Page 11 of 12 FirstFirst ... 9101112 LastLast
Results 101 to 110 of 118

Thread: BTZS + Modern Incident Meter w/ Digital Readout

  1. #101

    Join Date
    Sep 2003
    Location
    Massachusetts USA
    Posts
    8,476

    Re: BTZS + Modern Incident Meter w/ Digital Readout

    Quote Originally Posted by Steve Nicholls View Post
    Highlight = Dome facing away from camera pointing at the brightest part of the window. Examine the window to see what is the brightest part of the scene and get as close as you can so the meter is evaluating just the highlight. That is your highlight reading.

    Shadow = Dome at subject position pointed to the camera.


    Rangefinder Test, 2012
    Mamiya 7, 150mm Mamiya N
    TMY, Pyrocat HDC

    I just gave your recommendation a try, and I have to admit that it worked much better than I had anticipated. This photo was made to test the accuracy of the rangefinder at medium distance. I got the high reading from the window itself, but I exposed for the shadow reading, as recommended in the BTZS book.

    To me, this illustrates one of the advantages of incident metering: not having to pre-visualize. Almost everything in the subject is white, and there was nowhere obvious for me to place any of the zones. With this approach, the zones (many of which don't fall on a roman numeral if you know what I mean) took care of themselves.

  2. #102

    Join Date
    Mar 2008
    Location
    Beaverton, OR
    Posts
    653

    Re: BTZS + Modern Incident Meter w/ Digital Readout

    Quote Originally Posted by Ken Lee View Post
    To me, this illustrates one of the advantages of incident metering: not having to pre-visualize. Almost everything in the subject is white, and there was nowhere obvious for me to place any of the zones. With this approach, the zones (many of which don't fall on a roman numeral if you know what I mean) took care of themselves.


    I know the feeling.
    You can't depend on your eyes when your imagination is out of focus. ~ Mark Twain

  3. #103

    Join Date
    May 2011
    Posts
    29

    Re: BTZS + Modern Incident Meter w/ Digital Readout

    Quote Originally Posted by Stephen Benskin View Post
    While we're on the topic of the program, exactly how does it calculate the flare factor? Can you explain the math here?
    I can answer this one for both Plotter and ExpoDev:

    Flare in ExpoDev is rather simple. The Film Profiles that are exported from Plotter contain no flare information at all (it always gets stripped since Flare in Plotter is just a simulation). Flare as set in ExpoDev is used to to modify Average Gradient that is calculated (which is derived from the SBR and the Paper ES). It is simple a factor which gives a little boost to the calculated Average G. Average G is then used to look up the EFS value from the imported film curve so effectively, Flare typically (but not always depending on the curve) gives a slight boost to the EFS that will then be used to set the exposure. It is a pretty simple compensation factor.

    Flare in Plotter is used to simulate the effects of Flare when analyzing film curves. The math behind it there is quite a bit more complicated. Flare is a density value and Plotter tries to find the exposure value that would produce that density from the data points in a film curve. Once it finds the exposure value that produces that density, it then uses that as a exposure component to adjust all the points in the curve to get the flare adjusted density values.

    So if FE is the exposure value that produces a Flare density of FD and for each point on a curve an exposure value of IE produces a density of ID, then the flare adjusted density for each point in a curve becomes the ID density for an exposure value of FE + IE. Because we're working with log exposure values, the effects of any arithmetically added flare exposure is variable over the curve.

    The effect this has is, as the exposure value for each point on the curve increases the effects of flare become less and less. This is because exposure values increase in stops but flare is a minor exposure component arithmetically added to each point. In other words, flare's effect in increasing density are more pronounced in the toe of the curve than further up the curve. In fact up near the shoulder the effects become so small that the original and adjusted curve effectively merge.

    Since Flare increases toe density its practical effect is a little like pre-exposure, boosting the shadow density in negatives. This is why ExpoDev boosts the EFS when using Flare to counteract this increased negative shadow value density to prevent overexposure (i.e. boosting EFS reduces the overall exposure, which brings the density down).

