Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
10-07-2013, 02:32 PM   #31
Pentaxian
RobA_Oz's Avatar

Join Date: Nov 2008
Location: Tasmania, Australia
Photos: Gallery
Posts: 4,120
QuoteOriginally posted by CRPhoto Quote
I've never had a problem with moire with my K-5 IIs, if I ever buy the K-3, this featured would be off all the time! I love the sharp detailed pictures my IIs gives me!
I suspect it's aimed at those who want to avoid Moiré fringing in large-scale printing. The higher resolution means finer detail at high magnification, which is paradoxical, as the use of sensor shake will effectively reduce the resolution. I can't imagine that some complex algorithm will restore it, but we'll have to see what the output's like in reality, before passing judgment.

10-07-2013, 03:38 PM   #32
Senior Member
joergens.mi's Avatar

Join Date: Jan 2012
Posts: 229
QuoteOriginally posted by RobA_Oz Quote
I suspect it's aimed at those who want to avoid Moiré fringing in large-scale printing. The higher resolution means finer detail at high magnification, which is paradoxical, as the use of sensor shake will effectively reduce the resolution. I can't imagine that some complex algorithm will restore it, but we'll have to see what the output's like in reality, before passing judgment.
But not more then a AA-Filter, therefore the 24 Mpx resolution will be preserved. And you have the chance to get in 3 different steps
10-08-2013, 12:13 AM   #33
Pentaxian
bossa's Avatar

Join Date: Jul 2011
Location: Adelaide
Posts: 4,538
QuoteOriginally posted by audiobomber Quote
More megapixels means less chance of moire, for those who were asking.
I get moire on my D800E so I guess a 24MP camera will show "moire" (sorry ) but I also saw the samples from another post on this forum which clearly showed this tech works.

Sample shots of AA fliter at work from another member here with a pre-release K-3


As Dean Martin said all those years ago "That's a moire"
10-08-2013, 12:45 AM   #34
Pentaxian
gazonk's Avatar

Join Date: Feb 2009
Location: Oslo area, Norway
Photos: Gallery | Albums
Posts: 3,509
QuoteOriginally posted by bossa Quote
Sample shots of AA fliter at work from another member here with a pre-release K-3
Interesting! But I have a load of shots from my k10D with worse moire than that.

10-08-2013, 01:05 AM   #35
Pentaxian
bossa's Avatar

Join Date: Jul 2011
Location: Adelaide
Posts: 4,538
Yeah, it's not bad but as I don't own a K-3 I can only repost what I find elsewhere. My D800E does worse as well.
10-08-2013, 01:13 AM   #36
Veteran Member
Caat's Avatar

Join Date: Feb 2009
Location: United Kingdom
Photos: Albums
Posts: 896
I don't mind moire in photos most of the time because I find that my eyes seem to suffer from it quite a lot anyway!!
10-08-2013, 01:51 AM   #37
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,862
For the record ;)



Just for the record ...

Over the years, I had proposed this exact same "anti-aliasing simulator" several times on mutliple occasions, here in this forum and elsewhere.

I called it Bayer-AA filter implemented by the SR motor (or so).

At the time, I actually did computations if the SR motor is fast / strong enough to support this operation. I argued that the principle holds strong but would need a v2 SR motor to support shutter speeds as fast as 1/8000s (maybe somebody can still find the posts, I can't ...).

As it turns out now, Pentax engineers had the same idea (which I consider to be a very very obvious one). Just, they dismissed the caveat and kept the v1 SR motor, limiting the effect to 1/1000s and slower. I strongly believe that the K-5IIs was the test balloon to test for acceptability of a missing Bayer-AA filter. It is now obvious in the industry that many accept the idea.

So, the K-3 has an adjustable Bayer-AA filter for speeds 1/1000s and slower, and effectively none for faster speeds. Which is good enough for me. Still, a stronger SR motor would have been a more ambitious solution (possibly allowing for better SR too). But ok, compromises are everywhere in todays industries ...

Bravo Pentax.

And it is a shame for Sony and Olympus they didn't implement the idea first.

So, Bravissimo Pentax.


P.S.
If anybody has problems to understand how this thing works, I am ready to answer precise questions about.
10-08-2013, 02:26 AM   #38
Pentaxian
Clavius's Avatar

Join Date: Feb 2011
Location: De Klundert
Photos: Gallery
Posts: 4,115
QuoteOriginally posted by gazonk Quote
Interesting! But I have a load of shots from my k10D with worse moire than that.
QuoteOriginally posted by falconeye Quote


Just for the record ...

