Sony AR7IV vs GFX 50 for DSLR scanning.

"Ooooh, mummy. Someone's daring to question my hoard of precious knowledge (mmm, precious knowledge, dribble); suggesting some of it might be untrue!

I'm scared. I need my blanky.

I know, I'll throw a tantrum and call them names and claim they're stupid. That always shows I'm absolutely right."

Grow up Andrew! Or crawl back into your digital kennel. The grown ups were trying to hold a discussion - you might want to look that last word up in the dictionary.

"Ooooh, mummy. Someone's daring to question my hoard of precious knowledge (mmm, precious knowledge, dribble); suggesting some of it might be untrue!

 

I'm scared. I need my blanky.

I know, I'll throw a tantrum and call them names and claim they're stupid. That always shows I'm absolutely right."

 

Grow up Andrew! Or crawl back into your digital kennel. The grown ups were trying to hold a discussion - you might want to look that last word up in the dictionary.

"Wise men speak because they have something to say; Fools because they have to say something". -Plato

That quoted text above makes this abundantly clear.

The oft stated argument that Bayer patterning should have twice as many green sensors because the eye is most sensitive to green is a complete non-sequiter, and totally ridiculous.

This rationale doesn't really make good engineering sense to me, either, if it is the only reason for having a higher ratio of green pixels.

But in perusing Bayer's patent (readily available from the U.S. Patent Office: Patent Images ), he indicates that a doubling-up of the green pixels (which supply most of the luminance information) in an RGGB pattern permits a higher effective resolution because the pixel pattern then becomes "regular and uniform in two orthogonal directions":

Bryce Bayer:

The preferred luminance pattern [h]as especially desirable sampling qualities which result from the uniformity and orientation thereof. Of the possible patterns including only half of the element positions of a substantially rectangular array, the preferred pattern is the one that affords the largest useful region of frequency space, i.e., considering all directions on the array, the minimum Nyquist limit is largest. Moreover, because of the orientation of the preferred pattern to the major axes, this usable region proves more extensive in the horizontal and vertical directions...those directions where the human visual system is said to have greatest resolving power. (U.S. Patent #3,971,065, Column 4, lines 28-36)

Bryce Bayer:

those directions where the human visual system is said to have greatest resolving power

And that dismisses as well, RJ's actual nonsense:

The eye is most sensitive to green - agreed. Therefore it should be presented with twice as much green - whaaaaat!? That's a complete nonsense.

I don't have any desire to step into the cornfield between the Hatfields and McCoys, but I agree with rodeo_joe's skepticism about the usual (oversimplified) assertion that the green pixels are doubled only because the eye is more sensitive to green. If the only reason for having a higher proportion of green pixels were to mimic the human eye response, then about 70% of the pixels should be green, 21% red, and 9% blue. I suspect that a sensor with those proportions would not be too popular.

The patent makes clear that there are several good reasons for using the RGGB pattern. I think some of them are less important today because of technology improvements since the 70s; but if the RGGB pattern permits the highest resolution, that advantage is still as applicable as it was then.

70% of the pixels should be green, 21% red, and 9% blue.

Veery interestink [sic]! What would a 7:2:1 matrix look like? Like John Harrison's chronometer, it has to be both accurate and practicable.

Getting back to basics, the Bayer pattern has worked admirably for twenty years, and is incorporated into nearly every digital camera in existence. The problems so gravely cited are easily solved by including an anti-aliasing filter. The nature of the AA filter is such that most of the lost resolution can be restored mathematically. Light isn't destroyed in the process, merely rearranged.

Better heads than ours once argued how many angels could dance on the head of a pin. In many respects, those arguments made as much sense as this thread, given total disregard of things we can actually see.

I don't have any desire to step into the cornfield between the Hatfields and McCoys, but I agree with rodeo_joe's skepticism about the usual (oversimplified) assertion that the green pixels are doubled only because the eye is more sensitive to green.

Skepticism is the first step towards truth.-DENIS DIDEROT

Joe isn't a skeptic, he's provided, without a lick of fact or evidence, that the design as outlined by Bayer and explained by many, many others, a myth, a non-sequiter, and totally ridiculous.

Nothing wrong with being a skeptic. Until all the facts and science illustrate instead, the Earth isn't flat. Bayer isn't asserting, his patent and the results of his work is, as far as I know, disputed without any evidence by Cowboy Joe.

Again, he entitled his opinion. He's not entitled to informed opinions.

Yes, the patent and lots of other outside references by experts in the field makes clear that there are several good reasons for using the RGGB pattern. And one is outlined in the patent by Bayer that I have shown here that directly disagrees with RJ. Enough said.

There are two kinds of fools: those who can't change their opinions and those who won't. -Josh Billings

This must be all part of the myth? Who to believe (it's only a trick question for one guy here :D )

http://www.inf.fu-berlin.de/lehre/WS02/robotik/Vorlesungen/Vorlesung2/ComputerVision-2.pdf

This arrangement is called a Bayer Filter, because it was invented at Kodak by one Dr. Bayer. As can be seen, in the Bayer array half of the pixels have a green filter, the color for which the human eye is more sensitive. The red and blue filters are arranged in rows and columns, with one missing pixel in each direction. At the time Bayer registered his patent, he also proposed to use a cyan-magenta-yellow combination, that is another set of opposite colors. This arrangement was unpractical at the time because the necessary dyes did not exist, but is used in some new digital cameras.

https://www.cambridgeincolour.com/tutorials/camera-sensors.htm

A Bayer array consists of alternating rows of red-green and green-blue filters. Notice how the Bayer array contains twice as many green as red or blue sensors. Each primary color does not receive an equal fraction of the total area because the human eye is more sensitive to green light than both red and blue light. Redundancy with green pixels produces an image which appears less noisy and has finer detail than could be accomplished if each color were treated equally.

