AA-filters

[iconoclastica]

I don't need a new camera and neither do I have the budget to buy another one just for the fun of upgading. So I have been neglicent about the regular news flashes on new developments in sensorland. In spite of this, last week my attention was drawn to a second hand Eos 5ds-R. For the reasons stated above, I had missed that several cameras have come to market without the previously obligate filter alternatively known as anti-alias, low-pass, or moire-filter.
Googling for comparisons, I found a few rather badly executed tests showing a little improvement in fine detail. However, it looked like the kind of detail a deconvolution filter could do business with.

Do any of you have experience with this under the often diffraction limited conditions we are working mostly?

[ray_parkhurst]

Do any of you have experience with this under the often diffraction limited conditions we are working mostly?

I did a direct comparison recently between my long-modified HRT2i and an unmodified one, plus SR technique and 3XTC, see it here:

http://www.photomacrography.net/forum/v ... 2i#p257171

My conclusion is that there is significant increased detail visible with AA filter removed.

Note that I purchased (twice!) the 5DSR and was disappointed with it each time. I saw no improvement in IQ vs my HRT2i, and perhaps it was not as good. The 5DSR "fine detail" (or whatever they call it) mode is just an improved sharpening algo. Applying that algo to the HRT2i output yielded similar improvements in sharpness, but I am happier with RawTherapee's tools than the simple and relatively inflexible ones provided by Canon.

Further note that the 5DSR actually still has the AA filter stack, but Canon added a second stack to "cancel" the effect. IMO they were not completely successful.

[iconoclastica]

Thanks for that link, Ray. For some reason it hadn't shown up in my search, but that is what I wanted to read.

--Wim