acl <achilleaslazarides RemoveThis @yahoo.co.uk> wrote:
> On Feb 23, 6:42 am, Wolfgang Weisselberg <ozcvgt... RemoveThis @sneakemail.com>
>> acl <achilleaslazari... RemoveThis @yahoo.co.uk> wrote:
>> I am saying that unless you exactly specify what you mean by
>> "be made to produce aliasing", that term is completely
>> worthless.
> Well, I wrote aliasing and you interpreted it as stairstepping in
> lines
If your dictionary calls that not "aliasing", you need to
replace it.
> (as opposed to aliasing of frequencies above nyquist onto lower,
> causing moire).
Unspoken assumption: only "aliasing of frequencies above
nyquist onto lower" causes more. That assumption is wrong,
as I show below. And why didn't you say so first?
> So now it's my fault that you misunderstood. Typical.
If I'd drool "uga uga uga" and you misunderstood, yes, then it
would be *my* fault. Why should I cut yourself slack I don't
cut myself?
>> > For the last time: What I mean is that I can see that there is detail
>> > above the sampling frequency of the D200 sensor with all my lenses.
>> With which camera with a higher sampling frequency than your D200
>> do you see that?[1]
> No, I am claiming that you simply misunderstood what I was saying: I
> was saying that I see aliasing (ie artifacts caused by undersampling),
You see something you _think_ is "artifacts caused by
undersampling". That does not make it so. It'd be far more
convincing if you *showed* the raw data, how it was produced
and why you interpret what you interpret.
BTW: Stair stepping also is aliasing, caused by undersampling:
oversampling solves it. Try *correctly* scaling down and see
it disappear. QED.
> therefore conclude that there is detail above the sampling frequency.
> Being you, you'll now keep arguing rather than say "aha I see, I
> thought you meant something else".
Aha, I see, you _meant_ something else.
And yet! But Yet!
http://weissel.smugmug.com/photos/258125275_XhywQ-O.jpg
=> Lots of moiré --- the patterns that you see are simply not
there. But does it indicate that the lens resolution is better
than the camera?
http://weissel.smugmug.com/photos/258125277_8n7yy-O.jpg
=> Quite a bit moiré in th center ... does this lens outresolve
the camera?
http://weissel.smugmug.com/photos/258125270_5wbn6-O.jpg
=> I think everybody will agree that this is the sharpest, most
resolved one of the three. It's also what happens if you
focus the lens, instead of unfocussing it!!
(Oh, the vertical stripes? That's the monitor ... and if they
don't show up in the more blurred variants, that just might
tell you something.)[1]
Second --- how's your chroma subsampling these days?[2]
For bonus points:
http://en.wikipedia.org/wiki/Image:Moiré_fringes_IMG_3712.jpg
beautiful moiré, completely independent of sampling frequency.
>> But I still don't know your methology. I am very interested
>> at exactly which level of contrast you count your lp/mm or
>> lp/ph.
> No I've had quite enough. I just hate pompous idiots who never admit
> to being wrong.
It's not my job to show you without any data from your side how
you are misinterpreting things and show you some cheap parlor
tricks which produce the results you so eagerly interpret.
Yet I go that extra mile, league and marathon.
Then I get called names when I ask for clarification!
Well!
I must say, my dear chap, you are a bit over the line.
-Wolfgang
[1] How to reproduce: Take a Siemens Star. Shoot it (I did it
of a monitor). Center crop (these are 100% crops). For the
ones above, simply defocus your lens. I used a 20D + 50mm
f/1.4 at 1/125s, f/2.5.
[2] If you do not understand, don't worry, it's just another
factor to overlook ...
>> Stay informed about: Please name the sharpest Nikon lenses