Re: [pinhole-discussion] Reciprocity Problem
Hi Ballard, As John Farrias said, There is no curve to apply to reciprocity. I couldn't find one single curve that worked for all the times on the chart I have. I really have no idea what There is no curve to apply to reciprocity. It is a linear logarithmic function means. If it is a function, then it has to have a curve, so that sounds like it contradicts itself. I've never heard of linear logarithmic functions either, so my last sentence could easily be wrong. I played around for quite some time using adjusted times from an extensive reciprocity failure table for tmax 100, but couldn't find much. All the equations I found either matched short exposures or really long exposures. I can send the original chart I obtained these formulas from as soon as the author gives me permission. (i obtained it from a guy named Chris Cline through either rec.photo.darkroom or rec.photo.equipment.large-format) From 4 sec. to 4 min, the points are pretty close to the equation: y=(.00312782828831x^2) + 1.912257105111x -3.2132290636125 Where X is the metered time, and Y is the corrected time. From 4 minutes to one hour, the points are pretty close to: Y=((2.1721280231785 * 10 ^ -4) x^2) + 3.3843212260403x - 272.36563833739 Where X is the metered time, and Y is the corrected time. The leading coefficient is 2.172... times ten to the negative fourth power. it looks kinda confusing in the equation. Keep in mind this is for Tmax 100, not 400. The reciprocity failure for 400 is worse than 100. You really can't get much of an equation from three points. Hope this helps, John Yeo - Forwarded message -- From: Ballard Borich ballardbor...@juno.com To: pinhole-discussion@p at ??? Date: Thu, 21 Jun 2001 10:28:53 -0400 Subject: Re: [pinhole-discussion] Reciprocity Problem Message-ID: 20010621.102857.-222441.0.ballardbor...@juno.com Eric or John I have zero math apptitude or ability. I have always wondered what the actual curve for TMAX 400 film would be like. The guidelines from Kodak are @1sec +1/3 = 1 1/3 sec, @10sec + 1/2 = 15sec, and @100sec + 11/2 =300sec. I like the film just fine. It works well with my 120 film size toy cameras. I have just started using a 120 film size pinhole. I have a box of 5x7 film frozen to use with pinhole on a view camera. I have no darkroom and won't for a long time. I would like to cut down on the commercially processed (read exspensive) mistakes. I do understand that they are only guidelines. Thank you if you can help with the math because I just don't get it. Ballard On Tue, 19 Jun 2001 11:40:11 -0500 Eric Lawton lawtone...@hotmail.com writes: Hi Jeff, www.curvefit.com goes into a bit of detail if you're interested in the math - or maybe try a google search on linear regression and non-linear regression. Otherwise, I have a bit of mathematical ability and if you provide me with your data I could construct an equation for a curve. You might need a pocket calculator to evaluate the equation though. If you're interested in the math start by learning linear regression - it's fairly simple. You should use non-linear regression for your reciprocity failure data though (since reciprocity failure is, of course, a non-linear behaviour). Feel free to ask more questions. If you can't tell - I like math. Eric From: Jeff Dilcher r...@hiddenworld.net *** Warning *** *** Technical Question Follows ** ** press delete key if technical questions bother you ** I have a technical question, that, perhaps someone on the list can help me with. Please respond directly if you like. I am trying to develop a mathematical function which will help determine reciprocity compensations. Initially, I would like to be able to submit a small set of reference values (metered exposure times along with exposure time compensation for reciprocity). Using these, I would like be able to enter in any exposure value, and get a corresponding compensated value on the reciprocity curve. I am trying to boil this down into an algebraic equation as easily as is possible. I have quickly reached the endpoint of my math abilities. Can anyone provide a clue for me? _ Get your FREE download of MSN Explorer at http://explorer.msn.com ___ Pinhole-Discussion mailing list Pinhole-Discussion@p at ??? unsubscribe or change your account at http://www.???/discussion/
RE: [pinhole-discussion] Reciprocity Problem
