Fomapan 200 film development test results, also reciprocity

[alanrockwood]

Hi,

I tested Fomapan to figure out what development time I should use if using HC-110 unofficial dilution H (1 + 63). I also tested for Exposure index and contrast and reciprocity failure.

I did the development in a phototherm rotary processor at 73.9 F developed for 5 minutes 20 seconds. I didn't test other development times because this time gave me the contrast I was looking for. Attached is a characteristic curve, which some comments posted on it. I ended up with a G-bar of 0.54 and an E.I. of 75, which would round to the 80 as the closest standard speed. However, see an additional comment later in this post.

My reciprocity test gave the following result

ta = tm + 0.4 * tm^1.7

Where tm is the metered exposure time, and ta is the adjusted exposure time. Times are expressed in seconds. I used a modified Gainer equation, but I used an adjustable exponent instead of fixing it at 1.62 as Gainer did, and I also used an adjustable multiplier. The adjustable parameters were adjusted to give the best fit to the data, resulting in 0.4 and 1.7 as shown in the equation. This was based my experimental results on metered exposure times out to four seconds, but I suspect the equation would extrapolate fairly well out to at least tm=60 seconds, and maybe a bit higher, but exposures get to be pretty long very fast, so you probably don't want to try very long values for tm.

Now, back to the development test. Here's the curve.

The curve is basically upswept. I didn't go high enough in relative exposure to find the shoulder. The shape in the toe is a bit odd. It looks like it hits zero with a steep slope, whereas it should curve more and approach baseline more gradually. My theory is that In the toe region I was seeing the onset of short exposure reciprocity failure. (I varied the exposure by changing the shutter speed. The exposure times in the exposure series varied from 1/2000 to 1/8 seconds.) If this idea is true then the curve should nosedive into baseline with an unexpectedly steep slope, which is actually what we are seeing in the data. If this is true, and if short exposure reciprocity failure was starting to creep in a little bit by 1/750 seconds, the film speed would actually be a little faster than what I calculated, so a personal E.I of 100 us probably not unreasonable. For the G-bar calculation I used 1.5 relative exposure units above the speed point, calling that the G-bar point. For the film speed calculation I used an upper speed point of 3.3 stops of exposure above the speed point. That is approximately one relative exposure unit above the speed point. I call this point an upper speed point.

With the upswept curve the separation of highlight values should be very good and the separation of shadow values not so good. I read somewhere that Plux-X had an upswept curve. If so then Fomapan 200, if used with this development protocol, might be a reasonable substitute for Plus-X. I think this kind of curve could be good for portraiture of light-skinned people, but not so good for portraiture dark-skinned people. An S-shaped curve might be better for dark skinned people, or a straight line curve might be even better... TMax 100 anyone?

A note about developer volume: Dilution H gave ~6 ml of HC-110 concentrate, in the ~400 ml phototherm tank. More films in the tank would risk developer starvation.

Density of base + fog was 0.354, which was subtracted from the displayed curve.

I used a digital camera as a densitometer. The procedure was rather elaborate in order to get good results. I could explain how I did this, but it is lengthy, so I won't explain it here at this time. I used this method because my densitometer suddenly went kaput. The fact that the scatter of the data is low relative to the fitted line suggests that the densitometry using a digital camera in place of a real densitometer was actually pretty successful.

I lucked out on the development time. I got the results I wanted on the first try rather than having to try a bunch of different development times.

I just thought that some of these results could be useful to others.

Note added later: I should have done a better job labeling the x-axis. It is actually the log (base 10) of the exposure. For example, the furthest-left point on the graph was measured using a shutter speed on my digital camera (which is not a densitometer, but plays one on TV) was 1/2000. The log of 1/2000 is -3.301. I am replacing the original figure with one with a new label on the x-axis

 

[Raghu Kuvempunagar]

I used a digital camera as a densitometer. The procedure was rather elaborate in order to get good results. I could explain how I did this, but it is lengthy, so I won't explain it here at this time. I used this method because my densitometer suddenly went kaput. The fact that the scatter of the data is low relative to the fitted line suggests that the densitometry using a digital camera in place of a real densitometer was actually pretty successful.

This is very interesting and potentially very useful to many. I hope you'll write about it somewhere.

 

[laser]

Plotting on equally incremented log-log paper will not change the data but will yield curves that are more conventional. i.e. 1.0 density will be the same scale length as Log 1.0 exposure. Congratulations on using a digital camera as a densitometer.

 

[Paul Howell]

I don't use HC 110 and don't have a densitometer, but testing using a ring around I got an E.I of 160 in most developers I use, currently Clayton F79+, but got 200 with Rodinal and MCM 100. I assume I would also shoot at 160 with D76, maybe 200 with DDX.

 

[FotoD]

Thanks for sharing!

 

[alanrockwood]

Plotting on equally incremented log-log paper will not change the data but will yield curves that are more conventional. i.e. 1.0 density will be the same scale length as Log 1.0 exposure. Congratulations on using a digital camera as a densitometer.

