Velocity Reviews - Computer Hardware Reviews

Velocity Reviews > Newsgroups > Computing > Digital Photography > Re: The death of the Bayer filter? Maybe not.

Reply
Thread Tools

Re: The death of the Bayer filter? Maybe not.

 
 
TheRealSteve
Guest
Posts: n/a
 
      04-21-2012

On Wed, 18 Apr 2012 18:42:23 -0800, http://www.velocityreviews.com/forums/(E-Mail Removed) (Floyd L.
Davidson) wrote:

>Alfred Molon <(E-Mail Removed)> wrote:
>>In article <(E-Mail Removed)>, Floyd L. Davidson says...
>>> The data points are *calculated*, not guessed at.

>>
>>Sure. And you can also "calculate" where the stock market will be in one
>>month.

>
>That's a bit of abject silliness.
>
>The data points are not projected for some point in the future.
>And there is nothing close to a "guess".
>
>The usual way to to select 2 or more measured values from the
>point surrounding the desired point. From those a gradient is
>established, and the value of the gradient at the desired point
>is calculated. Generally it involves calculation of both
>horizontal and vertical gradients, and hence is obviously rather
>accurate!


But still a guess, albeit an educated guess, because you're not
calculating the true value of the data. You're calculating an estimate
of what the data may be or may not be.

>After all, just look at the results, where images produced by
>Bayer CFA encoded data are typically better than color film was
>ever able to accomplish!


Non sequitur.

Steve
 
Reply With Quote
 
 
 
 
TheRealSteve
Guest
Posts: n/a
 
      04-21-2012

On Wed, 18 Apr 2012 09:58:37 -0700, nospam <(E-Mail Removed)>
wrote:

>In article <(E-Mail Removed)>, TheRealSteve
><(E-Mail Removed)> wrote:
>
>> >are you actually saying a sensor has 3 different sampling rates because
>> >there are 3 different colour filters? you're *really* confused.

>>
>> Yes, that is almost exactly true. It has 2 different sampling rates.

>
>this shows how utterly clueless you are. there is *one* sampling rate
>for any given sensor.


Again, showing how utterly clueless you are. But I'll play along and
show an extreme case that proves you're wrong and leave it as an
exercise for you to prove to yourself that you're wrong for a bayer
cfa.

You said that there is *one* sampling rate for any given sensor. So
let's say you have a given 10MegaPixel sensor that has a color filter
array where every pixel except one samples light at a wavelength of
650nm (red) and a single pixel, say at the top right, that samples
light at 510nm (green).

Would you say this sensor has *one* sampling rate or possibly that it
just might have two sampling rates, one at 570nm and a different one
at 510nm?

Hint: the answer is two. If you say one, you're clueless.

Now that you know a given sensor can have two different sample rates
for the different colors of the color filter array, can you figure out
why a bayer cfa has different sample rates for the different colors of
it's color filter array?

I won't even give you a hint this time. But if you can't figure it
out, you're clueless.

I may be taking this too far but the next question, if you do happen
to come upon the correct answer (that the different colors of a bayer
cfa are sampled at different rates) is whether you can understand how
it's possible for the different sample rates of the colors of the cfa
to cause aliasing to begin at different input image resolutions for
each color of the cfa?

[snipped all of your clueless rants]

Steve
 
Reply With Quote
 
 
 
 
TheRealSteve
Guest
Posts: n/a
 
      04-21-2012

On Fri, 20 Apr 2012 18:29:29 +0200, Wolfgang Weisselberg
<(E-Mail Removed)> wrote:

>TheRealSteve <(E-Mail Removed)> wrote:
>> On Tue, 17 Apr 2012 20:40:34 +0200, Wolfgang Weisselberg
>>>TheRealSteve <(E-Mail Removed)> wrote:
>>>> On Mon, 16 Apr 2012 19:47:55 +0200, Wolfgang Weisselberg
>>>>>TheRealSteve <(E-Mail Removed)> wrote:

>
>>>>>> And finally, you need to understand how a 3 sensor camera system works
>>>>>> as opposed to a bayer filter. How a 3 sensor system uses 3
>>>>>> monochromatic sensors, each of which samples it's assigned color at
>>>>>> the full rate of the substrate under the filter while a bayer sensor
>>>>>> samples each of it's assigned colors at only 1/2 or 1/4 the rate of
>>>>>> the substrate under the filter.

>
>>>>>You're forgetting that a 3-sensor system has a rather narrow
>>>>>bandwith for each of the 3 colours. Bayer pattern sensors have
>>>>>a much wider bandwith in each colour filter.

