Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
12-04-2015, 03:50 PM   #1
New Member




Join Date: Sep 2012
Posts: 11
Is a non AA filtered sensor an advantage?

Hi,





I would like to know whether is the difference between the same sensor one with and one without antialiasing filter evident, especially in macro photography?

Thank you

12-04-2015, 05:06 PM   #3
Site Supporter




Join Date: Sep 2012
Location: Raleigh, NC, USA
Posts: 755
Thanks. I still wish I had gotten the k3 instead of the k5iis.
12-04-2015, 05:08 PM   #4
Administrator
Site Webmaster
Adam's Avatar

Join Date: Sep 2006
Location: Arizona
Photos: Gallery | Albums
Posts: 40,978
QuoteOriginally posted by Chosa Quote
Hi,





I would like to know whether is the difference between the same sensor one with and one without antialiasing filter evident, especially in macro photography?

Thank you
The difference is certainly appreciable. Newer Pentax cameras with the AA filter simulator mean that there is no good reason not to go for a filterless sensor.


Adam
PentaxForums.com Webmaster (Site Usage Guide | Site Help | My Photography)



PentaxForums.com's high server and development costs are user-supported. You can help cover those costs by donating. Or, buy your photo gear from our affiliates, Adorama, B&H Photo, or Topaz Labs, and get FREE Marketplace access - click here to see how! Trusted Pentax retailers:

12-04-2015, 06:07 PM   #5
Loyal Site Supporter
baro-nite's Avatar

Join Date: Aug 2011
Location: North Carolina, USA
Photos: Gallery | Albums
Posts: 4,298
QuoteOriginally posted by Adam Quote
Newer Pentax cameras with the AA filter simulator mean that there is no good reason not to go for a filterless sensor.
What about flash photography -- is the filter simulator still effective?
12-04-2015, 06:25 PM   #6
Administrator
Site Webmaster
Adam's Avatar

Join Date: Sep 2006
Location: Arizona
Photos: Gallery | Albums
Posts: 40,978
QuoteOriginally posted by baro-nite Quote
What about flash photography -- is the filter simulator still effective?
Good question- easy enough to test! I guess it would depend on how the sensor is vibrated.

Adam
PentaxForums.com Webmaster (Site Usage Guide | Site Help | My Photography)



PentaxForums.com's high server and development costs are user-supported. You can help cover those costs by donating. Or, buy your photo gear from our affiliates, Adorama, B&H Photo, or Topaz Labs, and get FREE Marketplace access - click here to see how! Trusted Pentax retailers:

12-05-2015, 05:03 AM   #7
Senior Member
fb_penpho's Avatar

Join Date: Jul 2009
Location: London
Photos: Gallery
Posts: 183
Superb question

QuoteOriginally posted by baro-nite Quote
What about flash photography -- is the filter simulator still effective?
Given the duration of the flash, there is a very good chance that on some/many occasions it copuld have any effect. Now having a K-S2, I have only once come across a pic that showed the moire effect. Then again I rarely use flash, so for me is unlikely to be a problem. Seeing the filter simulator uses the mechanical movement of the sensor to create the effect, then there is a lower upper limit of duration above which the filter simulator will not work. Suggests to me that possibly professional users who take mainly flash shots should take this issue into account with their selection. But these are the very people who might have the software that can remove it in the post processing. This is not just a Pentax problem as I believe all suppliers offer non AA sensors.

Last edited by fb_penpho; 12-05-2015 at 05:21 AM.
12-05-2015, 07:06 AM   #8
New Member




Join Date: Sep 2012
Posts: 11
Original Poster
Why does the flash can produce problems without aa?

12-05-2015, 07:21 AM   #9
Pentaxian




Join Date: Dec 2010
Location: Ontario
Photos: Gallery
Posts: 2,285
QuoteOriginally posted by Chosa Quote
Why does the flash can produce problems without aa?
He's asking about the K-3's mechanical simulation of an AA filter. The sensor has no physical AA filter, but is capable of vibrating during exposure to help reduce moire. A possible problem is the extremely short duration of some flash bursts and whether or not it's ever short enough that the sensor doesn't manage to vibrate during the exposure. There's more info about the AA filter simulator here: K-3 | RICOH IMAGING

They do say "The AA-filter effect is more evident when a shutter speed of 1/1000 second or slower is used." so I'd guess it could be an issue for some setups as flash duration can be much shorter than 1/1000s, but I don't have a k-3 to test.