Thanks for the note. I should have included on the x-axis label that the axis is actually log exposure, not linear exposure. I have swapped the original figure with one with a more correct x-axis description.

 

[alanrockwood]

I don't use HC 110 and don't have a densitometer, but testing using a ring around I got an E.I of 160 in most developers I use, currently Clayton F79+, but got 200 with Rodinal and MCM 100. I assume I would also shoot at 160 with D76, maybe 200 with DDX.

The "best" E.I. to use for fomapan 200 seems to be rather controversial. Most people suggest numbers similar to what you are using. Some suggest numbers in the range of 100 to 125. My results may have been influenced somewhat by the fact that HC-110 is often said to be a speed reducing developer. However, I was a bit surprised when my test came out so low (75 rounded to the closest conventional E.I. of 80). I expected it to come out somewhere in the 100 to 125 range. Of course, as I noted in my post, my tests might have been influenced a little bit by reciprocity failure in the low-exposure region, and if so it would push the measured E.I. up somewhat. How far, I don't know, but my wild guess would be that it would push the E.I. up to ~100.

I believe the criterion I used to determine the E.I. (~3.3 stops above the speed point) is at least somewhat favorable to preserving shadow detail. Some people might prefer better preservation of shadow detail. I think the zone-system aficionados like to use 4.0 stops above the speed point, which would favor preservation of shadow detail even more. On the other hand, some people would be satisfied with sacrificing a little bit of shadow detail in order to use a higher E.I. At least that's they way I interpret what I have read on the subject.

If film is scanned it might favor a bit higher E.I. in order to make a less dense negative in order to limit the amount of grain showing up in the image, and also because one could dig a little further down into the toe to recover detail in post processing.

 

[laser]

Thanks for the note. I should have included on the x-axis label that the axis is actually log exposure, not linear exposure. I have swapped the original figure with one with a more correct x-axis description.

I didn't explain myself clearly enough. Density and Log Exposure are both log base 10. To get the "proper" slope the density scale of 1.0 should have the same length as a Log E range of 1.0. To check it: The distance from Log 0.0 to Log 1.0 or any otehr Log 1.0) should equal the distance from density 0.0 to 1.0. (Or any otehr density range of 1.0. )The slope of a line from Log 0.0, density 0.0 to Log 1.0, density 1.0 = slope 1.0.

Briefly stated: when plotted, the distance for 1.0 log E = the distance for 1.0 density


With both log log scales the same it will be easier for you to interpret the results.

 

[laser]

Additional comment: It will be more accurate and easier if you use a step tablet (.15 or .3 density increments) instead of shutter speeds that are complicated by shutter inaccuracy and reciprocity. It will also use less film.

 

[alanrockwood]

I didn't explain myself clearly enough. Density and Log Exposure are both log base 10. To get the "proper" slope the density scale of 1.0 should have the same length as a Log E range of 1.0. To check it: The distance from Log 0.0 to Log 1.0 or any otehr Log 1.0) should equal the distance from density 0.0 to 1.0. (Or any otehr density range of 1.0. )The slope of a line from Log 0.0, density 0.0 to Log 1.0, density 1.0 = slope 1.0.

Briefly stated: when plotted, the distance for 1.0 log E = the distance for 1.0 density


With both log log scales the same it will be easier for you to interpret the results.

I calculated the slopes by taking the coordinates of the relevant points from the fitted result. It is a table of numbers, and I could just pick them off the table. That way it doesn't matter what the scaling of the coordinates is of the graph that displays the data.

 

[alanrockwood]

Additional comment: It will be more accurate and easier if you use a step tablet (.15 or .3 density increments) instead of shutter speeds that are complicated by shutter inaccuracy and reciprocity. It will also use less film.

That's right in theory, but it also requires a way to expose the film to the step tablet. I actually have a sensitometer that has a built in step tablet, but that entailed some additional complications. For one, the exposure time is rather long in that device, which risks getting onto the edge of the reciprocity failure time frame. Another is that the patches are small, which wouldn't be a problem if my densitometer were working, but might be too small to use in my digital-camera-as-a-sensitometer scheme.

The shutter speeds in my digital camera should be very accurate because they are electronically timed (I assume by a crystal-controlled oscillator in the camera). Most likely they are more accurate than a step tablet. The complication is that they are not continuously variable. However, I used a trick to get around that. I bracketed the exposure and then used interpolation to zero in on an accurate number. Also, the shutter in the film camera I used to expose the film is electronically controlled, so the same comment applies to that part of the process, except that there was no interpolation involved in that part of the process. That's a bit of an incomplete explanation, but if I get around to doing a more detailed write up it should become clearer.

Anyway, you are right in your comment, but it wouldn't quite work for me at the present time, so I improvised another solution.

 

[takilmaboxer]

I have a copy of Kodak's Publication F-8 regarding Plus-X, and it doesn't actually have an upswept curve. If I knew how I'd cut and paste the curves here
Your data certainly demonstrate that the ASA rating is optimistic! Thanks!