>
>>>>>Which in turn means that bayer pixels pick up more signal ---
>>>>>especially about luminance. In the real world high frequency
>>>>>luminance detail does not matter much to humans; humans are not
>>>>>equipped to detect it.

>
>>>> Nonsense. It has nothing to do with the bandwidth of the color
>>>> filters.

>
>>>Of course it has. The wider the filter, the more signal it
>>>can pick up. Or would you say a bandpass filter of just one
>>>wavelength would let through the same amount of light as a
>>>filter so wide that all of the light can pass?

>
>> I'm not saying anything about the amount of light let through.

>
>So you agree that the green pixels in a bayer sensor also
>pick up red and blue signals?


The green pixels may pick up a little red and blue signals because
like AA filters, real world color filters aren't perfect. But just
because the green pixel may pick up a tiny bit of unaliased red and
blue light doesn't mean that the majority of the red and blue light
coming from the aliased red and blue pixels aren't going to cause
major problems in the final image.

>> All I'm
>> saying is that any one of the color channels can be aliased and data
>> from the other color channels cannot completely solve that problem
>> without tedius, time consuming and intesive work at guessing just how
>> to remove the aliased information.

>
>That's the same "teduous, time consuming and intensive work"
>needed to realign 3 sensors, to which you only had to say
>that computers are fast enough?


If you actially believe that the complexity of the algorithms to
realign 3 sensors (something similar to registax) is anything near the
complexity needed to remove artifacts caused by aliasing (which really
can not be perfectly removed no matter what you do algorithmically)
then you better cut back on the meds.

>>>> The transmittance of the color filters may have something to
>>>> do with it but that is likely the exact same or extrememly close for a
>>>> bayer cfa vs. monochrome filter.

>
>>>Ah --- no. A dichromatic filter is more efficient. But not
>>>wide enough for bayer.

>
>>>> If anything, a monochrome filter is
>>>> likely more transparent because it's easier to produce.

>
>>>Black paint is easier to produce than a filter, is it more
>>>transparent? No? Then you have found your logic fails.

>
>> Flaws in your analogies don't equate to flaws in my logic.

>
>Flaws in your logic brought to their ultimate conclusion do
>not equate to flaws in my analogies.


But since your analogy is flawed it proves nothing about my logic.

>>>> And also, specious when it comes to what humans are or are not able to
>>>> detect. We're talking about aliasing artifacts like moire, which can
>>>> be detected easily by the human eye even if the real world high
>>>> frequency detail that caused the aliasing is not.

>
>>>Visible aliasing artifacts are low frequency details.

>
>>>The correct solution, then, is to use a properly blurry enough
>>>signal so all those high frequency details that can produce
>>>aliasing with the sensor sampling frequency are washed away.

>
>> Or, have a sensor resolution great enough that an AA filter isn't
>> necessary because the other parts of the camera system (like the lens,
>> focusing, etc.) can't resolve the high frequency details enough to
>> cause aliasing. That is the best of both worlds... you have the
>> maximum resolution that the rest of the camera system is capable of
>> without the blurring of an AA filter that reduces the resolution to
>> lower than what the rest of the camera system is capable of.

>
>Which does not preclude a Bayer system from being that
>sensor.


You finally said something that's correct.

[snipped some more wrong things of yours]

>> There are still
>> frequencies of light in the original image that can be aliased enough
>> in one color channel to cause visible artifacts in the final image,
>> but not in the other color channels.

>
>Q: does that happen in the real world or only in specially
>prepared experiments?


Yes, it happens in the real world if the camera system (lens,
focusing, etc.) is able to resolve well enough that the resolution of
the image projected on the sensor exceeds nyquist, even with the
imperfect AA filter.

Getting "technical", real world images, like that suit jacket that had
the moire pattern or maybe a picture of the crowd at a baseball game,
have a very widely distributed range of spatial frequencies across the
wideband temporal frequencies of light. Because you're sampling
different, narrower temporal frequency bands at different spatial
rates, the amount of aliased artifacts can be different for the
different temporal frequency bands.

Assuming that the sampled temporal bands have similar spatial
frequency content (a very good assumption for the baseball crowd,
maybe not so good for a field of green grass) the temporal bands that
are sampled at higher spatial rate will have less aliasing present
than those temporal bands that are sampled at a lower spatial rate
because less of the spatial frequency content will be aliased at the
higher sample rate.

[...]
>>>> Interpolating between not only 2x2 but 3x3 or more
>>>> sensels of a bayer sensor also dampens high frequency detail.

