Welcome to City-Data.com Forum!
U.S. CitiesCity-Data Forum Index
Go Back   City-Data Forum > General Forums > Photography
 [Register]
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
View detailed profile (Advanced) or search
site with Google Custom Search

Search Forums  (Advanced)
Reply Start New Thread
 
Old 02-12-2012, 12:01 PM
 
Location: Not far from Fairbanks, AK
20,275 posts, read 37,049,222 times
Reputation: 16391

Advertisements

Hmmm....there is something wrong here. mathjak107 started this thread, but he has not engaged in the arguments (discussion???) posted after.
Reply With Quote Quick reply to this message

 
Old 02-12-2012, 12:03 PM
 
106,242 posts, read 108,257,613 times
Reputation: 79786
Quote:
Originally Posted by RayinAK View Post
Hmmm....there is something wrong here. mathjak107 started this thread, but he has not engaged in the arguments (discussion???) posted after.
To be honest i dont know enough about all the ins and outs of the anti-aliasing filters. i have a basic idea but not enough to debate. im just letting ya'll know the new cameras are out in 2 versions.

to me the proof is in the pudding. we would have to see identical shots with both cameras to see whats what.
Reply With Quote Quick reply to this message
 
Old 02-12-2012, 12:05 PM
 
Location: Not far from Fairbanks, AK
20,275 posts, read 37,049,222 times
Reputation: 16391
Quote:
Originally Posted by mathjak107 View Post
To be honest i dont know enough about all the ins and outs of the anti-aliasing filters. i have a basic idea but not enough to debate. im just letting ya'll know the new cameras are out in 2 versions.

to me the proof is in the pudding. we would have to see identical shots with both cameras to see whats what.
Yes, that would make sense.
Reply With Quote Quick reply to this message
 
Old 02-12-2012, 01:46 PM
 
Location: Barrow, Alaska
3,539 posts, read 7,630,167 times
Reputation: 1836
Quote:
Originally Posted by mathjak107 View Post
to me the proof is in the pudding. we would have to see identical shots with both cameras to see whats what.
Very true! And it will be really interesting over the next year or so to see some of the practical demonstrations of various aspects that we've only been able to theorize about before. Not just the specific example of having one really high MP camera that is available with and without the anti-aliasing filter; but also because we now have high quality sensors all the way from 12 MP up to 36 MP to compare.

For a different venue I worked up an example that is interesting, which brings up just exactly which comparisons will be interesting. Absent a lot of detail, let me try to describe just the essential part.

Lets say we have a need to shoot some particular scene which exhibits a lot of aliasing distortion (color moiré for example) and it is determined that the required resolution for this image is 14 MP. That is, a 4600x3066 pixel image, which at 300 PPI would print at 15.3x10.2 inches. There are two basic high quality ways to get such an image. One is to use a 14 MP sensor that uses an appropriate anti-aliasing filter. A second method is called "over sampling", and that means using a camera with a higher resolution sensor and an appropriate anti-aliasing filter, and then downsampling the results to the desired size. What is the difference between those two methods?

A 4600x3066 image has a maximum detail of about 64 line pairs per millimeter horizontally. If we use a 14 MP camera the AA filter, because real "brick wall" filters do not exist, will need perhaps a -6 dB response at 64 lp/mm to insure no artifacts from detail above that frequency, and the filter might have a -3 dB cutoff at 44 lp/mm and have a 3 dB per 20 lp/mm rolloff. What that means is that detail at 24 lp/mm will not be affected at all, at 44 lp/mm the contrast will be reduced by 1/2 and at 64 lp/mm contrast will be reduced by 1/4. That is the range of "real" detail that can be recorded.