recently i saw an old kodak manual on this problem. There is no curve to apply to reciprocity. It is a linear logarithmic function. perfect exposure changes with time at a certain distance logrithmicaly with unchanging aperture. www.glassplatenegativehow-to.com -- Original Message - From: Jeff Dilcher Sent: Tue, Jun 19, 2001 9:03am To: pinhole-discussion@p at ??? Subject: [pinhole-discussion] Reciprocity Problem *** Warning *** *** Technical Question Follows ** ** press delete key if technical questions bother you ** I have a technical question, that, perhaps someone on the list can help me with. Please respond directly if you like. I am trying to develop a mathematical function which will help determine reciprocity compensations. Initially, I would like to be able to submit a small set of reference values (metered exposure times along with exposure time compensation for reciprocity). Using these, I would like be able to enter in any exposure value, and get a corresponding compensated value on the reciprocity curve. I am trying to boil this down into an algebraic equation as easily as is possible. I have quickly reached the endpoint of my math abilities. Can anyone provide a clue for me? ___ Pinhole-Discussion mailing list Pinhole-Discussion@p at ??? unsubscribe or change your account at http://www.???/discussion/
Re: [pinhole-discussion] Reciprocity Problem
- Original Message - From: G.Penate pen...@home.com Don't think so, if you download Regression Analyzer http://www.zdnet.com/downloads/stories/info/0,,000QE3,.html you just need to input any pair of values of KNOWN reciprocity corrections (Y axis) corresponding to METERED values (X axis), select a Regression Analysis method (I suggest 2nd or 3rd Polynomial) and magically the program draws the graph and tell you the algebraic equation. IT REALLY SOUND MORE COMPLICATED THAN IT IS, the fact is that you don't even have the remote idea on how it is done. I meant to write: the fact is that you don't even have TO HAVE the remote idea on how it is done. My apologies. Guillermo
Re: [pinhole-discussion] Reciprocity Problem
I would like to enter in a meter reading, and have it give methe number of seconds (or minutes) for the exposure with all filter factors and reciprocity factors adjusted in. Factoring filters is easy, but I am hung up on reciprocity. If you don't eyeball your exposures, which I could bet you don't do, just take your meter readings through the filter, that way the uncorrected exposure already includes the filter factor. Remember that filter factors are just a guide, different scenes with differente predominant colors would be affected differently by a filter. From the responses I have received, it appears that I need to do some sort of nonlinear regression curve. I did some quick research on the net for information on that, I quickly became bewildered with alot of information. I might be biting off more than I can chew! Don't think so, if you download Regression Analyzer http://www.zdnet.com/downloads/stories/info/0,,000QE3,.html you just need to input any pair of values of KNOWN reciprocity corrections (Y axis) corresponding to METERED values (X axis), select a Regression Analysis method (I suggest 2nd or 3rd Polynomial) and magically the program draws the graph and tell you the algebraic equation. IT REALLY SOUND MORE COMPLICATED THAN IT IS, the fact is that you don't even have the remote idea on how it is done. I am not sure if this can definately be constructed with the program I am using or not. Obviously an outside program might not be to helpful, as I am trying to get exposure information on the fly in the field. The OUTSIDE program will be used ONLY to get the equation that describes the reciprocity corrections of your film. Having the equation, you use it with your MATHPAD to calculate ON THE FLY any reciprocity corrections for a given metered exposure time. You could even print the GRAPH the Regression Analyzer gives you and use it as a back up in case your PDA fails. Guillermo
Re: [pinhole-discussion] Reciprocity Problem
