[froeschle]

These



are crops (AA-Filter-Simulator: OFF, TYPE 1, TYPE 2) from this scene



K-3 + FA 20-35 (ISO 100, 20mm, f7.1, distance about 1.2m)

[Adam]

That's one evil setup you've got there! Very nice

[TER-OR]

Purpose built for Moire patterning!

It really shows the effectiveness, too. The sharpness drop doesn't seem THAT severe, thanks for sharing this.

[akanarya]

great share, thanks
please link full image so that we can look in detail

[JimmyDranox]

Wow. Very good setup. I think it's very good for testing lens too. I want some prints like these. How do I get some?

[DeadJohn]

Purpose built for Moire patterning! It really shows the effectiveness, too. The sharpness drop doesn't seem THAT severe, thanks for sharing this.

Maybe it's my monitor, but I can't spot any sharpness drop. The cropped text looks equally readable regardless of the setting. I wonder if a different lens, perhaps an ultrasharp macro, would show a sharpness difference between AA settings.

@froeschle, thanks for posting this.

[froeschle]

Thanks, finally I am a "Pentaxian" (Veteran Member)!

please link full image so that we can look in detail

The three JPGs are about 17MB each.
Is it possible to upload such images here?
Nevertheless, the three depicted areas are representative.
As the overview shows, the false colors really can ruin a complete picture - or generate modern art.

I think it's very good for testing lens too. I want some prints like these. How do I get some?

There are probably better ways to test a lens.
However, one can use lens test charts from the internet to determine critical lp/mm parameters, where problems should occur.
After such a procedure I used suitable patterns as a stamp.
Note that useful prints for this "Color Moiré Art" depend on your setup.
I had something like this from my Ricoh GR test.
Thus, I used a similar focal length here.

but I can't spot any sharpness drop. The cropped text looks equally readable regardless of the setting. I wonder if a different lens, perhaps an ultrasharp macro, would show a sharpness difference between AA settings.

I am not really keen on trying that out.
The FA 20-35 @ 20mm & f7.1 should also be a quite solid performer in the image centre:
Photozone: Lab Test, MTF
In fact, the occurence of these artefacts shows that the lens is decently sharp.
A softer lens would act as an additional "AA" filter.
My point here was that it is quite easy to provoke moiré.
Often, it is stated that it can easily be removed.
Anyone is invited to try this out with the pictures above.

[TER-OR]

Maybe it's my monitor, but I can't spot any sharpness drop. The cropped text looks equally readable regardless of the setting. I wonder if a different lens, perhaps an ultrasharp macro, would show a sharpness difference between AA settings.

@froeschle, thanks for posting this.

Maybe I'm trying to find faults - I don't spot any real differences either upon closer examination.

[Adam]

Often, it is stated that it can easily be removed.

Anyone is invited to try this out with the pictures above.

Here's what software (ACR) can do. Fixing up the colors and eliminating the repeating patterns is tough, but it seems to do a good job of getting rid of the false color.

[froeschle]

Maybe the JPG settings of the comparative picture also mask the difference.
I have exceeded the maximum number of allowed file uploads.
Thus, the attachment now contains part of the comparison with quality setting at 100%.
There is a slight "sharpness" advantage without the simulated filter (with additional false colors, however).

but it seems to do a good job of getting rid of the false color

But luminance moiré is still there...

[danny09]

Maybe the JPG settings of the comparative picture also mask the difference.
I have exceeded the maximum number of allowed file uploads.
Thus, the attachment now contains part of the comparison with quality setting at 100%.
There is a slight "sharpness" advantage without the simulated filter (with additional false colors, however).

But luminance moiré is still there...

Thanks a lot for this useful information. I was eagerly looking forward to a 'real world' challenging test of AA filter simulator settings. Looks like AA filter OFF indeed increases the sharpness considerably I wonder how AA Filter OFF impacts low shutter speed handheld shots? Say 1/10th or say 1/6th of a second at 50 mm? If the subject is stationary I get sharp to reasonably sharp shots with SR turned ON.

[crussellsprout]

Nice samples! Can you also try out the in-camera post-processing moiré removal tool to see how that compares?

[froeschle]

Can you also try out the in-camera post-processing moiré removal tool

I have not tested that yet.
But it probably will not be better then what Adam posted.
I tried it with the Ricoh GR though (follow the red arrow): Ricoh GR moiré removal

Btw, here are some pictures of my current K-3 review (in German, work in progress).

I wonder if a different lens, perhaps an ultrasharp macro, would show a sharpness difference between AA settings.

Just to please you, I did some quick & dirty unscientific testing .
I printed out (DIN A4) these charts:
http://www.normankoren.com/Tutorials/Lenstarg_50_5906p_15g_25is.png
http://www.falklumo.com/lumolabs/articles/sharpness/images/siemensstern_pentax.jpg
Note that my printer was not able to fully resolve the originals.
The setup was also fast and simple - just to illustrate the magnitude of the difference.
I used the FA 50/2.8 Macro - one of the sharpest lenses available for Pentax.
AFAIK such a lens was used by Pentax Germany to calibrate DSLR bodies.
JPGs OOC (f5.6, ISO 100), crops (above) are from the full scene (below):

[DeadJohn]

The sharpness difference directly above this is more apparent yet still barely noticeable. That's good. The K-3 version of AA is strong enough to suppress moire, while weak enough to give sharp images. Even if you leave the AA-via-SR on all the time you can get very sharp images with the K-3.