    This is probably way more information than anyone wanted, but there it is - the short version at least.

  4. #104

    Join Date
    May 2011
    Posts
    29

    Re: BTZS + Modern Incident Meter w/ Digital Readout

    Quote Originally Posted by Ken Lee View Post
    To me, this illustrates one of the advantages of incident metering: not having to pre-visualize. Almost everything in the subject is white, and there was nowhere obvious for me to place any of the zones. With this approach, the zones (many of which don't fall on a roman numeral if you know what I mean) took care of themselves.
    From what I remember in talking with Phil at workshops, this was his point: accurate pre-visualization is tough to get right. BTZS Incident metering was developed to make it easier to get consistently good negatives that will be easier to print (both through contrast management and matching to your printing paper's ES). It makes some assumptions though, mainly that you want to render a scene realistically.

    If however you want to apply your own interpretation to a scene, you can do that my modifying either shadow placement or highlight placement, effectively altering exposure and/or compressing or expanding contrast. Placing a certain subject at a certain zone was not the goal. Capturing a good negative that was easier to work with was.

    The other part that most people leave out when talking about BTZS is that it is also about managing the translation of scene values to print values. This is more complicated than just placing a scene value on a certain zone in the negative via exposure and development. This is because printing papers have a curve too.

    You can think of it like this: scene values are translated to negative density by way of the film's curve. Negative density values are translated to print values by way of the paper's curve. It's the combination of the two that tells you how a scene value will become a print value. This also means you can leverage the interplay between a film and paper curve to give a different interpretation of a scene, perhaps by expanding mid-tones using one combination of film/paper or compression them using another combination. Some combinations can expand or compress shadow values while others can compress or expand highlights or any combinations in between.

    Plotter has a feature called the Matcher which graphs this out and lets you compare how different combinations of films and papers work together. You can see an example of that here, in the last screen shot on this page: http://btzs.org/Software/PlotterScreens.htm

  5. #105

    Join Date
    Jul 2008
    Location
    Los Angeles
    Posts
    212

    Re: BTZS + Modern Incident Meter w/ Digital Readout

    Quote Originally Posted by david.jade View Post
    Since Flare increases toe density its practical effect is a little like pre-exposure, boosting the shadow density in negatives. This is why ExpoDev boosts the EFS when using Flare to counteract this increased negative shadow value density to prevent overexposure (i.e. boosting EFS reduces the overall exposure, which brings the density down).

    This is probably way more information than anyone wanted, but there it is - the short version at least.
    Thanks David for your response. I've written my own series of plotting programs including a 4 quad reproduction diagram so it's not more than I wanted. To tell you the truth, I've never been a big fan of combining the flare and film curve together. It's a construct that I believe gives a false impression of how flare works. It also eliminates important information that can only be obtained by using a camera image / flare curve. It's just my personal preference.

    I take it the boosting of the EFS with the ExpoDev is only with anything above average flare. As you know, speed point, while determined using a no flare curve, has flare incorporated into its determination. Does the ExpoDev then basically adjust the EFS for the difference between average flare and the higher flare value?

    With the development calculations, does it use a fixed flare model or some sort of variable flare model?

    I'm curious as to know the exposure point, H, that you use to determine the flare value? Let me see if I can clarify the question some. A flare factor of 2, or 1 stop, means the doubling of the shadow exposure. That value is then added to the subsequent exposures along the film curve. Where that point is determined is important in correctly defining the value of flare. Let's say for a 125 speed film, it is determined at the speed point 0.8 / 125 = .0064 lxs. This value is then added to the all the points of exposure. But if flare is determined from the statistical average shadow exposure for a 125 speed film, 0.0032 lxs, this value will have quite a different influence. While both values technically have a flare factor of 2, they produce very different results. So at what point does ExpoDev determine the value for the flare exposure?