>
>>>Since the interpolation is not simply averaging --- as you
>>>seem to think for some reason --- your argument is not as
>>>valid as you think it is.

>
>> I never said it was simply averaging. Although it's close to that but
>> with different weights associated with different adjacent pixels.

>
>With different *dynamic* weights.


Doesn't matter whether weights are dynamic or not when it comes to
aliasing. In fact, the eitner algorithm doesn't really matter either
because none of them can remove the aliasing artifacts. The
differences are in how the artifacts are presented.

>> What
>> I'm saying is that it doesn't matter what the interpolation algorithm
>> is.

>
>Actually, it does matter.


Nope, it doesn't when it comes to whether the image has aliasing
artifacts. Only how they are presented.

>> If it's using sample data that is aliased to come up with
>> estimates of the missing data, the resultant estimates will also be
>> affected by the aliased data.

>
>And that's as true for Bayer as for Foveon as for 3-sensor as
>for any other sampling system. A proper low-pass filter
>guards against that problem.


Correct, except for one thing. A Fovean or 3-sensor system will have
less aliasing than a bayer sensor of the same spatial resolution.

>>>> However, all this discussion is leading far away from the original one
>>>> about aliasing.

>
>>>3-sensor systems are just as perceptible for aliasing, though
>>>they're usually not showing colour moire.

>
>> They're not "just as" susceptible to aliasing.

>
>They're just as.


No, they are not. If you believe they are then you don't understand
how the bayer sensor is sampled vs. a Fovean or 3-sensor system.

>> They are less
>> susceptible to aliasing although they still are susceptible.

>
>I can find an aliasing frequency for any given sensor.
>See, I can play your "but then I can buy 3 sensors instead of one"
>too ...


And the aliasing frequency you find for a bayer sensor is lower than
an equivalent (in terms of sensor size and pixel density) 3 sensor
system or fovean sensor. You can play whatever games you want but you
can't get away from that point.

>> And the
>> fact that 3 sensor systems usually don't show color moire but instead
>> show luma moire should give you a clue as to why they are overall less
>> susceptible to aliasing.

>
>They're just as susceptible. It only expresses itself differently.


No they are not just as susceptible. Obviously the clue hasn't led you
to the correct answer.

Steve
 
Reply With Quote
 
TheRealSteve
Guest
Posts: n/a
 
      04-21-2012

On Fri, 20 Apr 2012 20:29:29 -0700, nospam <(E-Mail Removed)>
wrote:

>In article <(E-Mail Removed)>, TheRealSteve
><(E-Mail Removed)> wrote:
>
>> >> >> The missing data from the bayer sensor is estimated.
>> >> >
>> >> >no it's not. it's calculated.
>> >>
>> >> No, it's not. It's estimated.
>> >
>> >you're playing word games.
>> >
>> >furthermore, if the answer is correct, then it doesn't matter how you
>> >arrived at it.

>>
>> But the answer will not be correct. It may be "close enough" for the
>> image quality *you* desire. But it will likely never be correct. And
>> even if it were correct, you would never know whether it was correct
>> or not. It's not word games. It's reality.

>
>word games.


Actually, you're playing word games. I'm just stating what's going on.

>
>> >> You won't find a single reputable
>> >> reference on bayer cfa demosaicing that says the values for the
>> >> missing colors are anything more than an estimate.
>> >
>> >oh yes i will, and in fact, many of them provide the formulas used in
>> >the calculations.

>>
>> The provided formulas are used to ESTIMATE the missing colors. Find me
>> a single reputable reference that says it can come up with the true
>> values of the amount of red or blue light that fell on a green sensor,
>> or any other missing data combination. They can't. Thay can only
>> estimate the missing colors.

>
>more word games.


It's not word games, it's reality,

>nothing can come up with the 'true value' no matter what you do, and if
>you think anything can, you're *really* confused. there is *always*
>going to be a certain amount of error, even in your beloved 3 sensor
>system.


Which is my point exactly! If you want to know what 2+2 is, you can
calculate it and come up with the true and correct answer, 4. If you
want to come up with the amount of red light on a green sensel, you
can not calculate that answer no matter what you do. You can only come
up with an estimate.

>> If you want to play word games, what is really done not is to
>> calculate the missing colors, but to calculate an estimate of the
>> missing colors. I.e., to estimate. Which is what I said when I said
>> the missing colors are estimated.

>
>even more word games.


Again, not word games at all. It's reality. The missing colors are
estimated, and that's a fact you can't deny no matter how hard you
try. And if you do deny it, you're denying facts just as solid as
2+2=4.