With the AA filter simulator turned off (or a camera with no AA-filter of any kind like the k5iis), it's business as usual with flashes.


Also, I've used a k5iis for macro stuff for a couple of years and it does an outstanding job. Mostly nature stuff, but I've noticed moire I think one time in an insects eye and it was easy to fix in post. I've no comparison with the AA-filter versions of this camera, but given the lack of moire I've encountered, I wouldn't hesitate to choose AA-free in the future.

Last edited by BrianR; 12-05-2015 at 07:27 AM.
12-05-2015, 08:13 AM   #10
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 8,937
QuoteOriginally posted by Chosa Quote
would like to know whether is the difference between the same sensor one with and one without antialiasing filter evident, especially in macro photography?
A Bayer-AA filter (or equivalent means of avoiding aliasing) is a necessary component in a camera.

Camera companies have been employing Bayer-AA filters for a good reason -- despite the increase in manufacturing cost this implies -- and only recently have been giving in to user demand.

It is incorrect to state (as the linked article does) that Bayer-AA filters "can even prevent some minute details from coming through". On the contrary, the aliasing that occurs with filterless designs destroys detail that can be recovered with deconvolution sharpening from a intentionally blurred capture.

All the seemingly improved acuity that looks like additional detail being captured by a filterless sensor is false detail. This can easily be seen by looking at captures of Siemens stars where a reconstructed image from an intentionally blurred capture retains detail for longer. The reason for this is basic information theory. The sensor cannot capture spatial frequencies beyond its Nyquist limit and if you let it anyhow (by not using a filter), you get false detail that looks "acute" but does not represent the original.

The only advantage of a filterless design is that it does not require capture sharpening. When one capture-sharpens the image from an intentionally blurred capture, one also sharpens image noise. So for images with some noise, e.g., shot at higher ISO settings, (and ideally some natural blur), a filterless design could be better.

It is sad that filterless cameras have become so popular because it means (at least in the case of Pentax) that the technically correct version (with a filter) is no longer available. When people don't see false detail, false colour, and colour moiré when using their filterless cameras, it is because their photos aren't sharp enough. Slight defocus, stopping down (diffraction blur), hand-shake, etc. can all be natural causes for image blur that obviate the need for a Bayer-AA filter.

In macro photography, one often deals with minute details, such as the facets of compound eyes, that can cause aliasing to occur. I'd say next to fashion photography (or other occasions where one could encounter highly regular textile patterns), I'd probably would want a Bayer-AA filter the most for macro photography.

QuoteOriginally posted by Adam Quote
Newer Pentax cameras with the AA filter simulator mean that there is no good reason not to go for a filterless sensor.
Well, the simulator becomes ineffective at fast shutter speeds (starting at 1/1000s).

Furthermore, the simulator's effect is not constant over the whole frame like that of a regular birefringent crystal, once the shutter speed exceeds the sync-speed (e.g., > 1/180s). Instead, it varies from top to bottom of the frame and interacts with camera and subject movement. This dynamic blurring cannot be undone as easily as the regular blurring of a birefringent crystal, i.e., an optimal capture sharpening will not be possible.

As long as the shutter speed does not exceed the sync speed and there is absolutely no movement then the simulator should be able to provide excellent captures. In other cases, I'd rather have a proper filter than the simulator replacement.

QuoteOriginally posted by baro-nite Quote
What about flash photography -- is the filter simulator still effective?
Not if the the flash duration becomes too short for the simulator to be effective.

If the simulator starts to become ineffective from 1/1000s upwards, that means that using, say a Metz 58 AF-2 at 1/4 power (or less) will cause the typical artefacts of a filterless capture to appear.

