On 18/03/2013 14:15, Doug McDonald wrote:
[]
OK, I've posted the seriously moired pictures from a Canon 7D photographing
"Cupcake Wars" on FoodTV network on an LED TV. This used a Canon 50 mm
f/1.7 lens at
either f/4.5 or f/5.6, if you care, look at the EXIF. There are two scenes,
the version beginning in an underscore was "developed" from raw with
Digital Photo Professional
and is full frame. The ones beginning with A are developed with
Photoshop and are
cropped quite a bit. Be sure to view them blown up to at least 100% or
200%. The ones
done in Photoshop have very low sharpening done to them.

https://dl.dropbox.com/u/5310360/_MG_3388.JPG
https://dl.dropbox.com/u/5310360/_MG_3391.JPG
https://dl.dropbox.com/u/5310360/AMG_3388.jpg
https://dl.dropbox.com/u/5310360/AMG_3390.jpg

Note that the distance to the screen is very critical to get the full
disastrous moire effect, as is focus. I focused using the Live View.
The camera was on a tripod.

Amazing, isn't it! These pictures are truly unusable.

YET ... I've never photographed a real scene, even with that or my even
better
100mm f/2.8 Macro lens where I saw any problem at all.

People here have asked for moired pictures ... here they are!

Doug McDonald

Thanks for posting those, Doug. What immediately strikes me is how much
worse the "A" images are, converting a relatively gentle patterning on
the horizontal part of the arm into a series of nasty-looking pink
splodges! Around X:661 Y:1718 in AMG_3388. I therefore discount these
photoshopped images.

Considering the _MG_3388 image, yes, there is moiré visible, but you
would expect some moiré as the simple AA filters used in cameras are not
perfect, and are not the "brick-wall" filters used in audio.

Of course, the TV image should have an optical low-pass filter in front
to the display in any case, to ensure that only optical frequencies up
to those contained in the source signal are displayed. This is normally
achieved by viewing the image at a distance where you cannot see the
dots - the MTF of your eye and brain filter out the dots and lines.

Your later comment about the quality of the TV image itself - the
artefacts caused by compression for transmission - are quite valid. We
do not notice then partially because the image is moving and we don't
have time to concentrate of the errors, perhaps sometimes because we
expect a particular channel to have a low-bandwidth low-quality signal,
but perhaps mainly because we are interested in the content of the image
(and the accompanying sound) rather than in its technical quality.
Those who view testcards all day may have a different opinion!

Will there still be testcards anywhere on the UK broadcast network after
next Tuesday?
--
Cheers,
David
Web: http://www.satsignal.eu