Most modern graphing calculators can find curves of best fit. I have a TI-83, and would be happy to provide more details on how to do it, or punch in the numbers if you don't have access to one. John - Original Message - From: Jeff Dilcher r...@hiddenworld.net To: pinhole-discussion@p at ??? Sent: Tuesday, June 19, 2001 8:03 AM Subject: [pinhole-discussion] Reciprocity Problem *** Warning *** *** Technical Question Follows ** ** press delete key if technical questions bother you ** I have a technical question, that, perhaps someone on the list can help me with. Please respond directly if you like. I am trying to develop a mathematical function which will help determine reciprocity compensations. Initially, I would like to be able to submit a small set of reference values (metered exposure times along with exposure time compensation for reciprocity). Using these, I would like be able to enter in any exposure value, and get a corresponding compensated value on the reciprocity curve. I am trying to boil this down into an algebraic equation as easily as is possible. I have quickly reached the endpoint of my math abilities. Can anyone provide a clue for me? ___ Pinhole-Discussion mailing list Pinhole-Discussion@p at ??? unsubscribe or change your account at http://www.???/discussion/
Re: [pinhole-discussion] Reciprocity Problem
Hi Jeff, www.curvefit.com goes into a bit of detail if you're interested in the math - or maybe try a google search on linear regression and non-linear regression. Otherwise, I have a bit of mathematical ability and if you provide me with your data I could construct an equation for a curve. You might need a pocket calculator to evaluate the equation though. If you're interested in the math start by learning linear regression - it's fairly simple. You should use non-linear regression for your reciprocity failure data though (since reciprocity failure is, of course, a non-linear behaviour). Feel free to ask more questions. If you can't tell - I like math. Eric From: Jeff Dilcher r...@hiddenworld.net *** Warning *** *** Technical Question Follows ** ** press delete key if technical questions bother you ** I have a technical question, that, perhaps someone on the list can help me with. Please respond directly if you like. I am trying to develop a mathematical function which will help determine reciprocity compensations. Initially, I would like to be able to submit a small set of reference values (metered exposure times along with exposure time compensation for reciprocity). Using these, I would like be able to enter in any exposure value, and get a corresponding compensated value on the reciprocity curve. I am trying to boil this down into an algebraic equation as easily as is possible. I have quickly reached the endpoint of my math abilities. Can anyone provide a clue for me? _ Get your FREE download of MSN Explorer at http://explorer.msn.com
Re: [pinhole-discussion] Reciprocity Problem
*** Warning *** *** Technical Answer Follows ** ** press delete key if technical questions bother you ** - Original Message - From: Jeff Dilcher r...@hiddenworld.net I am trying to develop a mathematical function which will help determine reciprocity compensations. As MS says, that is a good thing!! Initially, I would like to be able to submit a small set of reference values (metered exposure times along with exposure time compensation for reciprocity). Using these, I would like be able to enter in any exposure value, and get a corresponding compensated value on the reciprocity curve. That curve is known as BEST FIT. I am trying to boil this down into an algebraic equation as easily as is possible. I have quickly reached the endpoint of my math abilities. Can anyone provide a clue for me? Here is a clue: what you want to accomplish is known as REGRESSION ANALYSIS which give you as a result an equation for the BEST FIT CURVE, books on those topics will teach or refresh your knowledge on that. Since you mention the words as easily as possible, another way is to use one of the many REGRESSION ANALYSIS programs available (some shareware, some freeware) which will save you the tedious calculations needed. One program easy to use and very good IMO is Regression Analyzer, you can read about it here: http://www.mfsoft.com/equationgrapher/ Guillermo
[pinhole-discussion] Reciprocity Problem
*** Warning *** *** Technical Question Follows ** ** press delete key if technical questions bother you ** I have a technical question, that, perhaps someone on the list can help me with. Please respond directly if you like. I am trying to develop a mathematical function which will help determine reciprocity compensations. Initially, I would like to be able to submit a small set of reference values (metered exposure times along with exposure time compensation for reciprocity). Using these, I would like be able to enter in any exposure value, and get a corresponding compensated value on the reciprocity curve. I am trying to boil this down into an algebraic equation as easily as is possible. I have quickly reached the endpoint of my math abilities. Can anyone provide a clue for me?
