never installed SPP.
So - how should I find it?

Foveon is
quite clear they want the longer definition that's used by B&H.

Nice and balanced text. Written by an engineer I assume.

Ahhhhh ... but they ... even in this article where there is ample room for the longer definition ... calls it a 10 MPixel sensor several times. The longer definition IS cumbersome. Moreover - it is wrong - or simplified if you shall be nice. The X3 sensor do not have red, green and blue layers. If it had - I would have bought a camera a long time ago.

--
Roland

 

[Laurence Matson]

The X3 sensor do not have
red, green and blue layers. If it had - I would have bought a camera
a long time ago.

Am I missing something here? Are you? What are you trying to say?

--
Laurence

My idea of good company is the fellowship of clever, well-informed people, who have a great deal of conversation and liberality of ideas.

Jane Austen

http://www.pbase.com/lmatson/root
http://www.pbase.com/sigmadslr/root
http://www.pbase.com/cameras/sigma/sd10
http://www.pbase.com/cameras/sigma/sd9
http://www.beachbriss.com

 

[Zone8]

.... getting a print that looks sharp in the areas that need to be sharply defined.

All other criteria has no interest to me whatsoever. It's the end image and it's quality tonally that matters.
--
Zone8

The photograph isolates and perpetuates a moment of time: an important and revealing moment, or an unimportant and meaningless one, depending upon the photographer's understanding of his subject and mastery of his process. -Edward Weston

 

[Roland Karlsson]

The X3 sensor do not have
red, green and blue layers. If it had - I would have bought a camera
a long time ago.

Am I missing something here? Are you? What are you trying to say?

I hear you clearly ... but I do not understand you.

Do you believe that the X3 sensor has a red a green and a blue layer?

--
Roland

 

[Guido FORRIER]

yes 3 layers this is phisics not software) :

'Finally, Foveon has combined the best of what both film and digital have to offer. This is accomplished by the innovative design of the three layer Foveon X3 direct image sensor. Similar to the layers of chemical emulsion used in color film, Foveon X3 image sensors have three layers of pixels. The layers of pixels are embedded in silicon to take advantage of the fact that red, green, and blue light penetrate silicon to different depths – forming the first and only image sensor that captures full color at every point in the captured image.'
that is the physical base of foveon tech.

guido

 

[Tristan Cope]

Being "related" to resolution and "equating" to resolution are
different things. The pixel count in a picture only relates to
resolution via the number of discrete samplings and the quality of
these samplings. Duplicating existing datum points and moving them to
another location doesn't affect resolution in any positive way as is
evidenced by the results of testing a 3.4 megapixel count Foveon
sample against a 6.3 megapixel count CFA sample for true optical
measured b&w resolution.

Firstly, as you know, the points aren't "duplicated". Values in the grid are calculated from surrounding values. Secondly, a large part of the reason the 6.3MP CFA sensor doesn't perform as well as you might expect from the pixel count in the B&W resolution test is because of the AA filter.

That the 3.4 megapixel count Foveon actually
has "more" overall measured optical resolution across the spectrum
demonstrates rather conclusively that having a greater pixel count in
a picture does not equate to having greater resolution.

There are many things that effect the final "resolution" of the image (as measured for example by LPH). Pixel count is one of them, presence or absence of an AA filter, lens quality, even the raw conversion software used. I agree that pixel count certainly does not determine resolution, but it is a factor.

Foveon has the number of discrete
collection points they claim.

Yes, but the fact that they are stack on top of each other, rather than side-by-side makes some difference to the "resolution".

I'm not going to get into yet another
argument about what constitutes a "pixel" and regardless of what we
choose to call them, Foveon gets more "resolution" per output pixel
than CFA based technology.

Absolutely. But this isn't just about using a CFA - there is also the issue of the AA filter. Of course, sensors with a CFA tend to have an AA - but there are exceptions, and these cameras without AA filters produce higher resolution images than their counterparts with AA filters.

So using the output pixel matrix size to
"equate" to optical resolution is obviously false.

I don't believe any of us who use Sigma cameras really believe that
output pixel count isn't "related" to resolution, but it's only
meaningful to use this as a guideline for resolution when comparing
same type technology.

Frankly, I don't even think it's that useful for comparin the same technologies. Are the new 10 and 12MP compacts really producing extra resolution above 6 and 8MP ones? When we talk about resolution, we should be talking about something measurable in the final image - not how many nubbins the sensor has, because not all "nubbins" are equal (even on Bayer sensors). We should be talking about LPH or something similar.

If we talked about LPH instead of MP, people might be about as interested in "resolution" as they were in the days of film - which is to say not very much for most people. Of course all of this MP nonsense is mostly driven by the manufacturers because it sells cameras.

--
I think I'll change my signature

 

[Tristan Cope]

Call me old fashioned but perhaps resolution should be thrown out all
together. In my own mind i used defined detail capture. In other
words, if you photograph a bill board but can't read the words on it,
you haven't captured 'defined detail', meaning there isn't enough
detail for you to define what the object is. If another imaging
system can allow you to read the words, then it's--in practical
terms--higher resolution. It's gets more complicted when you throw in
landscapes, etc and then print at varying print sizes.

Isn't measuring resolution in LPH or perhaps more accurately looking at MTF just a standardised way of measuring exactly this?

--
I think I'll change my signature

 

[Roland Karlsson]

Three layers is not necessarily the same as a red a green and a blue layer.

--
Roland