Over the years, I had proposed this exact same "anti-aliasing simulator" several times on mutliple occasions, here in this forum and elsewhere.

I called it Bayer-AA filter implemented by the SR motor (or so).

At the time, I actually did computations if the SR motor is fast / strong enough to support this operation. I argued that the principle holds strong but would need a v2 SR motor to support shutter speeds as fast as 1/8000s (maybe somebody can still find the posts, I can't ...).

As it turns out now, Pentax engineers had the same idea (which I consider to be a very very obvious one). Just, they dismissed the caveat and kept the v1 SR motor, limiting the effect to 1/1000s and slower. I strongly believe that the K-5IIs was the test balloon to test for acceptability of a missing Bayer-AA filter. It is now obvious in the industry that many accept the idea.

So, the K-3 has an adjustable Bayer-AA filter for speeds 1/1000s and slower, and effectively none for faster speeds. Which is good enough for me. Still, a stronger SR motor would have been a more ambitious solution (possibly allowing for better SR too). But ok, compromises are everywhere in todays industries ...

Bravo Pentax.

And it is a shame for Sony and Olympus they didn't implement the idea first.

So, Bravissimo Pentax.


P.S.
If anybody has problems to understand how this thing works, I am ready to answer precise questions about.
We may even conclude that Pentax R&D reads PF forums. They could just as have picked up the idea from your posts.

10-08-2013, 02:30 AM   #39
Site Supporter
geomez's Avatar

Join Date: Mar 2013
Location: Preskit Arizona
Photos: Gallery
Posts: 1,557
QuoteOriginally posted by falconeye Quote
Just for the record ... Over the years, I had proposed this exact same "anti-aliasing simulator" several times on mutliple occasions, here in this forum and elsewhere. I called it Bayer-AA filter implemented by the SR motor (or so). At the time, I actually did computations if the SR motor is fast / strong enough to support this operation. I argued that the principle holds strong but would need a v2 SR motor to support shutter speeds as fast as 1/8000s (maybe somebody can still find the posts, I can't ...). As it turns out now, Pentax engineers had the same idea (which I consider to be a very very obvious one). Just, they dismissed the caveat and kept the v1 SR motor, limiting the effect to 1/1000s and slower. I strongly believe that the K-5IIs was the test balloon to test for acceptability of a missing Bayer-AA filter. It is now obvious in the industry that many accept the idea. So, the K-3 has an adjustable Bayer-AA filter for speeds 1/1000s and slower, and effectively none for faster speeds. Which is good enough for me. Still, a stronger SR motor would have been a more ambitious solution (possibly allowing for better SR too). But ok, compromises are everywhere in todays industries ... Bravo Pentax. And it is a shame for Sony and Olympus they didn't implement the idea first. So, Bravissimo Pentax. P.S. If anybody has problems to understand how this thing works, I am ready to answer precise questions about.
So that means Pentax could release a firmware update for my K-5 IIs and give me the functionality if an on-the-fly AA filter if they WANTED to. If so I can't really see why they wouldn't. I can't imagine this being the one feature motivating K-5 IIs owners to trade up to a K-3.
10-08-2013, 02:44 AM   #40
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,862
QuoteOriginally posted by Clavius Quote
We may even conclude that Pentax R&D reads PF forums. They could just as have picked up the idea from your posts.
Extremely unlikely.

QuoteOriginally posted by geomez Quote
So that means Pentax could release a firmware update for my K-5 IIs and give me the functionality
I don't think so. Besides strength of the SR motor, it requires the ability to send commands to / read positional data from the motor at a higher frequency ordinary SR operation would require. There are certainly subtle changes in the overall circuitry to support the new idea.
10-08-2013, 03:57 AM   #41
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 8,934
Imagine-Resource have published a nice description of the optional Bayer-AA-Filter approach of the K-3 along with an analysis of the working frequency of the system (500Hz).

They published a sound file which lets you hear the system at work.
10-08-2013, 08:34 AM   #42
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,862
QuoteOriginally posted by Class A Quote
Imagine-Resource have published a nice description of the optional Bayer-AA-Filter approach of the K-3 along with an analysis of the working frequency of the system (500Hz).