You are playing word games that try to imply that the values of the
missing colors can somehow be calculated just like 2+2 can be
calculated. But they can not be. They can only be estimated.

Steve
 
Reply With Quote
 
TheRealSteve
Guest
Posts: n/a
 
      04-21-2012

On Fri, 20 Apr 2012 20:29:36 -0700, nospam <(E-Mail Removed)>
wrote:

>In article <(E-Mail Removed)>, TheRealSteve
><(E-Mail Removed)> wrote:
>
>> >> An equivalent 3 sensor system will have 3 times the number of pixels
>> >> than the bayer cfa. It will, of course, be more expensive. But it's
>> >> that expense that buys you the extra resolution that you can't get
>> >> with the bayer sensor at whatever pixel density you choose as the
>> >> current technological limit. I.e., if current technology limits the
>> >> resolution of a monochrome sensor to X, A bayer sensor will have an
>> >> overall resolution of X but the individual color channels will be
>> >> sampled at 1/2 X for green and 1/4 X for red and blue. An equivalent 3
>> >> sensor system using 3 of the same sensors as the bayer sensor that is
>> >> at the current technological resolution limit will have a resolution
>> >> of X for each of the color channels. That's why 3 sensor systems are
>> >> used, to get greater resolution in the color channels without having
>> >> to go beyond the technological limit of sensor resolution.
>> >
>> >the only difference is higher chroma resolution and you can't see that
>> >difference because of the way the human eye works, which is why high
>> >end video cameras are moving away from 3 sensor systems.

>>
>> You keep bring up the human eye when it has nothing to do with
>> aliasing the color channels. Do you really thing that because a sensor
>> may have higher resolution than a human eye that aliasing is not
>> possible?

>
>two separate things. having 3 sensors gives you better chroma
>resolution which you can't see so why bother capturing it. aliasing is
>a different issue.


You are still totally confused. It's been explained to you many times
how aliasing can cause artifacts you can see from resolutions you
cannot see. You just don't seem to get it.

Time for you to go back to sampling theory 101.
 
Reply With Quote
 
nospam
Guest
Posts: n/a
 
      04-21-2012
In article <(E-Mail Removed)>, TheRealSteve
<(E-Mail Removed)> wrote:

> >nothing can come up with the 'true value' no matter what you do, and if
> >you think anything can, you're *really* confused. there is *always*
> >going to be a certain amount of error, even in your beloved 3 sensor
> >system.

>
> Which is my point exactly! If you want to know what 2+2 is, you can
> calculate it and come up with the true and correct answer, 4. If you
> want to come up with the amount of red light on a green sensel, you
> can not calculate that answer no matter what you do. You can only come
> up with an estimate.


so your 3 sensor system isn't any better than bayer. got it.

> >> If you want to play word games, what is really done not is to
> >> calculate the missing colors, but to calculate an estimate of the
> >> missing colors. I.e., to estimate. Which is what I said when I said
> >> the missing colors are estimated.

> >
> >even more word games.

>
> Again, not word games at all. It's reality. The missing colors are
> estimated, and that's a fact you can't deny no matter how hard you
> try. And if you do deny it, you're denying facts just as solid as
> 2+2=4.
>
> You are playing word games that try to imply that the values of the
> missing colors can somehow be calculated just like 2+2 can be
> calculated. But they can not be. They can only be estimated.


it's word games.

the data is precisely calculated and with a known error that is nearly
always imperceptible.
 
Reply With Quote
 
nospam
Guest
Posts: n/a
 
      04-21-2012
In article <(E-Mail Removed)>, Alfred
Molon <(E-Mail Removed)> wrote:

> > >Due to non-aligned sensors, it won't.

> >
> > If the sensors are aligned, it will. If they are not aligned and
> > instead are corrected in software,

>
> In fact a simple software calibration step, i.e. finding the correct
> mapping functions between the three sensors, will deliver perfect
> alignment, at the minimal cost of losing a few border pixels, which is
> irrelevant in a sensor with 4000 or 6000 pixels per row.


very hypocritical of you in suggesting a software calibration step that
fills in missing data due to misalignment and then reject bayer doing
essentially the same thing.
 
Reply With Quote
 
nospam
Guest
Posts: n/a
 
      04-21-2012
In article <(E-Mail Removed)>, TheRealSteve
<(E-Mail Removed)> wrote:

> >data on successive occasions. But in fact it is calculated,
> >and the same data will always produce the same result because
> >there is no guess work by the algorithm.