 

[alanrockwood]

I have a copy of Kodak's Publication F-8 regarding Plus-X, and it doesn't actually have an upswept curve. If I knew how I'd cut and paste the curves here
Your data certainly demonstrate that the ASA rating is optimistic! Thanks!

Thanks for the information about Plus-X. Question: was there more than one version of Plus-X, one for roll film and one for sheet film? If so the sheet film version might have had a different characteristic curve.

 

[Adrian Bacon]

I heavily shoot Foma 200. It’s 125-160 in pretty much every developer I’ve run it through, though to be fair, I haven’t run it through HC-110 as I’m just not an HC-110 user. I’ve posted a characteristic curve in the resources section if memory serves.

 

[takilmaboxer]

Re: Plus-X, back when I was using it a lot there were two basic versions: a short toe version (Plus-X) and a long toe version for studio use (Plus-X Pan Professional). Verchrome Pan was the short toe version but without a retouching surface on both sides. This was the late 70s into the 80s. I used them all in 120 and preferred Verichrome Pan. Tried Foma 200, didn't care for it, but I love the 100, grain and all.

 

[alanrockwood]

I heavily shoot Foma 200. It’s 125-160 in pretty much every developer I’ve run it through, though to be fair, I haven’t run it through HC-110 as I’m just not an HC-110 user. I’ve posted a characteristic curve in the resources section if memory serves.

Thanks Adrian. I don't know how I missed it, but before your post I have to admit I didn't even know the resources section existed. There is a wealth of information there, to which everyone else is saying, "well duh" to my ignorance of the existence of that section.

 

[laser]

Re: Plus-X, back when I was using it a lot there were two basic versions: a short toe version (Plus-X) and a long toe version for studio use (Plus-X Pan Professional). Verichrome Pan was the short toe version but without a retouching surface on both sides. This was the late 70s into the 80s. I used them all in 120 and preferred Verichrome Pan. Tried Foma 200, didn't care for it, but I love the 100, grain and all.

This is correct. The short toe originally was intended to accommodate the flare of consumer cameras. In the 1980s the primary difference between 120 PX and 120 VP was that PX had a anti-halation under coat (AHU) and VP didn't. Most VP was sold in developing countries so manufacturing cost was important. ,

 

[laser]

I calculated the slopes by taking the coordinates of the relevant points from the fitted result. It is a table of numbers, and I could just pick them off the table. That way it doesn't matter what the scaling of the coordinates is of the graph that displays the data.

I don't doubt that you can calculate the slope. Slope is a crude measure. If you are going to evaluate sensi consider looking at over-laid curves. Perceived differences are the width of a thin pencil line on full page plots. The proper spacing of the density and log exposure scales makes comparison easier. Having curves on separate sheets allows speed and fog compensation. The same thing can be done with graphing computer programs.

 

[alanrockwood]

I don't doubt that you can calculate the slope. Slope is a crude measure. If you are going to evaluate sensi consider looking at over-laid curves. Perceived differences are the width of a thin pencil line on full page plots. The proper spacing of the density and log exposure scales makes comparison easier. Having curves on separate sheets allows speed and fog compensation. The same thing can be done with graphing computer programs.

Regarding calculation of the slope, strictly speaking what I calculated was not the slope of the curve but was (mathematically speaking) the slope of a secant line. A secant line is a line between two points taken from the curve. The two points are indicated on the graph, the lower speed point and G-bar point. Note: "secant" as used here, is not the same thing as the trigonometric function that goes by the name "secant. "

 

[Adrian Bacon]

Thanks Adrian. I don't know how I missed it, but before your post I have to admit I didn't even know the resources section existed. There is a wealth of information there, to which everyone else is saying, "well duh" to my ignorance of the existence of that section.

Also, if you look at the spec sheet, what Foma considers normal exposure is between 1 second and 1/1000 of a second. Anything outside of that range is going to show some form of failure.

 

[alanrockwood]

I decided to check the short-exposure end of the reciprocity scale for Fomapan 200. I found an evenly illuminated object (sky on an overcast sky) and photographed it at shutter speeds 500, 1000, 2000, and 4000. The aperture was adjusted by so each frame had the same (midrange) exposure. The idea here was to see if there was reciprocity failure in this range. It was not to try to find adjustment factors to correct for reciprocity failure. Of course, adjustment would not be necessary if it were to turn out that there was no reciprocity failure

Here are the results in the form of a plot. The x-axis is the logarithm (base 10) of the shutter speed. The y-axis is density.

The left-most point in the plot is for 1/500 seconds (plotted as log(500)). The right hand point is is for 1/4000 seconds (plotted as log(4000)). I used the logarithm to linearize the x-axis to make evenly-spaced points.

As you can see, for a midrange tone there is no hint of reciprocity failure even at shutter speeds of 1/4000 second.. From the way reciprocity works, this says nothing about the highlights under high illumination (and consequently short shutter speed), one way or the other, but it does mean that there is no reciprocity failure for the shadows and mid-tones.

This is interesting because reciprocity at low image brightness (long shutter speeds) is rather bad for Fomapan 200, but it is really good for high image brightness (short shutter speeds).