Everything above that frequency will show up as aliasing distortion, which is what causes moiré. So it is reduced to 1/4 at the Nyquist Limit, but detail from 64 to 84 lp/mm causes aliasing at levels between 1/4 and 1/8th of what the real detail is. It will be aliased to the range from 44 to 64 lp/mm in the captured image. If sharpening is used to restore the correct levels of real detail in that range, the noise is increased right along with the desired signal. So a boost of 6 dB for the signal puts the alias distortion noise down at 44 lp/mm all the way back up to only 6 dB down from signal rather than 12 dB, and at 64 lp/m the noise is equal to the real signal. (All of which explains why we do sometimes see pretty bad moiré even in cameras with AA filters.)

But what happens if we shoot the scene with a 36 MP sensor, using a similar AA filter (3 dB per 20 lp/mm with the 3 dB cutoff located 20 lp/mm below the sensor's Nyquist Limit), and then resample to 14 MP?

The 36 MP image is about 7400x4933, so the image has about 103 line pairs per mm for a Nyquist Limit. Lets use 104 just to make the numbers easy. The AA filter described above would have 0 dB effect at 64 lp/mm, a -3 db cutoff at 84 lp/mm and would reduce 104 lp/mm detail by 6 dB. Note that the AA filter will have no effect at all on the real detail we are interested in at 64 lp/mm and below, and therefore no sharpening is required to offset the loss from the AA filter. But even more important, the frequencies that would be aliased down to 64 lp/mm and below start at (104 - 64) above the 104 lp/mm Nyquist Limit, or at 144 lp/mm. The AA filter is 6 dB down at 104mm and rolls off another 6 dB by 144 lp/mm (3 db per 20 lp/mm). So the frequencies that would be aliased down to 64 lp/mm are reduced 12 dB below the real detail at 64 lp/mm.

From the 14 MP sensor, noise at 44 lp/mm is 6 dB below signal and at 64 lp/mm is not only equal, but all non-aliased noise has also been boosted 6 dB by sharpening. From the 36 MP sensor noise at 44 lp/mm is 15 dB below signal, at 64 lp/mm it is 12 dB below signal, and other noise has not been boosted by sharpening.

Clearly a higher MP sensor has advantages for making smaller prints cleaner. That is as opposed to the usual consideration of making large prints sharper by using an higher MP sensor.
Reply With Quote Quick reply to this message
 
Old 02-12-2012, 03:31 PM
 
106,242 posts, read 108,257,613 times
Reputation: 79786
reminds me of when i was into high end audio equipment back in the 1970's.

the first cd players were released. the specs were perfection.but the ultimate test ,our ears said something was wrong.

the sound was fatiguing,flat and 2 dimensional yet the specs were as perfect as can be.

well as technology improved the problem turned out to be these brick wall filters in the digital to analog convertors were creating all kinds of nasty new distortions but up until now we had no way to measure them.

once they were able to identify them they began to work them out and used much more gentle filters the result were the best players yet.

time and comparisons will tell if these new cameras are the equal to early cd players and while specmanship looks great lets see how they perform.
Reply With Quote Quick reply to this message
 
Old 02-12-2012, 06:32 PM
 
106,242 posts, read 108,257,613 times
Reputation: 79786
Quote:
Originally Posted by Floyd_Davidson View Post
Until you absolutely need to get a shot of something that produces horrible moire'.

At that point you'll have to admit to not being capable of "using our cameras for taking good shots and perfecting our skills" unless you've spent some time discussing and learning all these technical topics that are prerequisites for making good photographs at will, rather than by accident.

Just having fun and making a hobby of photography is wonderful, and certainly to be encouraged. But the purpose of this forum includes serious photography too, whether for amateur or professional work.