They published a sound file which lets you hear the system at work.
Very nice analysis indeed. Thanks for the link.

I guess the linear mode is actually useful, e.g., if you work with shutter speeds near 1/90s where shutter-induced blur alone may suffice for AA purposes in the vertical direction.

500Hz means that the oscillation makes a half wave within 1/1000s. This may or may not blur full amplitude, depending on the phase of the curtain slit travelling across the image. So, the AA effect will vary across the image at fast shutter speeds. An interesting playground for geeks

P.S.
The Bayer-AA is audible. I remember when I proposed my idea, I opted to go for an ultrasonic operating frequency to make it both inaudible and fast enough for 1/8000s and video. However, the required forces might have become too large since Pentax moves an entire daughterboard with glass. Going from 500Hz to e.g., 25,000Hz requires 2500x stronger forces. Which would be a problem if IR's claim of 50x stronger motors were true. However, a 500Hz wave of 2µm amplitude has an acceleration of 2.0g exactly (if I am not wrong). This is about the force of Pentax' v1 SR motor (as determined by us when we studied shutter-induced blur). No idea where IR got their idea from that Pentax needed a stronger motor for this. However, to go ultrasonic, they would have needed a new 5000g motor and a much more lightweight sensor assembly
BTW, the v2 SR motor still needs to be a bit stronger than v1 SR in order to fully support SR when vibrating...

Last edited by falconeye; 10-08-2013 at 08:56 AM.
10-08-2013, 09:51 AM   #43
Pentaxian




Join Date: Apr 2007
Location: Romania
Posts: 8,406
QuoteOriginally posted by falconeye Quote
So, the K-3 has an adjustable Bayer-AA filter for speeds 1/1000s and slower, and effectively none for faster speeds. Which is good enough for me.
According to Pentax, the system works for speeds faster than 1/1000s, albeit it's less effective. Do you think it won't work at all?
10-08-2013, 12:27 PM   #44
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,862
OMG! Did you guys read about the AA on the K-3?

QuoteOriginally posted by Kunzite Quote
According to Pentax, the system works for speeds faster than 1/1000s ...
You mean, according to some marketing guys ...

As we do know now, the sensor vibrates at 500 Hz.

This means, it is 100% effective at 1/500s, blurring say with 1 pixel width.

At 1/1000s, it will blur for half a phase only, i.e., parts of the image will be blurred with 1 pixel width while others will be blurred with 1/2 pixel width only, depending on the direction and location of the image. Taking the square norm, the average blur will be 0.8 pixels wide, call it 80% effective. A formally more complete treatment yields 1/SQRT(2) effective.

At anything faster, the blur will be proportionally less. This gives the following average AA blur effect strengths (varying across the image plane!):

1/500s and slower: 100%
1/1000s: 71%
1/2000s: 35%
1/4000s: 18%
1/8000s: 9%

The effect will be still visible at 1/1000s but moiré will occur (similiar to cameras with a "weak" AA filter). Faster and the AA effect will be unnoticeable, even in the lab (similiar to cameras w/o AA filter).

To say the system is working at speeds faster than 1/1000s may be legally ok, but is a blatant lie in any other respect.

Last edited by falconeye; 10-08-2013 at 12:33 PM.
10-08-2013, 12:40 PM   #45
Pentaxian




Join Date: Apr 2007
Location: Romania
Posts: 8,406
QuoteOriginally posted by falconeye Quote
You mean, according to some marketing guys ...
I don't necessarily trust them, I guess it depends what's the real source for that (if it's their R&D, it might actually be true ) And of course, I value your opinion.

Could it be, that the 500Hz frequency is not for a linear movement, but for a circular one? (it's claimed that the system can work this way) Would that change anything?
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!
anti-aliasing, image, level, pentax news, pentax rumors
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Does the K-r AA battery holder works on K30? simbon4o Pentax K-30 & K-50 21 10-23-2012 09:06 AM
New K-30, did you remember the $20 2-year extended warranty? Docrwm Pentax K-30 & K-50 79 08-05-2012 03:17 PM
Features you found neat about the K-30 0rangepeel Pentax K-30 & K-50 13 08-03-2012 06:43 AM
what do you guys think about shutterfly? vimeto General Talk 8 12-22-2011 08:10 PM



All times are GMT -7. The time now is 05:43 PM. | 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