  6. #106

    Join Date
    May 2011
    Posts
    29

    Re: BTZS + Modern Incident Meter w/ Digital Readout

    Quote Originally Posted by Stephen Benskin View Post
    I take it the boosting of the EFS with the ExpoDev is only with anything above average flare. As you know, speed point, while determined using a no flare curve, has flare incorporated into its determination. Does the ExpoDev then basically adjust the EFS for the difference between average flare and the higher flare value?
    When the Flare amount is zero, ExpoDev just uses the EFS data as exported from Plotter (which has no Flare model applied since it was stripped before export). EFS points as exported by Plotter may be based on 0.1 over FB+F or may be based on the Approximate CI method. So in ExpoDev, Flare is just an additional compensation over what Plotter has determined for EFS at the calculated Average G of the scene (as metered).

    Quote Originally Posted by Stephen Benskin View Post
    With the development calculations, does it use a fixed flare model or some sort of variable flare model?
    ExpoDev uses a very simple model of Flare when adjusting development. Basically, exposure is decreased by a small amount and development is extended by a small amount. We're talking about very small values here with the goal of controlling shadow density a little better while maintaining roughly the same contrast.

    Quote Originally Posted by Stephen Benskin View Post
    I'm curious as to know the exposure point, H, that you use to determine the flare value? Let me see if I can clarify the question some. A flare factor of 2, or 1 stop, means the doubling of the shadow exposure. That value is then added to the subsequent exposures along the film curve. Where that point is determined is important in correctly defining the value of flare. Let's say for a 125 speed film, it is determined at the speed point 0.8 / 125 = .0064 lxs. This value is then added to the all the points of exposure. But if flare is determined from the statistical average shadow exposure for a 125 speed film, 0.0032 lxs, this value will have quite a different influence. While both values technically have a flare factor of 2, they produce very different results. So at what point does ExpoDev determine the value for the flare exposure?
    Since Plotter starts with a user-supplied Flare density value (0.02 is typical), for each film curve it finds the point that corresponds to that density on the curve. From that point it then knows how much additional exposure it takes to produce that density increase in a negative. This is the extra exposure given to each point in the curve to determine the new density that the combined exposure would yield.

    In practice it gets a little more complicated as Plotter actually simulates a new set of curve points based on a fixed interval (1/2 stop I believe). It then uses this new curve as a sort of look-up table to translate the real curve points into the new exposure/density space, generating the final translated flare adjusted curve.

  7. #107

    Join Date
    Nov 1999
    Posts
    763

    Re: BTZS + Modern Incident Meter w/ Digital Readout

    Quote Originally Posted by Ken Lee View Post


    Rangefinder Test, 2012
    Mamiya 7, 150mm Mamiya N
    TMY, Pyrocat HDC

    I just gave your recommendation a try, and I have to admit that it worked much better than I had anticipated. This photo was made to test the accuracy of the rangefinder at medium distance. I got the high reading from the window itself, but I exposed for the shadow reading, as recommended in the BTZS book.

    To me, this illustrates one of the advantages of incident metering: not having to pre-visualize. Almost everything in the subject is white, and there was nowhere obvious for me to place any of the zones. With this approach, the zones (many of which don't fall on a roman numeral if you know what I mean) took care of themselves.
    Nicely executed Ken. Incident when used as Phil intended with BTZS is very simple. Many/most people cannot or won't entertain the idea that an incident meter can be used in any method other than a single reading pegged to their preconceived 18% or z5 reading of general light falling on a subject. Phil has taken a tool, the incident meter and created a metering method that looks at light quality more than anything else. Your question concerning a backlit subject is one I'm asked many times regarding BTZS and Incident metering - by ZS users - as if it cannot be done in any manner other than with a spot meter. Thanks for posting your results. For me the actual light [the anchor points of shadow and highlight] is the way I now see my photography. Everything else will fall where it should based on your film, developer and paper test combination.

    This is a good discussion of BTZS. Thanks to all who are contributing and gaining some additional insights.

