Tabular: terrific or terrible? Your opinions, please.

[Alan Edward Klein]

Definitely not. You don't have to know anything about characteristic curves either, but some of us find them interesting and useful.

Just making a nuisance of myself.

 

[bernard_L]

Do I have to learn calculus to shoot a picture?

Maybe it escaped your attention that I had carefully avoided any equation in my explanation?

 

[Andrew O'Neill]

I was never crazy about TMX 100, but after doing my comparison video (with Panatomic-X), I'm now off the fence. I'll be picking up a few rolls...

 

[Chuck_P]

Do I have to learn calculus to shoot a picture?

There's all levels of interest here, pick one. I have my level of deepness that I go to but I don't go to all of them.

 

[aparat]

@aparat,
I have digitized (with suitable freeware) your curve for Delta100 in Xtol 8min, as shown in your post #210 above. Using basic software LibreOffice Calc) I have plotted

• all digitized points : small blue dots
• points selected by me as representative of the "straight-line portion"; choice did not seem critical or problematic
• a least-squares fit of a linear relation to the above-mentioned "straight-line portion", and the corresponding equation

View attachment 328734
Rounding up by 0.002, the gamma results as 0.70. Not 0.56. And identifying the straight-line portion is not problematic, as one can see at a glance from the near-perfect agreement between the fit line and the red crosses.

I don't know what is (what might be) a statistically derived value of gamma, when there is one data set to work with.

A separate point. In my opinion, CI and G-bar were not created because the straight-line portion of the D-logE curve was difficult to identify. But rather because in actual use for pictorial photography, the scene values are not generally placed entirely in the straight-line portion (except maybe for reproduction work). But rather with some shadow values in the toe; so that for the purpose of translating scene brightness range to density range, CI or G-bar are more useful.

Might sound like I'm nagging at your valuable work. Not so. You have provided lots of excellent and useful data. My remarks aim to help you correct what I perceive as errors.

Thank you for your thorough questions and comments. I appreciate them very much, especially as a learning opportunity! As @Anon Ymous pointed out, Gamma is a tricky parameter to calculate. I am sure you've seen lots of curves where it is not obvious where that straight line is, thus making it subject to interpretation. I decided to use a K-Means Clustering Algorithm that seeks a pattern (clusters) in an unknown data set, basically. Is it perfect? Absolutely not. My program does everything automatically, which is both a good thing and a bad thing. The model of the characteristic curve is often to blame. It works very nicely with most curves, but it fails with "lumpy" ones (especially with multiple local non-linearities) in sometimes producing spurious values. It's rare, but it happens. There will always be photographers who find this unacceptable, and that's perfectly fair. In my program, the user can manually specify which part of the curve to use, so you can just click and drag over the portion you specified, and you will obtain the same result as you did. But then, one opens another can of worms, so to speak. I prefer to not do that myself.

Here's a more detailed plot of that eight-minute curve you mentioned. It shows the Contrast Index calculation (in dotted blue lines) and Gamma (dashed green line). The green line does not coincide with the line you identified, but, as far as the algorithm is concerned, this line represents the "straight-line" portion of the curve. Which one is more correct? That depends on the point of view, unfortunately. I wish this was a simple thing. This is why, I almost always use the CI to estimate contrast. It's just a lot more unambiguous.


There is a lot more to it, and I would be more than happy to discuss it with you, but perhaps a new thread would be needed, or off-line. I am afraid, details like this aren't relevant to this particular thread. The last thing I want is for photographers to be dissuaded from quantitative analysis because it appears way too complex. The idea is to hide a lot of the complexity and present a view of the data that is easy to digest and implement in practice.

 

[aparat]

Just making a nuisance of myself.

No! You are doing the right thing asking a lot of questions. Do not stop! I really enjoy reading your questions because they are insightful. It's a very simple thing that never occurred to me: do I need to learn calculus to understand the characteristic curve? Of course not, and I am very sorry that you got that impression. The idea is to make this kind of analysis as easy to digest as possible. I am not there yet, and maybe I never will, but I appreciate your insight very much.