How Digital Cameras Work most common pattern of filters is the Bayer filter pattern. This pattern alternates a row of red and green filters with a row of blue and green filters. The pixels are not evenly divided -- there are as many green pixels as there are blue and red combined. This is because the human eye is not equally sensitive to all three colors. It's necessary to include more information from the green pixels in order to create an image that the eye will perceive as a "true color."

Bayer Filter

The pixels of the Bayer filter are arranged in a chequerboard pattern on the sensor. There are twice as many green as red and blue pixels in order to mimic the higher sensitivity of the human eye towards green light.

Bayer Pattern - an overview | ScienceDirect Topics

A Bayer pattern array [4] is shown in Figure 4.17. Half of the filter elements are green and the remainder are split between blue and red. This approximates human photopic vision where the M and L cones combine to produce a bias in the green spectral region.

Cameras use Bayer filters to attain true color

Bayer noticed that the human eye obtains most of its sharpness information from green light, which is why the pattern has more green pixels. By using two green-filtered pixels for every red or blue, the Bayer pattern is designed to maximize the perceived sharpness in the luminance channel, composed mostly of green data. However, since the image is undersampled, the full detail available is not attained.

Now we'll be told, all of the above is 'fake news" and "Alternative facts". :eek:

Here's the alternative facts:

The oft stated argument that Bayer patterning should have twice as many green sensors because the eye is most sensitive to green is a complete non-sequiter, and totally ridiculous.

It's time this stupid Bayer myth was exploded.

From my viewpoint, it is only necessary to read what Bayer himself wrote.

As I indicated above, if the reason for doubling the green pixels were to make the pixel counts mimic the human eye's relative color sensitivity, one should at least do it "right". The pixel ratios would then be 70% green, 21% red, and 9% blue. Sounds ridiculous to me, too.

Most of the conventional Internet explanations are non-sequiturs, like this one:

As can be seen, in the Bayer array half of the pixels have a green filter, the color for which the human eye is more sensitive.

The cambridgeincolour.com extract attempts an explanation, although it is somewhat muddled and the wrong part is emphasized. The important part is:

Redundancy with green pixels produces an image which appears less noisy and has finer detail

...by improving the quality of the pixel arrangement, which Bayer described in his patent (one relevant portion of which I've previously quoted).

We disagree they are non-sequiturs but I believe that Cowboy Joe's statements are absurd and without any evidence as expressed:

"It's time this stupid Bayer myth was exploded." And calling Bayers patient AS I have shown: "abundantly nonsensical."

His opinion expressed is abundantly nonsensical and foolish!

“There are two ways to be fooled. One is to believe what isn't true; the other is to refuse to believe what is true.” -Søren Kierkegaard

Wow! Mr. Kierkegaard's name spelled correctly for a change. We are making progress!

As for the RGB argument, how would you fill in a matrix of rows and columns? RRGB, RGGB and RGBB all fit nicely and once you realize that, you can easily come up with a justification for any of them. Some of the arguments in favor of RGGB even make sense. But I think any of the justifications are just after-thoughts; no more, no less,

Feel free to disagree, Andrew.

Wow! Mr. Kierkegaard's name spelled correctly for a change. We are making progress.

Return of the troll again.

As for the RGB argument, how would you fill in a matrix of rows and columns?

Don't ask questions when the correct answers will be ignored and misunderstood. Your agenda here is quite transparent. Maybe start a new post as a newbie asking about this. You are a Glutton for Punishment but apparently the punishment is worthwhile in order to get attention. And that's the posting agenda that can be seen here and in other photo forums, the list is sadly long but so consistent: Ask a question of the group, then immediately dismiss and push back on anything said that doesn't sync with the predetermined desired answer (notice the question marks!):

Flicker, eyestrain or headache when using a monitor with LED backlighting?

AND:

How do you like your 23" NEC P232W-BK-SV with LED backlighting?

AND:

Three different color calculators can't be wrong, can they?

AND:

Do SoLux bulbs meet color temp specs?

And (push back) again:

AND:

NEC SpectraView update doesn't work

AND:

NEC PA242W with SpectraView problems

AND:

NEC P242W with SpectraView weird flicker

AND:

Epson R2880 suddenly prints with reddish color cast

Nonsense about sensors overheating and being an issue in 10 pages of back and forth that resulted in it all getting deleted on the LuLa forums but still here onPhotoNet:

Are DSLRs doomed? Tony Northrup says no.

I do believe it illustrates a pattern: post about a 'problem' or ask a question as seen above, ask for consensus of others to agree the problem/answer exists (rarely if ever does consensus let alone any agreement occur), complain when what you are told doesn't match what he wants to hear, go away for X number of months and start all over again.

Some data today and maybe someone to contact about the Sony:

Getting the highest negative scan quality for under for under 10K