You still do not seem to get my point. I did not express myself clearly enough, sorry. "Automobile edges" - show me one automobile with really sharp edges. "For measurement we create lines" - these are not infinitely thin but have edges and the edges are not absolutely sharp. Then you say that MTF "is just a single point solution to the modulation transfer function where contrast = 0". Indeed! That was my original point: you cannot separate resolution and contrast. And in a real world you cannot separate your edges and lines with a contrast much higher than 0, the article you mention gives the limit at 20% - like in a photographic negative, print or slide. And the higher the contrast the more lines you can discern - visually or with any kind of equipment. And if we talk about "some silly notion about fuzzy molecules on the surface of the object" that�s what your lines - e.g. on a test target - are if we are talking about 200 lpm. All the best! Raimo Personal photography homepage at http://personal.inet.fi/private/raimo.korhonen
-----Alkuper�inen viesti----- L�hett�j�: Bob Blakely <[EMAIL PROTECTED]> Vastaanottaja: [EMAIL PROTECTED] <[EMAIL PROTECTED]> P�iv�: 17. tammikuuta 2002 20:58 Aihe: Re: And did I mention wrong >Raimo sez: "it (lpm) is effectively MTF" > >Bob sez: "No, it is not effectively MTF." > >Raimo insists: "Yes, it is. > >And Raimo continues: "Where do you find your points and lines made of points to >measure >your lpm?" > >Bob's Answer: For measurement, we create the lines. > >And Raimo continues: "In real world the edges of lines are not sharply defined if you >look >close enough." > >Bob's Answer: "Bullpucky! Consider a photo of an automobile (or an aerial photo of a >tank). The outline of the automobile (or an aerial photo of a tank) against the >background >is a sharp edge and its reality is sharply defined, unless you want to get into some >silly >notion about fuzzy molecules on the surface of the object. The world is full of edges. >Edges are how we recognize objects. In fact, the first layer of human visual >processing is >edge detection performed by neurons in the retina just behind the photo sensitive >cells. > >Go to http://www.photodo.com/art/articles.shtml# > >The graph shows contrast vs. spatial frequency. This is a graph of the modulation >transfer >function (MTF). Lines per millimeter is a single point on this graph. It is the point >where the MTF crosses 0 contrast. Anywhere just to the left of this crossing, items >can be >distinguished from one another. It may not look natural due to severe lack of >contrast, >but they can be distinguished. In some forms of photography, it is this ability to >distinguish separation that is all important. Note that the example, while sufficient >for >photographers such as yourself and most on this list for evaluation of lenses is >insufficient for others because it does not show where contrast becomes 0. To them, >this >is the all important single point on the graph. If they can get better contrast at the >lower spatial frequencies, that's icing on the cake. > >In other words, lpm is not "effectively the MTF". It is just a single point solution >to >the modulation transfer function where contrast = 0. > >Regards, >Bob... > > >----- Original Message ----- >From: "Raimo Korhonen" <[EMAIL PROTECTED]> >To: <[EMAIL PROTECTED]> >Sent: Thursday, January 17, 2002 8:42 AM >Subject: Vs: And did I mention wrong > > >> Yes, it is. Where do you find your points and lines made of points to measure your >lpm? >In real world the edges of lines are not sharply defined if you look close enough - >but I >see you point: if you set limits you can measure lpm - but if you set a different >limit, >the result will be different. >> All the best! >> Raimo >> Personal photography homepage at http://personal.inet.fi/private/raimo.korhonen >> >> -----Alkuper�inen viesti----- >> L�hett�j�: Bob Blakely <[EMAIL PROTECTED]> >> Vastaanottaja: [EMAIL PROTECTED] <[EMAIL PROTECTED]> >> P�iv�: 16. tammikuuta 2002 22:51 >> Aihe: Re: And did I mention wrong >> >> >> >No, it is not effectively MTF. In fact, lpm is not a transfer function. In >> >fact, being a single point, it does not meet the mathematical definition of >> >a function! >> > >> >Further, absolute black is available in nature. I did *NOT* claim that >> >absolute black exists or is