 

[DREW WILEY]

Alan - you have to learn about calculus if you're a dental assistant. I've made lots of densitometer plots, and other than the fact density is factored logarithmically, I ignore all the math. A simple look at the shape of the curve tells me everything. Of course, there are certain kinds of applications like making color separations where everything one is after has to be very precisely matched or replicated. But with respect to basic black and white film shooting applications, you basically need to differentiate where the toe at the bottom begins to climb the relatively straight line portion, and how far that goes until it shoulders off at the top. All of this is affected not only by film choice, but by amount of exposure and specific development too.

But otherwise, I personally ignore Contrast Index talk, "speed point", whatever kind of A-log-a-rhythms happen to be crackling in the fireplace, blah, blah. That's fine for those who enjoy thinking in that manner. But you don't really need it to visualize what any specific film curve implies, especially when comparing different ones,

 

[Chuck_P]

For me, it is a desire to engage in an effort that is both a combination of what is considered the "blah blah" and also what is considered just differentiating the parts of curve....... which there's absolutley nothing wrong with that. Everyone learns and comprehends in their own way. Go as deep as is comfortable. I personally feel compelled to want to be able to learn more about what Aparat is trying to show but I will not let it get in my way. I guess one has to be always on the lookout for.....analysis paralysis.

 

[MattKing]

Do I have to learn calculus to shoot a picture?

Alan - you have to learn about calculus if you're a dental assistant.

And you definitely need to learn a lot of math - including calculus - and have a lot of computer knowledge in order to become a trade certified auto mechanic.
There are actually some benefits to learning calculus when it comes to photography. The calculus is, among many other things, a method of constructing mathematical analogues of real world phenomena that enable us to examine the characteristics of those phenomena and predict other phenomena.
Sort of like taking a film photo and printing it . Or in Alan's case, showing it on his big TV.

 

[Craig]

There are actually some benefits to learning calculus when it comes to photography. The calculus is, among many other things, a method of constructing mathematical analogues of real world phenomena that enable us to examine the characteristics of those phenomena and predict other phenomena.

For example, taking the derivative of the lines plotted in the H&D charts gives the contrast index at any point on the line - a very useful thing to know.

 

[DREW WILEY]

Well Matt, I've had two close friends who were super math whiz types. One is a famous astrophysicist and the other got bored working at NASA and retired young. Neither can tie their shoelaces correctly.

 

[Sirius Glass]

Well Matt, I've had two close friends who were super math whiz types. One is a famous astrophysicist and the other got bored working at NASA and retired young. Neither can tie their shoelaces correctly.

... because they have always worn loafers as adults. And what is wrong with that?

 

[DREW WILEY]

One of them wanted only me to photograph his daughter's wedding; and it was a big University affair. Like the true Nobel Prize candidate he was, he arrived with his hair uncombed, his tie on backwards, and his shoelaces untied. But to actually win the Nobel Prize you have to show up at Lecture wearing a suit top, shoes, and only boxer shorts in between (pants forgotten). Both Einstein and Melvin Calvin are said to have done that once or twice apiece.

 

[MattKing]

Well Matt, I've had two close friends who were super math whiz types. One is a famous astrophysicist and the other got bored working at NASA and retired young. Neither can tie their shoelaces correctly.

Oh they probably can. It is just that they get distracted with other thoughts before finishing.
My first University degree was a major in physics, with a minor in mathematics. Some of the professors were amazing, brilliant people who communicated their enthusiasm and energized the interests of many students.
And some of my professors would show up to class with their shirt buttons in the wrong button hole.
And those two groups over-lapped.

 

[aparat]

I was going through my backlog of negatives and I found a 120 negative, including a test exposure in my sensitometer. The film is Fujifilm Acros 100, the older version, now discontinued. I decided to run the analysis, and here are the results.

Please, do not put too much stock into this result, as this is only one curve, and the ISO value is computed based on a hypothetical curve. Still, I thought this was a really interesting result, especially compared to the current version.

acros by Nick Mazur, on Flickr

 

[relistan]

I was going through my backlog of negatives and I found a 120 negative, including a test exposure in my sensitometer. The film is Fujifilm Acros 100, the older version, now discontinued. I decided to run the analysis, and here are the results.

