Archived posting to the Leica Users Group, 2009/11/20

[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]

Subject: [Leica] Thought question re antialiasing
From: dstella1 at ameritech.net (Dante Stella)
Date: Fri, 20 Nov 2009 08:36:22 -0500
References: <494971.52240.qm@web83908.mail.sp1.yahoo.com> <20091120032721.GA2535@d242.scdc1.swarthmore.edu> <4796FFE1-49FB-45D9-8E75-A381DCB355E3@ameritech.net> <20091120035749.GB2535@d242.scdc1.swarthmore.edu> <36172e5a0911192044s197d811ao879d833085c1a661@mail.gmail.com> <E8CE85D5-0958-4B15-B96E-11BDEABCCA9F@ameritech.net> <36172e5a0911192245u1298bc7et994c127d88e6b171@mail.gmail.com>

"Some" processing being the operative word...  I will probably put the 
question on microlenses directly to Kodak - which is usually a lot more 
forthcoming with technical information than Leica typically is.

Dante

On Nov 20, 2009, at 1:45 AM, Geoff Hopkinson wrote:

> hmm, well there is some correction going on in firmware even for DNG of
> course; the vignetting correction for example. I had thought Leica had
> already stated that they are dealing with moire in firmware.
> 
> 2009/11/20 Dante Stella <dstella1 at ameritech.net>
> 
>> Kodak is a little bit cagey with the info on the offset lenses on the
>> KAF-10500 and its bigger brother.  I think the microlenses are tailored to
>> the application.
>> 
>> What got me thinking about this is that I have an M8 and a Kodak 14n.
>> Neither has an AA filter, but the 14n also lacks microlenses.  If you
>> process RAW files for both through Lightroom (so you don't get 
>> contamination
>> from any in-camera correction), the 14n will produce insoluble moire much
>> more of the time - sometimes to the point where inducing diffraction by
>> stopping down the lens to f/11 is the solution.  On the other hand, the
>> Leica virtually never hits this point.  But given the tighter pixel pitch
>> and even higher-resolution lenses, one would expect it to.
>> 
>> My hypothesis (which might be wrong) is that the little lenses (completely
>> offset or not) are acting as something of an AA filter - and what we all
>> take to be sharper pictures from no AA filter is really higher-contrast
>> pictures from higher-contrast lenses.  Both the M8's and M9's resolutions
>> are 73lp/mm (at least according to Reichmann), and it should be routine to
>> cross that at wide apertures with M lenses.  But the incidence of moire at
>> any aperture is so low as to suggest that the barrier rarely gets crossed.
>> That suggests something acting as a limiting factor on resolution.
>> 
>> Something interesting, BTW, is the difference with and without AA filters.
>> If it is a relatively weak filter, as on the Nikon D3, removing the AA
>> filter actually does very little.  Go here:
>> 
>> http://bythom.com/nikond3xreview.htm
>> 
>> And take a look at the with/without pairs for the D3 and D3x.  Very
>> interesting.
>> 
>> Dante
>> 
>> 
>> 
>> On Nov 19, 2009, at 11:44 PM, Geoff Hopkinson wrote:
>> 
>>> My assumption also. Perhaps there is info on this available from Kodak,
>> or
>>> Dante has it already? Interesting technically anyway.
>>> 
>>> 2009/11/20 Tim Gray <tgray at 125px.com>
>>> 
>>>> On Thu 19, Nov'09 at 10:37 PM -0500, Dante Stella wrote:
>>>> 
>>>>> They do span more than one pixel - that's why they're called offset.
>> And
>>>>> that is why I asked.
>>>>> 
>>>> 
>>>> Ahh, I wasn't aware that they did that.  I interpreted the offset
>>>> microlenses as not being centered directly over a pixel, but still
>> funneling
>>>> in light into just one pixel.  Thus, enabling light to come in at more
>>>> oblique angles and still be captured.  I drew an ascii diagram, but I'm
>> sure
>>>> most everyone's mail client will mangle it, unless you have monospaced
>>>> fonts.
>>>> 
>>>> -------
>>>> \     |
>>>> \    | <- offset lens
>>>> \___|
>>>> |  | <- pixel
>>>> ----
>>>> As opposed to
>>>> 
>>>> -------
>>>> \     /  <- non-offset lens
>>>> \___/
>>>> |  |  <- pixel
>>>> ----
>>>> 
>>>> 
>>>> _______________________________________________
>>>> Leica Users Group.
>>>> See http://leica-users.org/mailman/listinfo/lug for more information
>>>> 
>>> 
>>> 
>>> 
>>> --
>>> Cheers
>>> Geoff
>>> http://www.pbase.com/hoppyman
>>> 
>>> _______________________________________________
>>> Leica Users Group.
>>> See http://leica-users.org/mailman/listinfo/lug for more information
>> 
>> 
>> _______________________________________________
>> Leica Users Group.
>> See http://leica-users.org/mailman/listinfo/lug for more information
>> 
> 
> 
> 
> -- 
> Cheers
> Geoff
> http://www.pbase.com/hoppyman
> 
> _______________________________________________
> Leica Users Group.
> See http://leica-users.org/mailman/listinfo/lug for more information



In reply to: Message from dstella1 at ameritech.net (Dante Stella) ([Leica] Thought question re antialiasing)
Message from tgray at 125px.com (Tim Gray) ([Leica] Thought question re antialiasing)
Message from dstella1 at ameritech.net (Dante Stella) ([Leica] Thought question re antialiasing)
Message from tgray at 125px.com (Tim Gray) ([Leica] Thought question re antialiasing)
Message from hopsternew at gmail.com (Geoff Hopkinson) ([Leica] Thought question re antialiasing)
Message from dstella1 at ameritech.net (Dante Stella) ([Leica] Thought question re antialiasing)
Message from hopsternew at gmail.com (Geoff Hopkinson) ([Leica] Thought question re antialiasing)