>
> Just because the result is repeatable doesn't mean it's not a guess.


if it's repeatable, it's *not* a guess.
 
Reply With Quote
 
nospam
Guest
Posts: n/a
 
      04-21-2012
In article <(E-Mail Removed)>, TheRealSteve
<(E-Mail Removed)> wrote:

> >> If it's using sample data that is aliased to come up with
> >> estimates of the missing data, the resultant estimates will also be
> >> affected by the aliased data.

> >
> >And that's as true for Bayer as for Foveon as for 3-sensor as
> >for any other sampling system. A proper low-pass filter
> >guards against that problem.

>
> Correct, except for one thing. A Fovean or 3-sensor system will have
> less aliasing than a bayer sensor of the same spatial resolution.


not true.

> >>>> However, all this discussion is leading far away from the original one
> >>>> about aliasing.

> >
> >>>3-sensor systems are just as perceptible for aliasing, though
> >>>they're usually not showing colour moire.

> >
> >> They're not "just as" susceptible to aliasing.

> >
> >They're just as.

>
> No, they are not. If you believe they are then you don't understand
> how the bayer sensor is sampled vs. a Fovean or 3-sensor system.


they're just as susceptible and it's clear you don't understand this as
much as you think you do.

> >> They are less
> >> susceptible to aliasing although they still are susceptible.

> >
> >I can find an aliasing frequency for any given sensor.
> >See, I can play your "but then I can buy 3 sensors instead of one"
> >too ...

>
> And the aliasing frequency you find for a bayer sensor is lower than
> an equivalent (in terms of sensor size and pixel density) 3 sensor
> system or fovean sensor. You can play whatever games you want but you
> can't get away from that point.


yes you can, since it's false.
 
Reply With Quote
 
TheRealSteve
Guest
Posts: n/a
 
      04-22-2012

On Sat, 21 Apr 2012 07:25:36 -0700, nospam <(E-Mail Removed)>
wrote:

>In article <(E-Mail Removed)>, TheRealSteve
><(E-Mail Removed)> wrote:
>
>> >nothing can come up with the 'true value' no matter what you do, and if
>> >you think anything can, you're *really* confused. there is *always*
>> >going to be a certain amount of error, even in your beloved 3 sensor
>> >system.

>>
>> Which is my point exactly! If you want to know what 2+2 is, you can
>> calculate it and come up with the true and correct answer, 4. If you
>> want to come up with the amount of red light on a green sensel, you
>> can not calculate that answer no matter what you do. You can only come
>> up with an estimate.

>
>so your 3 sensor system isn't any better than bayer. got it.


Your "logic" evades reason. Complete non-sequitur. I have no idea how
you got that from the statement above.

>> >> If you want to play word games, what is really done not is to
>> >> calculate the missing colors, but to calculate an estimate of the
>> >> missing colors. I.e., to estimate. Which is what I said when I said
>> >> the missing colors are estimated.
>> >
>> >even more word games.

>>
>> Again, not word games at all. It's reality. The missing colors are
>> estimated, and that's a fact you can't deny no matter how hard you
>> try. And if you do deny it, you're denying facts just as solid as
>> 2+2=4.
>>
>> You are playing word games that try to imply that the values of the
>> missing colors can somehow be calculated just like 2+2 can be
>> calculated. But they can not be. They can only be estimated.

>
>it's word games.
>
>the data is precisely calculated and with a known error that is nearly
>always imperceptible.


Now you're playing the word games. Precisely calculated with an known
error. You're not making sense. If it's precisely calculated, it
wouldn't have an error, known or otherwise. And if there is a known
error, it can be eliminated. The problem is that the error is unknown
within a statistical bounds. That's why it's only an estimate. You
either don't understand simple concepts or you're just playing games.

Steve
 
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
Re: The death of the Bayer filter? Maybe not. Bruce Digital Photography 3 04-04-2012 08:04 PM
Death, even when expected, is hard for anyone to deal with, but asuicidal death asim malik Perl Misc 1 09-13-2009 07:42 AM
Re: Crocodile Hunter - DEATH VIDEO - Steve Irwin - crocodile hunter - death video - steve irwin.exe (1/1) Mikey DVD Video 3 09-14-2006 05:52 PM
VoIP ebay & skype = DEATH (Offtopic ? Maybe) codefixer@gmail.com Cisco 0 09-08-2005 09:27 PM
Of Intrest ? Maybe Yes, Maybe No Old Gringo Computer Support 0 10-11-2004 01:35 AM



Advertisments