Last edited by Class A; 12-05-2015 at 08:19 AM.
12-05-2015, 09:31 AM   #11
Pentaxian
The Squirrel Mafia's Avatar

Join Date: Jan 2015
Posts: 648
QuoteOriginally posted by condor27596 Quote
Thanks. I still wish I had gotten the k3 instead of the k5iis.
The K-5IIs is an amazing camera. It pulled the most detail out of that 16MP Sony sensor, but at the same time it kept file sizes smaller than what the K-3 can do.

I would love to have a K-5IIs, but unfortunately they're going for an arm & a leg on eBay & I refuse to pay an arm & a leg for an older camera when the newer K-3 costs about the same or even less & has newer technology. But yeah. That K-5IIs is still very awesome. The Nikon D7000 also used the same sensor & is equally awesome, but same problem. They cost too much on eBay.
12-05-2015, 10:01 AM   #12
Pentaxian




Join Date: Dec 2010
Location: Ontario
Photos: Gallery
Posts: 2,285
QuoteOriginally posted by Class A Quote
In macro photography, one often deals with minute details, such as the facets of compound eyes, that can cause aliasing to occur. I'd say next to fashion photography (or other occasions where one could encounter highly regular textile patterns), I'd probably would want a Bayer-AA filter the most for macro photography.
This is my moire example from my k5iis. It's a 100% crop with the full image as an insert, dfa100mm @f/8, 1/30s, mirror lockup, sheltered from wind, etc. (processing was the lazy "Lightroom default" from raw...this ended up not being a keeper, for reasons other than the moire). For me at least, things virtually never line up in a way that compound eyes cause an issue, and I don't think I've ever seen a macro example with a problem over large areas like you may have with fabric. I've never seen any hint of moire when the eyes take up a good portion of the frame, in these cases the pattern isn't nearly so fine on the sensor to cause problems.

To be fair, I spend most of my time at f/11 to f/16 where diffraction will take that edge off with a little bit of blur in there (not lighting via low powered flashes was a rarity though). If I ever get around to getting into focus stacking so I can stick with the sharpest apertures, maybe I'll sing a different tune.

The advantage of the k5iis at the time I purchased it was largely availability over the k5ii. But I haven't run into any practical concerns with having no AA-filter.
Attached Images
 
12-05-2015, 12:11 PM   #13
Loyal Site Supporter




Join Date: Sep 2006
Location: Texas
Photos: Gallery | Albums
Posts: 23,767
The difference in my old K5 and my newer K5IIs seems to be in AF accuracy and maybe a little more sharpness. Looking at shots from both cameras, it is often hard for me to find much difference in IQ, so how much is real and how much is in my mind is questionable? The lack of the filter has not presented any problems for me, so I assume that any improvement in IQ is a free benefit.

The smaller file sizes is my only benefit over the K3...if you can call it a benefit?

Regards!
12-05-2015, 05:21 PM   #14
Pentaxian




Join Date: Nov 2013
Posts: 4,549
As measured by tools and reviewers using standard procedures, a filterless design is sharper than a design with ah low pass filter. The gain is quite significant. On DxO for example there more difference between K5 and K5-IIs than K5-IIs and K3 despite the K3 being 24MP. The gain seems to be 10-20% resolution.

I don't adhere to the whole theory stuff of false detail because otherwise the measurement tests would not conclude that actually the sensor can allow to distinguish more line pairs.

Now on some occasion you can have moire. On 24MP this is not that often. You need a high quality lens with a subject with the proper pattern. Moire is mostly visbible at 100% viewing size and a low pass filter can be applied in post processing. Contrary to the hardware low pass filter this one is configurable in DxO and I would gess Lightroom too.

In practical term a filterless sensor will provide sharper results, in most occasion there will be no moire and in the rare occasion were it is present you can remove it in post processing. From time to time you could encounter a subject that would not correct well in post processing. Honestly this never occured to me and I have never seen somebody in that situation.

To me that's the difference between theory and practice. In practice it yield better results.
12-05-2015, 09:34 PM   #15
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 8,937
QuoteOriginally posted by BrianR Quote
To be fair, I spend most of my time at f/11 to f/16 where diffraction will take that edge off with a little bit of blur in there
Sure, diffraction blur at f/16, for instance, means you could capture the detail of the image projected by the lens with a ~5MP sensor, i.e., a lot of detail in the scene is lost and the need for a Bayer AA-filter is heavily reduced (and becomes completely unecessary for a ~25MP camera).

