Olympus-OM
[Top] [All Lists]

Re: [OM] Blur from camera shake

Subject: Re: [OM] Blur from camera shake
From: whunter <whunterjr@xxxxxxxx>
Date: Sat, 9 Aug 2003 16:51:35 -0400
Fine, that would be my choice. PhotoShop is too expensive and too stressful unless doing color separation work. To avoid endless talk about preferences, I cited PhotoShop which constitutes a paradigm for desktop computing. I also said "or equivalent" which includes many other excellent software applications. While I have reviewed the FFT equations for research applications - - and they are IMPRESSIVE - - I am not a mathematician. I offer two simple concepts to understand them in totality without mathematics: - FFT, PhotoShop, etc, are mathematical 'Filters' analogous to the use of optical filters on a lens which impacts what you see. - UNLIKE optical filters, etc., which can alter and directly determine the S/N of data collected, the mathematical filters can only filter this fixed set of data, not alter it. (I should add that the REAL value of FFT is in the design of optimized imaging systems including lenses to maximize S/N of collected data and in some complex systems as an online monitor of data being collected. Use of FFT to make 'purty' pictures is but an afterthought. GI/GO is the Law of God!!!)

'Ideal' data (picture, etc.) can be enhanced with mathematical filters to improve our perception. Likewise, data with low S/N (shake, poor focus, etc.) can be filtered to improve our perception from this data. Where reality breaks down, is the 'public perception' that garbage in can mysteriously be converted to pearls out.

Finally, to accent the general principles above: Scattered photons of light can seriously degrade the S/N, i.e. 'sharpness', of a photo. You are experienced in the use of polarizing filters to markedly enhance the S/N at the time of data acquisition when a scattered photon (N) can be separated by identifiable physical characteristics. However, once all of the photons (data bits) are collected in one 'bucket', either film or digital, there is no identifiable characteristic of a scattered photon. Placing a 'polarizing' mathematical filter on the 'bucket' of collected data will not alter the S/N. 'Sharpening' an image with mathematical filtering of low frequency data blurring edges is not even in the same vernacular as 'sharpening' an image by exclusion of scattered light at the time of acquisition.
Hope that helps.
Bill Hunter

On Saturday, August 9, 2003, at 03:25  PM, Danrich wrote:

What about PhotoShop Element 2.0 made for photo work specifically?
I guess sharpen is not sufficient to overly blurred picture?
Daniel



< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >


<Prev in Thread] Current Thread [Next in Thread>
Sponsored by Tako
Impressum | Datenschutz