Please, do not put too much stock into this result, as this is only one curve, and the ISO value is computed based on a hypothetical curve. Still, I thought this was a really interesting result, especially compared to the current version.

acros by Nick Mazur, on Flickr

That is interesting! I looked on your Flickr stream but didn’t see if you have a curve for Acros II in D-76 1+1. Wonder if the developer is important in the difference.

 

[aparat]

That is interesting! I looked on your Flickr stream but didn’t see if you have a curve for Acros II in D-76 1+1. Wonder if the developer is important in the difference.

I'm still working on it. I am re-testing Acros II, in D-76, because other people got a higher speed in their tests than I did in mine. It's going to be at least a couple of weeks. Acros II is such a popular film, I need to be sure I get it right so it can be a useful test for other photographers.

 

[titrisol]

I was going through my backlog of negatives and I found a 120 negative, including a test exposure in my sensitometer. The film is Fujifilm Acros 100, the older version, now discontinued. I decided to run the analysis, and here are the results.

Please, do not put too much stock into this result, as this is only one curve, and the ISO value is computed based on a hypothetical curve. Still, I thought this was a really interesting result, especially compared to the current version.

the old Acros was a wonderful film, it worked great in Rodinal and DDX

 

[aparat]

the old Acros was a wonderful film, it worked great in Rodinal and DDX

Yeah, I remember it fondly. I had one five-pack left a few months ago. I did some pinhole photography with it. It was an amazing film. Unfortunately, I didn't run a comprehensive test with it while I still had it. The new version is also very good. It's a little different, in my experience, but one can still make amazing photographs with it.

 

[relistan]

I'm still working on it. I am re-testing Acros II, in D-76, because other people got a higher speed in their tests than I did in mine. It's going to be at least a couple of weeks. Acros II is such a popular film, I need to be sure I get it right so it can be a useful test for other photographers.

No worries, no pressure from me! Appreciate everything you have done already.

 

[BHuij]

Only time I ever played with Acros II was to try it in a pinhole, where its low reciprocity failure should have been a big advantage. The results came out fine, I didn't see anything special about the film (it was a roll of 120 in a curved-plane 6x17 camera). Of course, a random pinhole isn't exactly a fair shake for that film to shine. I know it has a large fanbase.

I guess I'm just happy enough with Delta 100 that I can comfortably sit up on my high horse and continue my low-key boycott of Fuji without worrying too much about what I might be missing out on with Acros II

 

[aparat]

Only time I ever played with Acros II was to try it in a pinhole, where its low reciprocity failure should have been a big advantage. The results came out fine, I didn't see anything special about the film (it was a roll of 120 in a curved-plane 6x17 camera). Of course, a random pinhole isn't exactly a fair shake for that film to shine. I know it has a large fanbase.

I guess I'm just happy enough with Delta 100 that I can comfortably sit up on my high horse and continue my low-key boycott of Fuji without worrying too much about what I might be missing out on with Acros II

Cool. I enjoy Delta 100 a lot, too. Speaking of pinholes, I used my last supply of Acros on pinhole photographs. In terms of exposure, they came of just as predicted, with virtually no reciprocity issues. I'm glad to hear that Acros II has the same property.

 

[bluechromis]

One other thing I want to mention is that a hypo-clearing agent can reduce the amount of the pinkish cast from the film base pretty effectively, in my experience. In fact, after washing the tabular grain films, my Kodak Hypo Clearing Agent has turned a nice shade of pink.

I found an interesting quote in Woods (Zone System Craftbook, 1993) regarding pushing and pulling t-grain films: "This variable zone placement must be used with care and only when the T-Max developers are used. When some conventional developers are used, the extreme high values produced by overexposure will "shoulder" quickly," and later "I again stress the need to use the special developers designed for this film to achieve the maximum benefits from it."

I wonder if statements such as this bear out in your experience. In my tests, the Kodak t-grain films do not shoulder particularly easily with XTOL or D-76, but perhaps they do in your experience? Do we really need to use the T-MAX Developer with tabular grain films? Or was that just a marketing ploy to make sure we buy those special developers?

I think the idea that Tmax films are ideally suited to Tmax developer can be overstated. I saw a recent review of a Tmax film that said that since Tmax films and developer both have the same name, they must have been designed to suit each other. The reviewer said that Tmax films are best developed in Tmax developer for this reason. Xtol did not exist when Tmax films first came out. My understanding is that now Kodak says that Xtol is the best developer for Tmax films.

Kodak has a history of using similar brand names for entirely unrelated products. Think of all of the products that have Ekta or Ektar in the name. I wonder if Kodak was thinking that Tmax was their hot new brand name, and they wanted to slap it on as many products as possible.

According to Kodak's developer comparison chart, Tmax developer has the best shadow detail (film speed), but only by hair over Xtol. But in grain size, it was the worst. Maybe there was a rationale that if the initial versions of Tmax films were touchy about exposure, some extra speed would allow a greater margin of error when shooting at box speed. Since Tmax films were fine-grained, maybe it would be okay to lose something in grain size.

But that rationale only goes so far. I have not used Tmax developer, but my impression is that it is a speed-increasing developer in a convenient liquid form. So, it may be roughly similar to other speed-increasing developers like Ilford Microphen. So one might use Tmax developer for the same reason they would use other speed-increasing developers, but it not may not be categorically different with Tmax films.

 

[bluechromis]

Nicholas - Tech Pan was pathetic at trying to mimic 4X5 results. Not only was it incapable of handling highlights and shadows anywhere near as well as ordinary film, even using special developers, but if enlarged directly from 35mm much, exhibited enlarged little dots in the sky too. I had a backpacking pal addicted to it, who used very expensive Zeiss 6x6 lenses. Sharp, yup, but with that proverbial "soot and chalk" look which AA derided, and lots of tiny sky blemishes. He eventually got frustrated and switched to FP4, and while it was still availlable used Efke 25 roll film on my recommendation, which gave him something super-fine grained, but without TP's headaches.

The edge effect of Tech Pan was abominably mediocre. So yeah, doing the very best that could be done using 35mm TP, and comparing the resultant print side by side to the worst that can be done using actual 4X5, by some klutz with a warped film back, dumpy lens, and the grainiest film he can find, devoid of a focus loupe, and that old BS Kodak ad sorta, vaguely, remotely rings true. But I've got 8x10 Tech Pan film on hand - so that must be equivalent to traditional film ten feet across if you want to believe Kodak's old marketing ploy!

My biggest complaint with TMX100 is its own disappointing edge effect despite very fine grain and otherwise exceptional detail capacity over a very long contrast scale. That's why I develop small sizes of it (including 120 roll film shots) using Perceptol 1:3, which gives me a very nice boost in edge effect, yet only a minimal increase in graininess. TMY400 has very good edge effect, so in its case, all kinds of developers work well, though I lean toward PMK pyro as my standard option. Sometimes, like for portraiture, I actually prefer to keep the softer-edged look of TMX100 as-is. Note that this has nothing to do with overall detail capacity or the reproduction of subtle microtonality, but just the Mackie line edge effect itself.

Geoffry Crawley had the view that TMX 100 was so fine-grained and the edge boundaries were so smooth that viewers perceived it as less sharp than it really was. There may be some reason for thinking that. Perceived sharpness is based upon both resolution and boundary contrast. It may be the viewers are sensitive to boundary (edge) effects more than resolution. Crawley developed FX-37 and FX-39 to meet the needs of tab grain films, though they can be used with traditional films. They do not produce fine grain, but Crawley felt that they brought out the best characteristics in tab grain films.

 

[MattKing]

Perceived sharpness is based upon both resolution and boundary contrast.

Actually, perceived sharpness is mostly influence by edge contrast (essentially acutance), followed by overall contrast, followed distantly by resolution.
Which is why grainy films like Tri-X often are perceived to be "sharper" than very fine grain films like T-Max 100 or, in previous times, Panatomic-X.
And T-Max developer shares the name with the T-Max films not because they were designed exclusively for each other, but they were introduced as part of the same program. The developers offer/offered full speed in an efficient and easy to use liquid packaging. As such, they have/had many of the advantages in a commercial lab setting that HC-110 offered, without imposing the speed penalty.
So yes, the developer and the films were/are well suited to each other, but the films don't require the developer.