    Steve

  8. #108

    Join Date
    Jul 2008
    Location
    Los Angeles
    Posts
    212

    Re: BTZS + Modern Incident Meter w/ Digital Readout

    Quote Originally Posted by david.jade View Post
    When the Flare amount is zero, ExpoDev just uses the EFS data as exported from Plotter (which has no Flare model applied since it was stripped before export). EFS points as exported by Plotter may be based on 0.1 over FB+F or may be based on the Approximate CI method. So in ExpoDev, Flare is just an additional compensation over what Plotter has determined for EFS at the calculated Average G of the scene (as metered).
    I'm probably not being very clear with my question. The example below shows the ratio between the camera exposure and speed point of 0.10 over fb+f (point B) in addition to where the average scene's shadows fall (point A). The speed point is approximately one stop above where the shadows fall. Average flare brings the shadows (A) up from to around the speed point (B). Now, if I understand your response correctly, speed data is imported into ExpoDev where it compensates the EFS for flare. The way I'm interpreting what you are saying is that the ExpoDev assumes the speed data is from a no flare model and any flare value input into ExpoDev will then affect the EFS. If this is correct, then ExpoDev is adjusting the EFS for flare, up to average flare, which has already been compensated for.

    Click image for larger version. 

Name:	Speed Point - Standard Model.jpg 
Views:	10 
Size:	184.5 KB 
ID:	70366

    What I'm saying is that the placement of the speed point Hm at a difference of 1.0 log-H from P (metered exposure point 8 / ISO) assumes flare as part of the speed equation as the below example illustrates using the standard exposure model for a 125 speed based exposure. In effect, the below example is essentially the exposure condition that the ISO black and white film standard assumes when determining film speed at the 0.10 over Fb+f speed point. If the ExpoDev assumes the addition of any flare on top of that, up to the value of average flare, then any compensation to the EFS would be compensating for double the effect of flare and under rate the film. Am I correct in assuming that when 1/2 stop of flare is input into ExpoDev, it will then adjust the EFS based only on that value?

    Click image for larger version. 

Name:	1 stop flare - shadow.jpg 
Views:	10 
Size:	45.3 KB 
ID:	70371

  9. #109

    Join Date
    Dec 2009
    Location
    Pacifica, CA
    Posts
    1,710

    Re: BTZS + Modern Incident Meter w/ Digital Readout

    Quote Originally Posted by david.jade View Post
    The other part that most people leave out when talking about BTZS is that it is also about managing the translation of scene values to print values. This is more complicated than just placing a scene value on a certain zone in the negative via exposure and development. This is because printing papers have a curve too.
    This is where the original Power Dial fascinated me, because the steps to make it involved some intense graphic arts knife work (which I could feasibly do given my graphic arts background). Then you expose some of your paper at varying times. The result is a real estimation of the tone for the exposure given the development. Amazing idea. Why I wonder if anyone ever made one of these...

  10. #110

    Join Date
    Jul 2008
    Location
    Los Angeles
    Posts
    212

    Re: BTZS + Modern Incident Meter w/ Digital Readout

    This is what I was talking about concerning where the value of flare is calculated. The example below has the standard model where a flare factor of 2 (1 stop) is calculated from the shadow exposure (0.0034) making it (0.0068). The other example calculates the flare value from just a little over the speed point (0.0068) making the shadow exposure (0.0098).

    Click image for larger version. 

Name:	Where the Value of Flare is measured.jpg 
Views:	11 
Size:	93.1 KB 
ID:	70372

    These are the equations I used to calculate H for a one stop flare factor. In my program, a number of the values are variables.

    Click image for larger version. 

Name:	Calculating Flare.jpg 
Views:	9 
Size:	126.2 KB 
ID:	70373

Similar Threads

  1. BTZS: Incident Readings in Flat Light
    By Ken Lee in forum Style & Technique
    Replies: 25
    Last Post: 2-Jan-2013, 19:12
  2. Incident Metering Again. BTZS.
    By Sinar-Man in forum Style & Technique
    Replies: 8
    Last Post: 31-Aug-2010, 22:13
  3. BTZS vs Zone? Incident v
    By cyrus in forum Style & Technique
    Replies: 34
    Last Post: 29-Aug-2006, 20:33

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •