Moiré disaster on the E-5

Started Jul 10, 2013 | Discussions thread
John Sheehy Forum Pro • Posts: 20,094
Re: Not nearly the same ...

rovingtim wrote:

Airmel wrote:

The technology simply doesn't exist to build a cheap optical low pass filter which has anywhere near the performance of its electronic signal processing counterpart. For this reason you might rightly refer to all optical anti-alias filters as "weak".

Perhaps optical AA's aren't as efficient as their electronic counterpart, but I think the key is more likely to be "The technology simply doesn't exist to build a cheap optical low pass filter ..."

I suspect this has more to do with why the industry has campaigned to wean photographers off AA's. Additionally, I think it could be argued that the 'carrot' they used is the appearance of higher resolution that results from the artifacts.

That's the big problem.  The mob, unfortunately, with better eyes than the parts of the brain that process the information provided by the eyes, conflate the visual acuity of having a small fraction of neighboring pixels having high contrast between them with "detail".  Yes it is detail, but not detail of just the subject, but also the detail of the pixel structure and a distortion of the subject detail, which ideally, one would not want to see at all, and have a virtually analog image.

Short of cameras that have hundreds of MPs, which may not happen for a long, long, time, a godsend would be a cheap method of variable AA filtering, based on user preferences, possibly including lens data (no one needs AA filtering when shooting with a moderately-sharp telephoto and a 2x TC stopped down to its sharpest at f/16, or doing 1:1 macro at f/16, where the airy disk is much larger than it is at f/16 and infinity).

I have a bad feeling that the false perception of the masses is going to hold back the progress of pixel density, which is the real solution to so many imaging problems.  A future camera that shot 200MP+ under the hood, and used user-selected RAW resampling methods and parameters for those who don't want to be bothered with all the data, would serve everyone.  Those who like the false detail of point-sampling could have that, even at fairly high resolutions.  Those who like properly-filtered images could use downsampling methods that included pixels outside the area covered by the output pixels.  The camera could do geometric distortion and CA corrections with a very minimum of image damage.  Doing CA correction with low pixel densities is a disaster, as any pixel contrast in edges in the capture will be gone if the scaling required by the correction for the red or blue channel might be a half pixel, in which case the transient goes from one pixel border to two pixels inclusive, while in another part of the image, the correction might be a whole pixel, losing no contrast - apparent sharpness occurring in a modulating pattern across the image.  When there is a lot of white or green, this is not a big issue, because green dominates the sense of micro-contrast, but for blues and reds against black, this can be significant.

Post (hide subjects) Posted by
(unknown member)
(unknown member)
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow