Ilford FP4+ Characteristic Curve

[Paul Howell]

If you can live not having as much air flow as you might want think about a strip of weather seal on the bottom of the door and the inside of the door jams. Last option is a darkroom curtain made of black out cloth, large enough to over cover the door, close fitting curtain rod above the door so it can pulled aside when not in use.

In terms of Zone testing, I don't have access to a densitomter, have testing my EI in a precise way in decades, my memory is a little rusty, but is Phil Davis's BTZ method more accurate? At one there were a number of members who were very committed, have seen many discussions in past few years.

 

[Bill Burk]

Adrian. Instead of just trying to hit a target development time, I am hoping you’ll start to build time/contrast charts from all the testing you are doing.

Then you can be flexible determining development times for different purposes.

 

[Adrian Bacon]

Adrian. Instead of just trying to hit a target development time, I am hoping you’ll start to build time/contrast charts from all the testing you are doing.

Then you can be flexible determining development times for different purposes.

Great minds think alike.

Yes, that is a long term goal. I’ve not discarded the times I’ve got so far that where not zone normal, so as I have time to make the exposures and run the film, it’ll build out. I’ve been trying to stick to full even minute increments for the time slots for the collection points. It should be easy enough to interpolate between them for partial minutes.

 

[Bill Burk]

There’s a set people like to test, I can’t remember the sequence right now. Something like a minute or two apart for the quickest times and then spread out further as developing time gets longer.

Something like 4, 6, 9, 13, 20

 

[Adrian Bacon]

There’s a set people like to test, I can’t remember the sequence right now. Something like a minute or two apart for the quickest times and then spread out further as developing time gets longer.

Something like 4, 6, 9, 13, 20

That sounds about right. I've been doing mine a little bit different, where I try to keep it based off of a 10 minute start time. For example, start at 10 minutes, cut the time by 25%, puts you at 7:30, cut another 25% (from 10 minutes, 50% total), puts you at 5 minutes. The time between 7:30 and 5 is 6:15, but I round down to 6 minutes flat (6 minutes is a pretty common dev time). Going the other way from 10 minutes, adding 25% is 12:30, adding another 25% (from the 10 minutes, 50% total) is 15 minutes, then 17:30, then 20 (assuming I need to go that far), etc..

I know it sounds strange to do it that way, but it only takes two or three data points where you changed the time by a certain percentage to see how much the middle gray density moved for that percentage of time change. From there, it actually starts to get very predictable, where I can guesstimate with a pretty high degree of accuracy how much time I need to have to put my correctly exposed middle grey at a certain density. From there, if I know it's under exposed, by a stop lets say, it's not that hard to start at one of my data points, look at the density for 1 or two stops over/under and make a relatively accurate estimate for what time it will take to get that up (or down) to where middle grey would normally be if correctly exposed.

You also know what the contrast is, though, the way I scan, that's less useful to me. My setup always puts the film base plus fog 6 stops below middle gray, then I have a standard tone curve that I use for most films that would put a correctly exposed grey card at 46.6% luminance in Adobe Lightroom, with the highlights topping out at 100% luminance 7 stops up from middle grey, for a total of 13-14 stops from pure white to as black as it's going to get. From there, it's a matter of knowing what time you need to have to make a grey card exposed however it was exposed sit at middle grey. If it was under exposed, then instead of 5-6 stops between middle grey and film base plus fog, you'll have 4-5, or less if you really under exposed it. The same goes the other way for over exposing. Your visible contrast will contract and expand as you adjust the time.

For films that have wonky tone curves, or if I need to do something very specific, I just make a tone curve for that situation and use that instead. The end goal is to always make the scanned image look normal no matter how it was shot or developed. Most people that send film in to be processed can't even tell you how they exposed it. It's usually something along the lines of "I put it in the camera and it worked". In those instances, I run normal time, then adjust the scanning tone curve to more or less contrast on the frames that need it. I have contrast tone curves from 0.24 to 0.92 gamma in 0.02 increments, with 0.40 to 0.72 gamma in 0.01 increments, so once you put the FB+F at -6 down, if the standard contrast doesn't look right (too dark or too bright), you just crank it up or down until it does. It's basically the digital equivalent to variable contrast paper where you get your max paper black then adjust the contrast until it looks the way you want it to. You can even get nutty and do things like make multiple versions at different contrast levels, stack them as layers in photoshop with layer masks then mask in and out the different parts you want to make a combination image that for example could have clouds where you dialed down the contrast to bring them down to less brightness, medium middle tones, and punchy as all get out shadows and blacks. Total sacrilege, I know. But if you do it right on that special image, it looks pretty awesome. I've probably gone too far outside of analog land for this thread, but this is very much the reason for doing a lot of this. A well executed analog print is a beautiful thing, but digital gives a lot more flexibility (and complexity), and with modern high resolution pigment printers and inkjet papers, well executed digital output quality can be very high. I've said enough.

 

[Bill Burk]

It sounds funny to hear you talk of gray values. @DREW WILEY where are you? Help me explain this. You should think of it as a whole trip to the mountains. Sometimes you go up the East side where the road just goes up from the get go (TMAX etc.) other times you go up the West side (Tri-X). The road starts in the foothills and goes gently until you see bear clover, then it takes a steep trajectory. The whole trip matters.

So take any two points and find how far you go up in that distance. That’s your contrast as @Stephen Benskin says - that’s what matters. I really like Kodak’s CI (Contrast Index) because it looks at two points you are really going to use.

But you can use Gamma or Ilford’s measure, but what’s important is how steep the grade, not any one point.

 

[DREW WILEY]

Let's talk about real mountains instead, and the non -hypothetical challenge of getting discernable deep shadow values in black volcanic rock all the way up to sparking specular highlight in ice crystals atop brightly sunlit snow, and keeping it all sparkly in the print, and not all mashed flat through minus development. That ability separates the men from the boys when it comes to film linearity. FP4 just couldn't handle the extreme range like Tax films or Bergger 200, though it is one of my favorite sheet films. I like ACROS in the mtns for a different reason; it has less scale, but it's orthopan sensitivity renders foliage and so forth more naturally. I don't know where you're from, Bill. But I lived on the west side of the Sierra and it has some of the steepest terrain on the continent, as well as the deepest canyons. From a brief walk up the road from my house I could view up-canyon where it is 8500 feet deep. Sadly, it just got to be too much work dealing with all the forest fire prevention and maintenance issues approaching retirement, so I sold the place. The opposite direction there was about a hundred square miles down-river just too rugged to be inhabited at all. Quite different from the gradual west-side road approaches to Lk. Tahoe etc. way further north, and twice as high at the top. Not a Tri-X curve at all; more like lith film in lots of places.

 

[Bill Burk]

I had a cabin in Holy Jim Canyon (still standing but feel sad for my Trabuco neighbors who just got burned twice. First their cabins burned. Now the Forest Service is playing the “return it to nature” card).

At the same time I lived in Camp Nelson which is up one of those steep roads. Thirty miles across for a mile up.

Neither of these homes presented difficult photographic challenges. Not much water or rock. Lots of dirt and trees. So everything’s gray and you have to look for subtlety.

In my analogy though, I had Kings Canyon in mind.

 

[DREW WILEY]

I lived right on the edge of the San Joaquin Can. The N.Fork of the Kings was right uphill and the combined main lower Kings 15 mi south. I took a 100 mi backpack in the Kings 2 yrs ago, and a shorter one last fall. There's still part of the Middle Fork I'd like to see, but nobody past their early 60's has ever gotten in there, and I'll be 70 in a few months, so dunno. A real ornery storm stopped me last time. All that country was basically my back yard. The Sierras are mainly a fault block range tipped west, so has that classic steep east front above Owens Valley and Mono Basin. But the uplift was so quick after the Pliocene that the most extensive glaciation in the subsequent Pleistocene was on the more extensive, ocean-facing west slope, so that's where the deepest canyons are. Haven't made up my mind which film I'll take with me this summer and fall. Too many good choices out there.

 

[Adrian Bacon]

It sounds funny to hear you talk of gray values. @DREW WILEY where are you? Help me explain this. You should think of it as a whole trip to the mountains. Sometimes you go up the East side where the road just goes up from the get go (TMAX etc.) other times you go up the West side (Tri-X). The road starts in the foothills and goes gently until you see bear clover, then it takes a steep trajectory. The whole trip matters.

So take any two points and find how far you go up in that distance. That’s your contrast as @Stephen Benskin says - that’s what matters. I really like Kodak’s CI (Contrast Index) because it looks at two points you are really going to use.

But you can use Gamma or Ilford’s measure, but what’s important is how steep the grade, not any one point.

Perhaps you misunderstood my post. This is why I make test exposures and characteristic curves. I want to see what the film is doing so that I can decide if a generic curve will be fine, or if I need to make something more specific. More often than not, a generic curve works just fine, but if you want to do something specific, then you do need to work out where the inflection points in the curve are, and how much contrast there is between the inflection points. Describing with words tends to lead to misunderstandings, so it's easier to just show results: CatLABS X film 80, developed for 9 minutes, gives a touch over ISO contrast, relatively straight, with a pretty generic looking toe. All the inflection points are about where I'd expect them to be, so it's the standard generic ISO contrast curve.

Contrast wise, I have full shade on the underside of the building, with full afternoon sky and clouds. I can see details on the underside of the building, the sky and clouds aren't blown out, and I'm getting a nice reflection of the sky and clouds off the glass face of the building (which is mostly west facing and also has the sun beating down on it). It's not particularly flat, nor is it particularly punchy. Perfect for tuning up into something if I were so inclined to do so.

The key takeaway here is that there's more than one way to skin a cat. If it's not the way you'd skin it, that doesn't mean it's wrong, it's just a different way.

 

[DREW WILEY]

FP4 is fairly predictable unless you want to do something exotic, and the published curves are representative : A brief but distinct toe, then a long moderate-gamma straight line; but it can shoulder off if you go too far. I have densitometer plotted many families of curves for FP4 for special applications, however, with reference to a variety of developer tweaks. The main piece of advice I'd give for general shooting is, if encountering a high-contrast scene, significantly boost exposure by bringing the deep shadow values up onto the straight line of the characteristic curve. I generally use ASA 50 for FP4, and only in mild contrast scenes use ASA 100. Of course, the rules can be bent for interpretive expression, for example, if you deliberately wanted blacked-out shade in the above picture. It's not a film I'd choose out in the redwoods under open sun, where the contrast range can run 12 stops. But otherwise, FP4 is an exceptionally versatile and easy-to-use film.

 

[Vaughn]

Dang...did someone just mention redwoods?! I love FP4+...I cook it in paper developer.

 

[DREW WILEY]

Carbon printing is obviously a little different in terms of responding to the entire film curve. My favorite film out in the woods was Bergger 200, once Super-XX was gone. Now that it's gone too, TMax films have the longest straight line available, with the exception of Foma 200, which is just too slow and afflicted with poor recip characteristics to be of much use in deep woods, not to mention its dicey quality control.

 

[Bill Burk]

Perhaps you misunderstood my post.
Contrast wise..., I have full shade on the underside of the building, with full afternoon sky and clouds.

I was thinking of our entire conversation not just one post. What I mean by contrast isn't in the context of how contrasty the final result is, but the degree of development measured as a decimal fraction with a couple decimal points of accuracy.

I get the feeling you are aiming for a point of negative density of 0.72 for metered gray as an aim. You can skin your cat that way, because you are graphing a complete curve and checking how well you fit the entire curve to the reference line.

Your graphs are perfect. We're talking now about interpreting them. And I want you to expand your test times to cover a wide range of aim contrasts.

Contrast for that purpose is the fraction "rise over run" so in my mountain example where I drove 30 miles from work to my cabin and went from sea level to a mile elevation, that number is 1/30 or 0.03

In developing film you know the ASA contrast figures out near 0.62

For my Zone System work, I use the following contrast aims: N+2 0.95 CI, N+1 0.75 CI, N 0.62 CI, N-1 0.53 CI, N-2 0.46 CI
I picked these off this chart starting with just saying 7 stops is Zone System N: (http://beefalobill.com/imgs/Contrast Indexes - Kodak.jpg)

The main point I want to make is that rather than aiming for a perfect Zone System N time or ASA fit, you should test a family of development times.

Try to get a spread of results in the range of contrasts from 0.95 to 0.46

From that you can make a Time/Contrast curve that will give you development times for any contrast you want between your test times (and you can extrapolate past the test times a little bit).

Then if Vaughn brings you a box of film labeled N+1 and you know he wants to carbon print... You can work it out and develop for a time that will hit 1.21 he needs. Just a guess I don't know how much density range Vaughn needs but it's way more than I usually need.

 

[Stephen Benskin]

Then if Vaughn brings you a box of film labeled N+1 and you know he wants to carbon print... You can work it out and develop for a time that will hit 1.21 he needs. Just a guess I don't know how much density range Vaughn needs but it's way more than I usually need.

This is where the true superiority of a sensitometry and tone reproduction theory approach shows. All you need is a film's curve family and a developmental model and development can be tailored to a mired of situations.

Conceptually, the developmental model is the most difficult to accurately define. The idea should be simple, divide the aim negative density range or paper LER over the scene's subject Luminance range. This will give the CI or Average gradient need to match the scene to the print. A scene with the statistically average Luminance range is what is considered Normal development.

Unfortunately, there is another variable to consider. The actual Illuminance range striking the film consists of the subject Luminance range plus flare. Veiling flare adds illumination to the shadows, effectively reducing the subject's Luminance range at the film plane. How much depends on whether the lens elements are coated, how many elements in the lens, and the interior structure of the camera. On average flare under statistically average conditions are around 1 to 1 1/3 stops. The equation to determine aim CI becomes NDR / (LSLR - Flare).

The chart that Bill references was given to me by Dick Dickerson, head of Kodak's B&W R&D Department. His team invented the T-Max films and Xtol. It is based on a fixed flare model with 0.40 as the value for flare. While flare tends to increase 1/3 of a stop per stop increase in Luminance range and 1/3 decrease with each stop reduction in the subject Luminance range, the model is good enough as flare is variable based on not only the camera's optical system or the scene's Luminance range, but also the tonal distribution within the scene. A greater degree of lighter tones will produce a higher degree of flare. And since flare is extremely difficult to measure, calculating the influence of flare for any condition is limited to best guess.

Any degree of flare doesn't matter too much until developing for more extreme situations like -2 and below and +2 and above. Arguments can be made for a variable flare model, and I've come up with one that is a balance between the two and that conforms to psychophysical preferences.

Anybody that has worked in a lab knows that when a customer asks for +1 development, they mean the film was rated one stop under the ISO speed. Development for "speed" is different than developing for contrast. Different but not impossible to determine. The model Kodak uses is almost identical to the variable flare model.

All possible from one series of film tests.

 

[Adrian Bacon]

I get the feeling you are aiming for a point of negative density of 0.72 for metered gray as an aim.

For a film correctly exposed at box speed and developed to normal contrast, yes. As previously stated, I also plan to chart other development times so ultimately there will be a family of curves for every film I can buy today.

Part of what I suspect may be causing some “churn” if you will is we use the data differently from each other. You use it primarily for zone work, I use it primarily to scan. Things that are important to you are less so to me, and things that I care about a lot doesn’t really register with you. If I can make it work for both then I’m happy to do so, however the number of people who send me film who can even tell me how they exposed and how they’d like it developed is infinitely small, and even fewer of those people actually want prints, so in some respects we’re going to be pretty far apart, as there are times where I break from what you’d expect to see from a zone system technique and do what gives me better results in a digital lab environment. There’s nothing wrong with either way, they serve different purposes.

 

[Lachlan Young]

For a film correctly exposed at box speed and developed to normal contrast, yes. As previously stated, I also plan to chart other development times so ultimately there will be a family of curves for every film I can buy today.

Part of what I suspect may be causing some “churn” if you will is we use the data differently from each other. You use it primarily for zone work, I use it primarily to scan. Things that are important to you are less so to me, and things that I care about a lot doesn’t really register with you. If I can make it work for both then I’m happy to do so, however the number of people who send me film who can even tell me how they exposed and how they’d like it developed is infinitely small, and even fewer of those people actually want prints, so in some respects we’re going to be pretty far apart, as there are times where I break from what you’d expect to see from a zone system technique and do what gives me better results in a digital lab environment. There’s nothing wrong with either way, they serve different purposes.

There's no reason to compromise negatives for perceived scanning 'improvement' - any vaguely competent high end scanner (essentially anything other than a consumer grade flatbed) & any contemporary CMOS sensor is going to have no problems with a darkroom printable negative & can largely even handle those intended for alternative processes. In other words, aim the negatives for a darkroom print & they'll scan excellently.

 

[Bill Burk]

Adrian,

My personal feelings about printing on silver gelatin paper are not key to this discussion. I’m describing a device independent workflow. I use Zone System vocabulary because it’s widely understood.

One thing you could do in a scanning workflow is apply a transfer function. I don’t know why more people don’t talk about that. If you want to have that discussion we would need a new thread on the hybrid side of Photrio.

As long as we stay device independent, there is no limit to what output the target is. I give Vaughn’s example as an extreme on the high end of what you might be asked to develop.

Stephen’s charts are worth pasting into your lab book. Someday someone will hand you a roll of Tri-x they shot at 1250 and it would be nice to know to develop it to 0.75 contrast index.

 

[Adrian Bacon]

There's no reason to compromise negatives for perceived scanning 'improvement' - any vaguely competent high end scanner (essentially anything other than a consumer grade flatbed) & any contemporary CMOS sensor is going to have no problems with a darkroom printable negative & can largely even handle those intended for alternative processes. In other words, aim the negatives for a darkroom print & they'll scan excellently.

This is true most of the time when exposing normally. However, there are times when the person way over exposed or way under exposed. If they way overexposed with a film that can handle a lot of density and I know that ahead of time, I’m going to do what I need to do to keep the density range on the part of the scanner that has the most bits.

I’ll give an example. I recently had a high schooler rent my darkroom to make some prints for a school project. They didn’t have me develop their film (they didn’t know about my services at that time), but instead sent it to a lab down in San Francisco. The negatives they got back (tmax 400), were easily the densest negatives I’d ever seen. I asked the student how they exposed it, and their response was they wanted to make sure they got enough exposure, so they set the camera exposure compensation to +3. The shots where all middle of the day, full daylight. I asked them if the lab they sent it to asked them how they exposed it, and they said that the lab did not.

So, we basically had a set of negatives that the lab ran through at whatever time they had listed for tmax 400. They where basically black.

Had the student asked me to develop the film instead, the first question I would have asked was how was it exposed? Knowing they shot it at EI50 is useful. In this case, they got scans and wanted prints, however 99%+ of the time they want just scans and can’t or won’t tell me how it was exposed, and want good results. 99%+ of the time, they paid ahead of time online, and their roll (or two) shows up in a padded mailer with nothing else and I have to go look up the return address and match it up with a paid order to figure out who’s film it even is. So how do you even handle that film? Zone system stuff is great, but the reality is most of what comes through my lab isn’t exposed correctly, and believe me, I take notes about that sort of stuff. If a client regularly sends me film that is regularly under or over exposed, I’m going to change what I do to make it easier to deliver acceptable scans when handling that film. Am I compromising that negative in doing so? Nope. It’s not exposed correctly to begin with, despite my guidance to the client to adjust exposure. I know that there’s an extremely low likelihood that they’ll ever make a wet print of that negative because people who do that generally develop their own film, and the ones that don’t send their film in with very explicit instructions.