QuoteOriginally posted by Nicolas06 Quote
As measured by tools and reviewers using standard procedures, a filterless design is sharper than a design with ah low pass filter. ... The gain seems to be 10-20% resolution.
You are incorrectly relying on data that was produced without providing a level playing field.

The use of Bayer-AA filter mandates that the image is capture-sharpened. The intentional blur can be, and should be, removed in post-processing.

A filterless design does not require this capture-sharpening (and would respond to it by exaggerating the artefacts that are already present).

If you measure resolution without using extra capture-sharpening for a camera like the K-5 II compared to a K-5 IIs then you get incorrect results, suggesting a resolution advantage that isn't real.

Have a look at this comparison of filtered, filterless, and filtered-sharpened versions of the same scene and you'll see that there is not more detail in the filterless version.

QuoteOriginally posted by Nicolas06 Quote
Now on some occasion you can have moire. On 24MP this is not that often.
How often it occurs, depends on shooting conditions. In less than ideal conditions there are a number of sources for image blur that avoid major artefacts.

I'm not saying that a filterless design will give you a ton of unusable shots. I'm just saying if you are interested in getting the most resolution possible and invest into creating the right shooting conditions to get it then a filterless design is the opposite of what you want.

Filterless designs are typically promoted as resolution enhancers, whereas they only really work when shooting conditions are adverse to high resolution captures.

QuoteOriginally posted by Nicolas06 Quote
You need a high quality lens with a subject with the proper pattern.
True, a bad lens will step in for a missing Bayer-AA filter, but why use a filterless design when you are not going to use high quality lenses? Obviating the need for capture-sharpening is a good reason, but getting more detail is not.

Regarding the "proper pattern" that is only needed to excite moiré. As long as there is small enough detail in the scence, it will cause damage to an image captured without a Bayer-AA filter in the form of false detail and false colour. These effects will admittedly only be visible at high magnification, but again, this is contrary to the claim that a filterless design increases resolution.

QuoteOriginally posted by Nicolas06 Quote
Moire is mostly visbible at 100% viewing size and a low pass filter can be applied in post processing.
Applying the low pass filter in post-processing is too late. The damage has been done and you can only hide it. Check out this comparison here, for instance. You will never be able to reconstruct the details of the scene as you can when using a Bayer-AA filter with optimal capture sharpening. Part of the reason is that a Bayer-AA filter does not "blur", it "splits".

QuoteOriginally posted by Nicolas06 Quote
To me that's the difference between theory and practice. In practice it yield better results.
The theory predicts that most shooters will not get problems from using a filterless camera in most situations because they are not taking images under conditions that allow critical sharpness to develop.

In other words, most of the time, the majority of shooters don't need an extremely resolving sensor, so they shouldn't get a filterless camera for reasons of capturing high levels of detail. On the contrary, when shooting conditions are such that high levels of detail can be captured, the Achilles heel of the filterless design kicks in.

FWIW, I found an earlier post, I've written, about why omitting a Bayer-AA filter is not a good idea that goes into more detail, for instance, the fact that most demosaicing algorithms rely on the "splitting" (not "blurring") effect of the Bayer-AA filter.

Last edited by Class A; 12-05-2015 at 09:41 PM.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
camera, dslr, photography, sensor
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Is ther a K-x or other AA Pentax for sale on the Forum kamicota Pentax DSLR Discussion 21 08-17-2015 07:33 AM
Is a K-3 with a full frame sensor enough for you? jake14mw Pentax Full Frame 58 03-28-2015 11:39 AM
Pentax staying small. An advantage? stillshot2 Photographic Industry and Professionals 27 12-22-2013 02:34 AM
Bayer filtered camera sensor brian101 Troubleshooting and Beginner Help 4 05-08-2013 01:10 PM
What is the advantage of 21/35/40 prime vs a f2.8 zoom lens raider Troubleshooting and Beginner Help 23 06-21-2010 10:57 AM



All times are GMT -7. The time now is 03:23 AM. | See also: NikonForums.com, part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top