Disparaging discussion because it is above your current level lacks class.
to tell you the truth i disagree. there is a certain amount of knowledge thats needed but the deep dark technical stuff of how things may or may not work on a digital level is no more needed to produce great photography than knowing how every aspect of how a baby is formed is needed to be good parents.
Reply With Quote Quick reply to this message
 
Old 02-12-2012, 07:44 PM
 
Location: Moon Over Palmettos
5,978 posts, read 19,865,269 times
Reputation: 5102
^^^ Nailed it, Mathjak!
Reply With Quote Quick reply to this message
 
Old 02-12-2012, 08:47 PM
 
Location: Barrow, Alaska
3,539 posts, read 7,630,167 times
Reputation: 1836
Quote:
Originally Posted by mathjak107 View Post
to tell you the truth i disagree. there is a certain amount of knowledge thats needed but the deep dark technical stuff of how things may or may not work on a digital level is no more needed to produce great photography [...]
Of course it isn't absolutely necessary in order "to produce great photography". But the consistency and the breadth of "great photography" is directly associated with how deep and how broad the photographer's technical knowledge is.

Great photographs can happen at random and be accidental. It's like monkeys with typewriters producing a line or two from Shakespeare! Except the odds are much better for photographs. Anyone can take a half decent camera and at some point in the next 10,000 exposures there will be a few dozen or so images that have terrific emotional appeal to most people. "Great photography", eh? That may be a few great photographs, but it is neither great photography nor a great photographer.

But it is absolutely good enough for most people who use photography as a hobby. Being able to print a couple dozen really attractive images a year is very satisfying!

But the mark of a great photographer is being able to produce great photographs at will. That is great photography!

If you want to get a specific great photograph, which might be one that pushes the limits of your camera, it's like I said in the previous post: you either have already studied and understand the technical aspects required, or you simply have to admit you can't do that shot. Randomly producing great photographs isn't the point because what is required is efficient work that produces precisely the needed results without wasting time waiting for a random great photograph to happen.

Quote:
than knowing how every aspect of how a baby is formed is needed to be good parents.
That is a particularly unfortunate analogy. Children die because parents are unaware of "how a baby is formed". Or they are damaged, and suffer for the rest of their life, because parents are not fully aware...
Reply With Quote Quick reply to this message
 
Old 02-12-2012, 09:01 PM
 
Location: Barrow, Alaska
3,539 posts, read 7,630,167 times
Reputation: 1836
Quote:
Originally Posted by bibit612 View Post
^^^ Nailed it, Mathjak!
Think about really great photographers... Adams, Lange, Eisenstadt, Weston, Gowland, Shaw, Cartier-Bresson, ... whatever field. There is one single common characteristic that connects all of these photographers that stand out as the best: they were all extremely technically knowledgeable. They wrote books about it, they designed their own cameras, and they drove their technical staff up the wall with it.

But you all say that isn't important?
Reply With Quote Quick reply to this message
 
Old 02-13-2012, 02:06 AM
 
106,242 posts, read 108,257,613 times
Reputation: 79786
they are technical in their photography techniques and understood the things they had to understand to crank out great photographs. they werent camera designers or electronic engineers. they were photographers.

understanding the mechanics of what goes on in a chip or even nyquist theory isnt a pre-req for cranking out great photos.

understanding the controls on your camera, lighting ,composition are the the things that make great pictures , understanding the design of rgb sensors and how the photons move on a digital level is not.

im a pretty technical guy but i find no use for alot of what goes on in the inner workings of a digital camera ,all i care about are the results it produces and what those benefits mean to me . .

you may have an interest in knowing the engineering side but ill say it again. most great photographers couldnt care less, all they care about is the results of those benefits .

Last edited by mathjak107; 02-13-2012 at 02:24 AM..
Reply With Quote Quick reply to this message
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.

Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.


Reply
Please update this thread with any new information or opinions. This open thread is still read by thousands of people, so we encourage all additional points of view.

Quick Reply
Message:

Over $104,000 in prizes was already given out to active posters on our forum and additional giveaways are planned!

Go Back   City-Data Forum > General Forums > Photography
Similar Threads

All times are GMT -6. The time now is 12:44 AM.

© 2005-2024, Advameg, Inc. · Please obey Forum Rules · Terms of Use and Privacy Policy · Bug Bounty

City-Data.com - Contact Us - Archive 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37 - Top