necessary, but it exists. Stop into an Afghan >> >terrorist cave, have an F-14 seal it up with a missile, and you will >> >"discover" absolute black in the visual spectrum. >> > >> >Further, no mythical absolute white is necessary to measure lpm *NOR* did I >> >claim that it exists. I have no idea where you came up with this nonsense. >> > >> >All that is necessary for the measurement is that the power in the white be >> >significantly higher that the power in the "black". This is why there is a >> >reference to 3 dB or half power in what I said. This is a _relative_ >> >_power_, relative to the peak power in the white. How high is significantly >> >higher? well, that depends on how accurate you want the measurement. Just 15 >> >dB or 5 stops is certainly enough. >> > >> >Measurement of MTF is subject to analogous constraints. >> > >> >Regards, >> >Bob... >> > >> >From: "Raimo Korhonen" <[EMAIL PROTECTED]> >> > >> > >> >> No - I am not wrong - when it is done the way you say, it is effectively >> >MTF. There is no absolute black or white in the nature, or in the world if >> >you will. >> > >> >> L�hett�j�: Bob Blakely <[EMAIL PROTECTED]> >> >> >> >> >No. You are wrong. Lines Per Millimeter is objective. It's often read >> >with >> >> >the human eye, but can be read by instruments. Lines Per Millimeter can >> >also >> >> >be calculated mathematically from the rise from black to white of a >> >single >> >> >edge (or fall from white to black). It is identical to the concept of >> >> >measuring the resolution of a radar (in this case the measure is >> >degrees). >> >> >Generally, the figure is not measured but calculated from the parameters >> >of >> >> >the antenna of which gain is the greatest determining factor. The edge of >> >> >resolution is where (for example) the light (from white to black) falls >> >off >> >> >by 3 dB or half power. It's at this point that two separate objects >> >cannot >> >> >be separated mathematically, let alone visually. For photo surveyors and >> >> >aircraft/spacecraft spies, this is the ONLY measure that counts. FYI, lpm >> >> >for these lenses are not measured by eye but by instruments. Stay tuned >> >for >> >> >an installment on how contrast affects this number. >> >> > >> >> >I do err, but not often and not in areas where I have worked. >> >> > >> >> >Regards, >> >> >Bob... >> >> > >> >> >From: "Raimo Korhonen" <[EMAIL PROTECTED]> >> >> > >> >> > >> >> >> Nope - lpm is measured by human eye and it is quite subjective. I think >> >> >that MTF is the most objective method, it takes into account both >> >contrast >> >> >and sharpness objectively. >> >> >> All the best! >> >> >> Raimo >> >> >> Personal photography homepage at >> >> >http://personal.inet.fi/private/raimo.korhonen >> >> >> >> >> >> L�hett�j�: Bob Blakely <[EMAIL PROTECTED]> >> >> >> >> >> >> >"Sharpness" and resolution are closely tied together, so I will >> >discuss >> >> >them both. >> >> >> > >> >> >> >Resolution is objective and quantifiable. It is a measure of the >> >ability >> >> >to separate two >> >> >> >identical objects as they are brought close together. In photography, >> >> >this is measured in >> >> >> >lines per millimeter where the lines are black the spaces between the >> >> >lines are white. >> >> >> <snip> >> >> >> >Regards, >> >> >> >Bob... >> >> - >> >> This message is from the Pentax-Discuss Mail List. To unsubscribe, >> >> go to http://www.pdml.net and follow the directions. Don't forget to >> >> visit the Pentax Users' Gallery at http://pug.komkon.org . >> >- >> >This message is from the Pentax-Discuss Mail List. To unsubscribe, >> >go to http://www.pdml.net and follow the directions. Don't forget to >> >visit the Pentax Users' Gallery at http://pug.komkon.org . >> - >> This message is from the Pentax-Discuss Mail List. To unsubscribe, >> go to http://www.pdml.net and follow the directions. Don't forget to >> visit the Pentax Users' Gallery at http://pug.komkon.org . >- >This message is from the Pentax-Discuss Mail List. To unsubscribe, >go to http://www.pdml.net and follow the directions. Don't forget to >visit the Pentax Users' Gallery at http://pug.komkon.org . - This message is from the Pentax-Discuss Mail List. To unsubscribe, go to http://www.pdml.net and follow the directions. Don't forget to visit the Pentax Users' Gallery at http://pug.komkon.org .