To use Bill’s parlance, if Vaughn sent me a box of film that he cared about, he’d expect to have a conversation with me about how he’d like to have it developed and what developer he wants me to use, and would send the film in with instructions indicating so. The reality is, Vaughn likely would never send me his film, but would either just process it himself in his darkroom, or if he didn’t have one, rent mine for a couple hours to do it, assuming he was local to me.

 

[Adrian Bacon]

I use Zone System vocabulary because it’s widely understood.

Here on Photrio, yes. See post 94. The reality is most film that comes through my door has no instructions and the exposure is all over the place.

Funny you should mention transfer functions. You’d probably find my collection of lookup tables to be pretty interesting.

 

[Bill Burk]

Here’s what I would do if I could introduce transfer functions... I would try to end up with this curve at the print.

If you ever wondered why paper has the strong s-curve it has, aside from that being generally characteristic of silver gelatin, the paper manufacturers aim at this final curve... (Todd-Zakia).

 

[Adrian Bacon]

View attachment 225391

Here’s what I would do if I could introduce transfer functions... I would try to end up with this curve at the print.

If you ever wondered why paper has the strong s-curve it has, aside from that being generally characteristic of silver gelatin, the paper manufacturers aim at this final curve... (Todd-Zakia).

For an output, that’s a good place to end up, though, if the film is digitized, there’s another set of transforms that happens between the negative and that. The Adobe Digital Negative files I deliver are actually an intermediate, not the final product. My transforms go from the film H + D curve to floating point scene referred linear gamma. From there you transform it to the desired output curve, so for example:

Film H+D -> linear gamma -> Printer paper profile -> inkjet print

Or

Film H+D -> linear gamma -> Printer transparency paper profile -> transparency material -> contact print

Or

Film H+D -> linear gamma -> sRGB gamma -> display/jpeg

You get the idea.

 

[Stephen Benskin]

Nice call Bill. I was thinking about showing a couple of tone reproduction diagrams and now I will. Here are two tone reproduction diagrams. One is normal exposure, development, and printed to match the paper LER. The other has the negative two stops over-exposed. The negative picked up a little extra density range as it moved off the toe, but as with short toed curves, it wasn't much. The negative was printed to fit the same paper. The results are a little extra contrast in the shadows, slightly lighter tones through the mid-tone range, and slightly more compression of the higher values than in the highlights.

If the negative processing was reduced, it would have produced a flatter negative which would have needed a higher grade of paper. Now, I have a Howtech Scanmaster 4500 drum scanner, but haven't really used it yet, so I don't have experience scanning through the additional density from over-exposed negatives. I guess this would be a question for a different forum. For here in the analog section, as Loyd Jones said, it's not about density, it's about contrast.

 

[Adrian Bacon]

Now, I have a Howtech Scanmaster 4500 drum scanner, but haven't really used it yet, so I don't have experience scanning through the additional density from over-exposed negatives. I guess this would be a question for a different forum

If it's over exposed, you have to be careful about exceeding the density range that the scanner can handle without introducing posterization artifacts in what will ultimately become the highlights. The problem with over exposed negatives is you still usually have tone values down to film base plus fog, as there's almost always something in the scene that has so little exposure that you'll have tone values down there, so you end up with a negative with a *huge* density range (depending on the film, some films just shoulder off and you end up with all your exposure crunched up in the shoulder of the film). If you know that ahead of time and know what density range give the best results for your equipment, you can pull the development time in so that the higher density areas get more bits.

 

[DREW WILEY]

Bill - I have certainly done more than my fair share of densitometer plotting. And I am comfortable with quite a variety of films and developers. But when I take pictures (vs specialized lab applications of film), I am actually visualizing where metered placement falls on the specific part of a real curve, and not deferring to some generic formula for how to interpolate a line. It takes experience to do this; but for me it works MORE precisely, and allows me to intuitively bend the rules as desired esthetically. I don't want to be futzing around with numbers in the field. A simple spot meter is adequate.